----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 30 Aug 97 11:37 #2 By: Ward Dossche To: All Re: An IP-conference St: ------------------------------------------------------------------------------ Dear all, The feeling in my fingers tells me we are going to have IP introduced in the Fidonet nodelist. Now there will be at least 3 parallel discussions going on (REGCON, REGCON.EUR, ENET.SYSOP) in Europe with others probable. I think it is logical to have only one discussion-area and deflect all IP-related traffic to it. It would benefit productivity and signal/noise-ratio I think. After all there is little use for others and myself scanning and crossposting at least 3 conferences to deal with a single subject. Therefor as of immediately I am offering a conference called IP_CONNECT to discuss and deal with IP-connectivity. (If someone has a better suggestion please step forward) Take care, \x/ard --- DB 1.58/001877 * Origin: The pianoplayer (2:292/854) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 876 335/533 341/70 362/21 SEEN-BY: 380/106 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 31 Aug 97 21:19 #3 By: Ward Dossche To: Pedro Lima Re: IP-connectivity St: ------------------------------------------------------------------------------ Yow, > I'd prefer to have the straight IP address in the phone field (to be > more general, the "natural" address of the technology being used) and > that differentiation is carried out via flags. I dissagree to proceed like that! The whole idea of preceding the IP-address by an inexistant country-code is to avoid that mailers will dial numbers in another technology-realm. Playing with flags does not work. You don't have to believe me, I can forward your disbelief to a number of sysops operating ISDN-lines properly flagged which are receiving analog-calls around the clock anyway. By using inexistant country-codes at the same time we avoid an unending discussion about not being able to dial such a node during ZMH. I don't see why you would advocate to leave the possibility for mistakes open when we can close the loophole altogether. > A phone number would then > be only one kind of address (which is what it is, after all), and > capability differences have been traditionally expressed with flags. And traditionally with the majority of people this doesn't work. > We're also allowing > ISDN-only nodes, and this seems not to be the intention with IP nodes. It definitely is my intention to allow IP-only nodes. Take care, \x/ard --- DB 1.58/001877 * Origin: R40 missing from nodelist - freq "REGION40" (2:292/854) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 876 335/533 341/70 362/21 SEEN-BY: 380/106 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 31 Aug 97 14:55 #4 By: Andrew Mowatt To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ To: ward.dossche@medelec.uia.ac.be From: Andrew_Mowatt@nemesis.ie (Andrew Mowatt) In a message dated Sat 30 Aug 97 18:20, Ward Dossche It seems to me that it would be completely wrong to not embrace IP- WD> connectivity Agreed. WD> Life can be much simpler than creating a separate zone/region/net. WD> What Denis proposes is to put the IP-number in the phone-number-field WD> but have WD> it preceded by a fictitious non-existant country-code. Sounds good. WD> can be used for other future technologies as well and, why not, ISDN Yes. WD> Comments please? You get my vote! -- Via DLG Pro v1.16 /\ndy - andy@nemesis.ie CBR600F3 fidonet: 2:263/150 BBS: +353-1-6211360 All opinions I express are MY OWN and not necessarily those of anyone else! --- DB 1.58/001877 * Origin: R40 missing from nodelist - freq "REGION40" (2:292/854) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 876 335/533 341/70 362/21 SEEN-BY: 380/106 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 3 Sep 97 8:19 #5 By: Mats Wallin To: Pedro Lima Re: IP-connectivity St: ------------------------------------------------------------------------------ * Message copied from "Regional Coordinators [Intl]" PL> Perhaps we could go for the time being with IP as an PL> extension for PSTN nodes and get the nodelist scenario solved PL> first? It's of course necessary to solve the nodelist problem, but it is IMHO more important to make it possible for IP-only nodes to be able to join FidoNet, then it is to make it possible for existing nodes to have both IP and phone connections. Those people are already in FidoNet, it is the others that wants to join FidoNet that we should try not to lose. Mats mw@defsol.se --- * Origin: Definite Solutions, Stockholm, Sweden (2:201/329.11) SEEN-BY: 201/300 329 330 203/614 212/8 254/60 280/100 292/854 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 3 Sep 97 10:54 #6 By: Joaquim Homrighausen To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ wd> By using inexistant country-codes at the same time we avoid an wd> unending discussion about not being able to dial such a node wd> during ZMH. What's wrong with using a simple prefix character.. (e.g. "#198.32.3.112" or "#198-32-3-112")? Or does that break MakeNL/other nodelist tools? wd> It definitely is my intention to allow IP-only nodes. Excellent! %JoHo% joho@defsol.se * Origin: Definite Solutions ~/Stockholm, Sweden (2:201/330.1) SEEN-BY: 201/300 329 330 203/614 212/8 254/60 280/100 292/854 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 4 Sep 97 8:27 #7 By: Ward Dossche To: Joaquim Homrighausen Re: IP-connectivity St: ------------------------------------------------------------------------------ Joaquim, > What's wrong with using a simple prefix character.. (e.g. > "#198.32.3.112" or "#198-32-3-112")? Or does that break MakeNL/other > nodelist tools? Maybe, if I find the time before tomorrow, I will include a few entries in the nodelist to enable people to test their software. I tend to look upon this as a technical issue, and that's the way to get it quickly solved. There will be a political debate too, but I'm not very good at that. \x/ard --- DB 1.58/001877 * Origin: R40 missing from nodelist - freq "REGION40" (2:292/854) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 876 335/533 341/70 362/21 SEEN-BY: 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 4 Sep 97 11:52 #8 By: Joaquim Homrighausen To: Mats Wallin Re: IP-connectivity St: ------------------------------------------------------------------------------ * Reply to message in "z.REGCON" * On Thu 4 Sep 97 8.02, wrote: JS>> I repeat my question: is there any existing IP-based mailer JS>> that uses the Version 7 compiled nodelist? mw> Every mailer that uses a FOSSIL can use VModem for it's mw> IP-communication. [..] Uh, could we move this to IP_CONNECT .. ? %JoHo% joho@defsol.se * Origin: Definite Solutions ~/Stockholm, Sweden (2:201/330.1) SEEN-BY: 201/300 329 330 203/614 212/8 254/60 280/100 292/854 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 3 Sep 97 18:21 #9 By: Jan Ceuleers To: All Re: Quiet St: ------------------------------------------------------------------------------ It's awfully quiet in here. Looks like the discussion is continuing in several places instead of being concentrated here as per Ward's suggestion. Jan --- timEd-A10a * Origin: Experimenter Board, Antwerp, Belgium (2:292/857) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 857 876 335/533 341/70 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 3 Sep 97 18:23 #10 By: Jan Ceuleers To: All Re: IP-access St: ------------------------------------------------------------------------------ --> Note: Forwarded (from: ENET.SYSOP) by Jan Ceuleers using timEd. Originally from Jan Ceuleers (2:292/857.0) to Bill Birrell. Original dated: Sep 03 '97, 17:55 I quote Bill Birrell: > Permanent IP connectivity though cable modems is being > introduced here. The price is under 9000 BEF per > annum. We'll have to see what the throughput will > actually be (the marketing types obviously promise > heaven) but throuhput isn't of prime importance if you > don't have a metered connection anyway. BB> Only throughput could justify the exorbitant prices for BB> a leased line in UK. I take it that yu do mean leased lines, BB> Jan, with IP virtual modems? No, I mean cable modems. Modems that make use of a dedicated frequency range on the coaxial cable TV network (that is upgraded to bidirectional operation for the purpose). The reason why I mentioned throughput as an issue is that all subscribers attached to a particular segment of the coax cable have to share the bandwidth. The more people simultaneously surfing, the slower each user's responses and transfers will be. Belgium is one of the world's most densely coax cabled countries. Using this network for providing services other than cable TV is therefore quite a natural thing to do. In countries or areas where cable TV isn't as commonplace this makes less sense. An alternative is ADSL which is also being introduced in various parts of the world but I haven't got a price reference for that. Jan ___ timEd-A10a - Origin: Experimenter Board, Antwerp, Belgium (2:292/857) --- timEd-A10a * Origin: Experimenter Board, Antwerp, Belgium (2:292/857) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 857 876 335/533 341/70 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 3 Sep 97 18:23 #11 By: Jan Ceuleers To: All Re: IP-connectivity St: ------------------------------------------------------------------------------ --> Note: Forwarded (from: ENET.SYSOP) by Jan Ceuleers using timEd. Originally from Jan Ceuleers (2:292/857.0) to Frank Ellermann. Original dated: Sep 03 '97, 18:06 I quote Frank Ellermann: FE> Here too, several weeks (or even months ? ;-) in NODES.024 FE> last year... "Country code" 0 is probably not the best FE> solution. Of course it will never exist, but mailers or FE> nodelist compilers could mistake it as legal dial code... FE> Not funny, if someone tries to dial 00-nnn-... (instead of FE> IP 0-nnn.nnn.nnn.nnn) for example. If you insist on a FE> zero-based-prefix, how about 000 ? That can't happen. The nodelist currently does not contain any prefixes; the telephone numbers start with the country code. All nodes in the world, if they're to be able to even poll their regular uplink, must have correctly set up their system to do things like recognize their own country code, recognize their own local calling area, prepend long distance prefixes and international access codes as necessary etc. In Belgium (and most of Europe) the international access code is 00. Country code 0 will therefore be expanded to 000 which isn't dialable. Any modern local switch will reject such calls, without those calls even making it to the international gateway exchange. In any event, the international gateway exchange would reject them as well. Anyway, there is no way any of us can protect anyone from their own stupidity. JC> We should check with the ITU which fake country code to use. FE> Infos posted in NET_DEV by Leonard Erickson a few weeks ago: (...) FE> The only sure (?) gaps are 882-, 883-, 884-, 885-, 887-, FE> 888- and 889- in this numbering scheme... That's not good enough. I take it you don't subscribe to the ITU's Operational Bulletin? (Neither do I, but a subscription to the Telecom Digest on Usenet is a good substitute if you're not actually responsible for the maintenance of number translators of a telephone switch or network). New countries are being created all the time. Changes in the political situation of a country or area are frequently reflected shortly thereafter by a request from the new government to the ITU for a country code. There are many recent examples in Eastern Europe, but also in the Far East. It's therefore not good enough to just use a country code that isn't currently being used, because you can't be sure that it won't be allocated by the ITU to some obscure island tomorrow. FE> and of course FE> anything FE> starting with 0- That's the only-one that isn't likely to ever be used. I'll need to look at the ITU recommendations on the numbering plan (probably the annex to E.164) to see what this is flagged as. Jan ___ timEd-A10a - Origin: Experimenter Board, Antwerp, Belgium (2:292/857) --- timEd-A10a * Origin: Experimenter Board, Antwerp, Belgium (2:292/857) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 857 876 335/533 341/70 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 4 Sep 97 17:52 #12 By: Jan Ceuleers To: All Re: IP-connectivity St: ------------------------------------------------------------------------------ --> Note: Forwarded (from: REGCON) by Jan Ceuleers using timEd. Originally from Jan Ceuleers (2:292/857.0) to Jerry Schwartz. Original dated: Sep 04 '97, 07:11 I quote Jerry Schwartz: JC> We should check with the ITU which fake country code to use. It should JC> be a JC> country code that does not now and will never in the future exist. One JC> way JC> to ensure that would perhaps be to just use a 0 as the first digit in JC> order JC> to avoid collisions. The only thing that is likely to be used for is as JC> the JC> prefix for interplanetary telephone calls ;-) JS> In this country, it is not uncommon to use 0-xxxx to select JS> your long distance carrier for a particular call. If I had JS> a specific carrier that I wanted to use for international JS> calls, for example, I might have my nodelist compiler stick JS> 01339 in front of all "other-country" calls. I might even So what you actually dial is 011 (the international prefix) 01339 (carrier selection code) 32 (country code, in this case Belgium) and the rest of the number? No you don't. The carrier selection prefix is actually 101339 (not 01339), isn't it? Please keep in mind that in North America the "1" prefix is just a prefix for long distance dialing, not the prefix for international dialing. It just happens to be the same as the country code. You don't normally call my BBS through the default long-distance or international carrier by just dialing 32-3-320-9129 do you? This number needs to be prepended with the international access code (011). Which in turn needs to be prepended with the carrier selection code (in your case 101339). Right? So, to get back to my proposal: are there really any circumstances under which you need to dial a 0 after the 011 prefix? Even if the answer to that is yes, then the meaning of that 0 is local to North America. I'm not saying that we therefore needn't care, but it means that this 0 can't be a part of the actual telephone number (if it was it would have to have worldwide significance). That is: it can't be playing a role in _identifying_ the called party. Perhaps it is used in routing (carrier selection) or for switching off echo cancellers (feature selection) or something of that nature. None of those kinds of thing need to be listed in the nodelist because if any of them are necessary for actually dialing that node from your particular area, then it is your responsibility to add those codes in your own setup. Those codes would very likely be different in other parts of the world anyway. JS> Also, 0 is used for operator-assisted calls as well. Aha, so 011-0 get you the international operator? Is that right away or after a timeout? JS> 0 is not a good choice here. You are confusing the local numbering plan (the NANP) with the world-wide numbering plan. There is no collision between the two; the "0" country code is not in use. For more information on the North-American numbering plan, check out http://www.bellcore.com/NANP Jan ___ timEd-A10a - Origin: Experimenter Board, Antwerp, Belgium (2:292/857) --- timEd-A10a * Origin: Experimenter Board, Antwerp, Belgium (2:292/857) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 857 876 335/533 341/70 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 4 Sep 97 18:15 #13 By: Jan Ceuleers To: All Re: IP-connectivity St: ------------------------------------------------------------------------------ --> Note: Forwarded (from: REGCON.EUR) by Jan Ceuleers using timEd. Originally from Jan Ceuleers (2:292/857.0) to Wim Van Sebroeck. Original dated: Sep 04 '97, 18:00 I quote Wim Van Sebroeck: > At the time, I proposed at least the following address spaces: IPv4 (the > current 32-bit IP addressing system), IPv6 (the next-generation IP > addressing > system), X.121 (the addressing system used by X.25) and perhaps others. > (I've > been searching through my archives but couldn't find the earlier > correspondence > I had about this). WVS> Nice idea. But what whit nodes that have a pstn or isdn WVS> number and TCP/IP connectivity? Those nodes will need extra node numbers. Unfortunately, the current nodelist format does not realistically permit a node to advertise all of its capabilities if it has a multitude of them. Not even in the case modem/ISDN. It may be that a node has a modem on-line at phone number x, and an ISDN device (which does not have a modem) at phone number y. Even if these two devices can be reached at the same phone number, and hence only one node number is listed, then there may be uncertainties as to whether a particular capability flag refers to the modem, to the ISDN device or to both (V42B comes to mind). The most elegant solution would indeed be something like what Steve Woodmore proposed, but that requires a change in all software that generates or uses the nodelist. Steve proposed the addition of "Extra," lines which would indicate additional capabilities or co-ordinates of the last non-Extra line preceding them. The main merit of this solution is that only one node number is needed, so that a node is unequivocally identified by the single node number, regardless of the connectivity method used to reach him. The (in my view) most important disadvantage is that it will no longer be possible to select a technology through which to connect to a multi-technology node by selecting the corresponding AKA of that node. Also, there would not be much difference in the resulting size of the nodelist between Steve's proposal and simply handing out extra node numbers. Jan ___ timEd-A10a - Origin: Experimenter Board, Antwerp, Belgium (2:292/857) --- timEd-A10a * Origin: Experimenter Board, Antwerp, Belgium (2:292/857) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 857 876 335/533 341/70 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 4 Sep 97 18:20 #14 By: Jan Ceuleers To: All Re: IP-access St: ------------------------------------------------------------------------------ --> Note: Forwarded (from: ENET.SYSOP) by Jan Ceuleers using timEd. Originally from Jan Ceuleers (2:292/857.0) to Paul Walker. Original dated: Sep 04 '97, 18:16 I quote Paul Walker: mg> for me, the only smart way is adding flags that would carry ip number mg> and the port number/inet mailer type. we cannot provide ip address PW> That's not practical - a lot of the time I don't get the PW> same IP address, it varies. Then how do you realize incoming connections now? How do your correspondents get hold of your current IP address (since it may vary from one session to the next)? Unless you have been assigned a DNS entry with an incredibly short time-to-live parameter, I can't see how dynamic IP address allocation can be made compatible with the notion of incoming connections. What does Wim think about this? Jan ___ timEd-A10a - Origin: Experimenter Board, Antwerp, Belgium (2:292/857) --- timEd-A10a * Origin: Experimenter Board, Antwerp, Belgium (2:292/857) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 857 876 335/533 341/70 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 5 Sep 97 10:25 #15 By: Anton Kuznetsov To: Ward Dossche Re: Q. St: ------------------------------------------------------------------------------ Buenas dias, amigo Ward Dossche ! Who is allowed to write to this conference? Anthony. Donetsk, Ukraine --- E-mail: tony@tony.donetsk.ua * Origin: Everything you say may be used against you. (2:465/50) SEEN-BY: 50/0 201/329 212/8 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 450/97 460/111 461/35 463/94 690 465/11 SEEN-BY: 465/50 70 85 87 467/67 480/33 2432/200 4625/1 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 5 Sep 97 10:35 #16 By: Joaquim Homrighausen To: Ward Dossche Re: Using a character-prefix St: ------------------------------------------------------------------------------ wd> Maybe, if I find the time before tomorrow, I will include a few wd> entries in the nodelist to enable people to test their software. Great. wd> I tend to look upon this as a technical issue, and that's the way wd> to get it quickly solved. As most things are.. :-) wd> There will be a political debate too, but I'm not very good at wd> that. Who is ? :-) %JoHo% joho@defsol.se * Origin: who dares wins (2:201/330.1) SEEN-BY: 201/300 329 330 203/614 212/8 254/60 280/100 292/854 857 SEEN-BY: 292/876 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 6 Sep 97 4:14 #17 By: Pedro Lima To: Mats Wallin Re: IP-connectivity St: ------------------------------------------------------------------------------ Hi, MW> It's of course necessary to solve the nodelist problem, but it is IMHO MW> more important to make it possible for IP-only nodes to be able to MW> join FidoNet, then it is to make it possible for existing nodes to MW> have both IP and phone connections. Those people are already in MW> FidoNet, it is the others that wants to join FidoNet that we should MW> try not to lose. If we solve the nodelist problem of IP-only nodes i'm sure we're also solving the problem of nodes having both phone and IP connections. Thing is, the problem of IP-only nodes is much wider than the format of the nodelist entry. How does an IP-only system request for a node? How is netmail traffic routed to that node? And so on... All this seems to me much more complicated than agreeing on the nodelist entry format. Besides, it's not useless work even when considering IP-only nodes. So I'd suggest we focused (without loosing sight of the rest, however) on this particular problem. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 857 876 335/533 341/70 SEEN-BY: 361/7 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 7 Sep 97 9:40 #18 By: Ward Dossche To: Jan Ceuleers Re: Quiet St: ------------------------------------------------------------------------------ > It's awfully quiet in here. Looks like the discussion is continuing in > several places instead of being concentrated here as per Ward's > suggestion. Give or take 2 weeks for people to become aware and get this stuff distributed. Just look at seen-by and you see where it goes for the moment. We're equally attempting distribution to other zones. Z1 will be OK soon, the rest will then follow ... if they care. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 857 876 335/533 341/70 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 7 Sep 97 9:40 #19 By: Ward Dossche To: Anton Kuznetsov Re: Q. St: ------------------------------------------------------------------------------ Anton, > Who is allowed to write to this conference? I would say anyone who has a distinct contribution to make. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 857 876 335/533 341/70 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 7 Sep 97 12:06 #20 By: Ward Dossche To: Joaquim Homrighausen Re: Using a character-prefix St: ------------------------------------------------------------------------------ > wd> There will be a political debate too, but I'm not very good at > wd> that. > Who is ? :-) Most of whom which haven't anything else significant to contribute. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 857 876 335/533 341/70 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 8 Sep 97 8:51 #22 By: Mats Wallin To: Pedro Lima Re: IP-connectivity St: ------------------------------------------------------------------------------ PL> Thing is, the problem of IP-only nodes is much wider than the PL> format of the nodelist entry. How does an IP-only system PL> request for a node? How is netmail traffic routed to that PL> node? And so on... None of these questions should be very hard to answer, IMHO. One requirement that has to be met, if we should be able to accept IP-only nodes, is that they can pickup their mail from another FidoNet system. This means that there somewhere in the region has to be a system with both IP and modem access. What I would propose is that a system that has both IP and modem access is setup as the Host of either a new net, or an existing net (existing net is probably only possible if the existing host already have IP access). Alternatively, the system with both IP and modem access could of course be a hub instead. I think this should solve all your questions. The request for a nodenumber should of course be sent to this host or hub that has IP access, just any other modem-only node sends their nodenumber request to a host or hub. Netmail traffic will be routed via the host/hub system, just as most other netmail is routed, especially netmails to systems that you can't call dir directly. So I really don't see that part as a problem at all. Mats mw@defsol.se --- * Origin: Definite Solutions, Stockholm, Sweden (2:201/329.11) SEEN-BY: 201/300 329 330 203/614 212/8 254/60 280/100 292/854 857 SEEN-BY: 292/876 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 8 Sep 97 8:55 #23 By: Mats Wallin To: Ward Dossche Re: Test of an IP-number in the nodelist St: ------------------------------------------------------------------------------ WD> ,3000,IP-test,somewhere_unimportant,Lothar_Behet,0000-194.231 WD> .142.17,300,U,IP As far as I can tell, MakeNL does not accept that as a phonenumber. (I just tried it here, and it complained about it). I think the "phonenumber" will have to be written as 0000-194-231-142-17, which of course will require some type of translation before it can be used. Or we will need to replace MakeNL with another program. Mats mw@defsol.se --- * Origin: Definite APXw (2:201/329.11) SEEN-BY: 201/300 329 330 203/614 212/8 254/60 280/100 292/854 857 SEEN-BY: 292/876 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 8 Sep 97 9:01 #24 By: Mats Wallin To: Jerry Schwartz Re: IP-connectivity St: ------------------------------------------------------------------------------ * Message copied from "Regional Coordinators [Intl]" JS> That rather presumes something about your operating system, JS> doesn't it? Is VModem available for Mac OS7? Windows 95? I didn't know that there were a FOSSIL for Mac OS, but if there is, I guess I were wrong. But I think there is a VModem implementation for Windows 95. There is one for NT, and there is one for OS/2. However, what I did wrote was that any mailer that used a FOSSIL could use VModem. I might be wrong about Mac OS, but with that exception, what I wrote is correct, because any DOS based mailer (which are the only ones I knew used FOSSILs), can be used under OS/2, and can therefor use VModem. Mats mw@defsol.se --- * Origin: Stockholm, Sweden (2:201/329.11) SEEN-BY: 201/300 329 330 203/614 212/8 254/60 280/100 292/854 857 SEEN-BY: 292/876 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 8 Sep 97 9:12 #25 By: Mats Wallin To: Lothar Behet Re: IP-access St: ------------------------------------------------------------------------------ * Message copied from "SysOps [Z2]" LB> Concerning to the fact, that most of the IP-capable nodes LB> only have a dialup-connection to the internet, this makes LB> sense. In this case they should notice their connct times via LB> the T-Userflag, which is presumably quite different from LB> their normal online times. IMHO, a node that only have a dialup-connection to the internet should not be listed in the nodelist. To be listed in the nodelist as an IP-node, it is necessary to have a static IP-address that can be specified in the nodelist. In most cases, you do not have a static IP-address when you use a dialup connection to your ISP. LB> How do you handle IP-nodes with a variable IP-address (most LB> dialup-connections will be assigend an address from that pool LB> by the provider)? Supporting DNS might be a solution to this. I don't think DNS would help in that case, since DNS is used to point to a static IP address (as far as I know anyway). I doubt you can create a new DNS name each time you connect to the Internet, and have it to point to a new IP-address. And remember, that DNS name would have to be spread all over the world at once, which you hardly can do. Mats mw@defsol.se --- * Origin: Definite Solutions, Stockholm, Sweden (2:201/329.11) SEEN-BY: 201/300 329 330 203/614 212/8 254/60 280/100 292/854 857 SEEN-BY: 292/876 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 8 Sep 97 8:58 #26 By: Joaquim Homrighausen To: Ward Dossche Re: Using a character-prefix St: ------------------------------------------------------------------------------ wd> Most of whom which haven't anything else significant to wd> contribute. Ah.. yes.. "them" :-) %JoHo% joho@defsol.se * Origin: Definite Solutions ~/Stockholm, Sweden (2:201/330.1) SEEN-BY: 201/300 329 330 203/614 212/8 254/60 280/100 292/854 857 SEEN-BY: 292/876 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 8 Sep 97 7:15 #27 By: Jan Ceuleers To: All Re: IP-access St: ------------------------------------------------------------------------------ --> Note: Forwarded (from: ENET.SYSOP) by Jan Ceuleers using timEd. Originally from Jan Ceuleers (2:292/857.0) to Alto Speckhardt. Original dated: Sep 08 '97, 07:08 I quote Alto Speckhardt: AS> I understood you said there would already be such a AS> nameserver for zone two, is that right? If so, what would AS> that server be and isn't one single nameserver a bit AS> overtaxed if the whole region asks him for DNS-entries? Quite obviously, the name server for zone 2 is z2.fidonet.org. As to it being over-taxed, that needn't be the case. It's not necessary to query this name server each and every time you need the IP address of a particular node: queries are cached in various name servers (such as that of your ISP or even your own name server if you have one) along the way. Also: it's quite possible to delegate the role of name server for sub-domains. In this case, a sub-domain is a net or a series of nets. My main problem is that due to the caching function I mentioned above, it doesn't seem to be possible to accomodate nodes whose ISP dynamically assigns them an IP address as they log on. I don't think DNS can cope with that, but I'd be happy to be proved wrong. Jan ___ timEd-A10a - Origin: Experimenter Board, Antwerp, Belgium (2:292/857) --- timEd-A10a * Origin: Experimenter Board, Antwerp, Belgium (2:292/857) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 857 876 335/533 341/70 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 9 Sep 97 0:31 #28 By: Pedro Lima To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ Hi, WD> However, should it bother an NC if an IP-only node cannot apply for a WD> node-number? I think it's the node's problem to figure thatone out ... WD> it's the same now too. What I'm saying is it should be contemplated in the set of procedures. Like ISDN nodes, IP nodes don't require a modem to operate. Thing is, ISDN works over the PSTN, like modems do, and that fact simplified, IMHO, the whole issue of ISDN nodes and their inclusion in FidoNet. This may not be the case with IP nodes. I'll say it in a different way. We need to realize that we're being forced to admit certain situations that are breaching Policy. That almost certainly happens in Nets where the host doesn't have ISDN connectivity, although he addmitted ISDN-only nodes in the net. I'm sure the same will happen with IP-only nodes. I'm not worried about breaching Policy where a clear need exists for it, what I'm saying is that Policy has to be changed and that this gets more urgent by the minute. Unless we address the problem ASAP, we'll soon be at a complete mess, and even if we end up doing little or nothing about it for the time being, I think it's important that the problem isn't ignored and that it gets really discussed (not Flame Exchange Inc.), or we'll end up loosing perspective on what we're doing. WD> I suspect that when IP-only nodes will spring-up they will deal with WD> that themselves, i.e. create an IP-hub of which at least the WD> hub-system has both PSTN- and IP-access. I agree. However, obtaining an IP address is generally more difficult than getting an ISDN line, moreover if we think that the IP address space is nearly exausted. WD> I think most of the problem is just technical and psychological. For now, I'd agree, but we must have the eyes on the future, and stop patching things up as urgent needs arise. I'm not saying we should predict everything, but there are things which we can safely "predict", and it's easier to build "from scratch" than "patching up". So perhaps it should be in order to create a document with some procedures to help prospective non-modem nodes to join up, even if it cannot be stated as an official document. WD> Let's just face it ... most evolutions in Fido are technology-driven Actually, I wouldn't say that... The need for evolution is social. The satisfaction of that need is usually technical. WD> ... and, yes, P4 is a pain in the ass. :-) Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 9 Sep 97 11:52 #29 By: Ward Dossche To: Mats Wallin Re: IP-connectivity St: ------------------------------------------------------------------------------ Mats, > What I would propose is that a system that has both IP and modem access > is setup as the Host of either a new net, or an existing net (existing > net is probably only possible if the existing host already have IP > access). Alternatively, the system with both IP and modem access could of > course be a hub instead. I suppose this is something which will solve itselve, how else can they be included in the nodelist and/or participate in Fidonet. > So I really don't see that part as a problem at all. Me neither ... this is just something technical again which, in this particular case, is quite easy. \x/ard --- DB 1.58/001874 * Origin: His master's voice (2:292/854.1) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 9 Sep 97 11:54 #30 By: Ward Dossche To: Mats Wallin Re: Test of an IP-number in the nodelist St: ------------------------------------------------------------------------------ > WD> ,3000,IP-test,somewhere_unimportant,Lothar_Behet,0000-194.231 > WD> .142.17,300,U,IP > As far as I can tell, MakeNL does not accept that as a phonenumber. (I > just tried it here, and it complained about it). Thank you, hadn't gone that far yet. I did try compilations with several products and none of those did complain. > Or we will need to replace MakeNL with another program. That is a totally different matter, but quite appropriate. No updates on MAKENL have been forthcoming for about 3,5 years I think. \x/ard --- DB 1.58/001874 * Origin: His master's voice (2:292/854.1) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 9 Sep 97 11:56 #31 By: Ward Dossche To: Mats Wallin Re: IP-access St: ------------------------------------------------------------------------------ Mats, > IMHO, a node that only have a dialup-connection to the internet should > not be listed in the nodelist. That is indeed logic itself and the whole point of attempting this exercise. \x/ard --- DB 1.58/001874 * Origin: His master's voice (2:292/854.1) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 9 Sep 97 7:15 #32 By: Jan Ceuleers To: All Re: IP-access St: ------------------------------------------------------------------------------ --> Note: Forwarded (from: REGCON.EUR) by Jan Ceuleers using timEd. Originally from Jan Ceuleers (2:292/857.0) to Jan Vermeulen. Original dated: Sep 09 '97, 07:09 I quote Jan Vermeulen: [Listing host@fqdn in BBS name entry] JV> This not only introduces the possibility to list nodes JV> that are allocated on the fly a different IP number each JV> time they connect to their ISP but also eliminates the JV> problem of lack of room in the flags field for flagging non JV> standard IP ports at the destination node as presumably this JV> would be solved by the destination's uplink. Please tell me how this bit about dynamic IP addresses works. Is it possible for an ISP to allocate a DNS entry to a dialup customer who has a dynamically-assigned IP address? Where does that DNS entry point to while that customer isn't on-line? How does the ISP avoid this DNS entry being cached by name servers across the world, thus avoiding outdated and therefore incorrect data to be served to people querying a non-authoritative name server (which is the normal thing to do)? Jan ___ timEd-A10a - Origin: Experimenter Board, Antwerp, Belgium (2:292/857) --- timEd-A10a * Origin: Experimenter Board, Antwerp, Belgium (2:292/857) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 9 Sep 97 9:13 #33 By: Mats Wallin To: Rune Johansen Re: IP-access St: ------------------------------------------------------------------------------ * Message copied from "SysOps [Z2]" (A copy of this message has been posted in IP_CONNECT, which is the conference where discussions about IP-access is supposed to be, to make it easy for all interested parties to take part of the discussion). RJ> So, this would also mean that IP-only nodes should be listed RJ> as Pvt, with "-Unpublished-" in their phone field, and DNS/IP RJ> address in BBS name field? This is not a good idea, IMHO. IP-nodes has to be included in the nodelist as full FidoNet members, and should not be listed as Pvt nodes. Using the BBS name field is not a good idea either, due to two reasons: * It would require a modification of every existing software before we can start to use IP-nodes. Adding the information in the phone number fields would make it possible to use IP today, with current software. * Many people like to see the name of the remote system, adding a DNS name, or worse, an IP number, is definitely not very user-friendly. RJ> Non-standard ports can also be indicated via "standard" URL RJ> type with colon as identifier, like this: RJ> f20.n210.z2.fidonet.org:28954 in the name field. Since some RJ> people would suggest that the port number for VModem should RJ> be standard, and some would say that telnet port should be RJ> standard, and some say that ifmail port should be standard, I RJ> would propose that the portnumber always should be RJ> included... If the port number is required, it would mean that a node that is able to accept both Telnet and Vmodem sessions is required to have two different nodenumbers. Not a very good idea. Port numbers should not be specified unless it is different from the default one. I recommend that a User flag is used to indicate which protocol(s) are supported, and that should also indicate which port that is used. RJ> I notice that you say "IP-nodes", meaning those that can be RJ> reached directly by using a connection-oriented protocol over RJ> IP, that runs over TCP again. That means that we currently RJ> have no solution for email-able nodes, using gating software RJ> for tunneling packets via email or other kinds of RJ> to-be-relayed-somehow transportation. I'm not objecting against email-able nodes, but IMHO, it is a rather different stories. A VModem/Telnet capable node is much closer to the requirements of a FidoNet node, which for instance is to support FTS-1, to be online and reachable during ZMH. All Vmodem/Telnet capable nodes are able to fulfill these requirements, while an email-able node is not. Mats mw@defsol.se --- * Origin: Stockholm, Sweden (2:201/329.11) SEEN-BY: 201/300 329 330 203/614 212/8 235/100 254/60 280/100 292/854 SEEN-BY: 292/857 876 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 9 Sep 97 23:11 #34 By: Denis McMahon To: All Re: Hi St: ------------------------------------------------------------------------------ Hello All! I'm here. Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 10 Sep 97 22:32 #35 By: Ward Dossche To: Denis McMahon Re: Hi St: ------------------------------------------------------------------------------ > I'm here. So you made it back through the tunnel? \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 10 Sep 97 0:57 #36 By: Pedro Lima To: Mats Wallin Re: IP-connectivity St: ------------------------------------------------------------------------------ Hello, MW> One requirement that has to be met, if we should be able to accept MW> IP-only nodes, is that they can pickup their mail from another FidoNet MW> system. This means that there somewhere in the region has to be a MW> system with both IP and modem access. MW> What I would propose is that a system that has both IP and modem MW> access is setup as the Host of either a new net, or an existing net MW> (existing net is probably only possible if the existing host already MW> have IP access). Alternatively, the system with both IP and modem MW> access could of course be a hub instead. Yes, that's some of what I wanted to hear. You see, even if the NC has both modem and IP access, if we're going (as it seems, and I agree) for the alternative of using the phone number field for the address (be it IP or FQDN), then an aka has to be used. This aka can be a hub in the net, then, and naturally, all the nodes listed below it are fed by him. If the IP hub and the NC aren't the same system, then we need the NC and the hub to route netmail between them, but this isn't forceably implied by Policy, it's left to the discretion of the NC, and there's still the odd case where there isn't a host with both modem and IP access. All in all, I think we need to start considering defining some standard procedures for the situations of non-modem nodes. Ideally, these would be covered by Policy, but for the time being it's possible to write a non-official document about it (much in the way gatewaying between zones and other nets is "standardized"). MW> I think this should solve all your questions. Not all of them, but your reply was very encouraging. MW> So I really don't see that part as a problem at all. You're an optimist, and that's good. :-) Best regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 10 Sep 97 3:45 #37 By: Pedro Lima To: Mats Wallin Re: IP-access St: ------------------------------------------------------------------------------ Hi, MW> IMHO, a node that only have a dialup-connection to the internet should MW> not be listed in the nodelist. I agree. LB> How do you handle IP-nodes with a variable IP-address (...) LB> Supporting DNS might be a solution to this. MW> And remember, that DNS name would have to be spread all over the MW> world at once, which you hardly can do. This reasoning is valid for the top levels of DNS, like "whatever.org" or even "whocares.whatever.org", but I doubt that it's forceably valid for a domain like "myhosts.whocares.whatever.org". The basis of the DNS working is to have top level name servers to list the name servers of the sub-domains, then the more specific name server is consulted, and so on, so there's the possibility of a domain served by only one name server, and although I think it might be possible in such a situation, probably with the complicity of the ISP, to change the DNS entries "on the fly", this could easily fail by the caching function of name servers, as Jan pointed out. It's much simpler and quite possible to have a fixed address whenever the system logs on (and possibly a FQDN assigned to it). Nonetheless, I don't see how this brings any advantage for FidoNet purposes. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 10 Sep 97 2:34 #38 By: Pedro Lima To: Jan Ceuleers Re: IP-access St: ------------------------------------------------------------------------------ Hi, JC> Quite obviously, the name server for zone 2 is z2.fidonet.org. Actually, it isn't obvious. A probable name could perfectly be ns.z2.fidonet.org. JC> My main problem is that due to the caching function I mentioned above, JC> it doesn't seem to be possible to accomodate nodes whose ISP JC> dynamically assigns them an IP address as they log on. I don't think JC> DNS can cope with that, but I'd be happy to be proved wrong. I believe you're absolutely right. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 11 Sep 97 2:51 #39 By: Pedro Lima To: Mats Wallin Re: IP-access St: ------------------------------------------------------------------------------ Hi, [IP-nodes not listed as Pvt] [Existant software already dealing with address in phone field] [Maintaining user-friendliness of system name field] [Default ports not listed] [Non-default ports listed in the flags] [Differentiation between direct and indirect connectivity] Just to say, FWIW, that I fully agree with the whole of your message. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 11 Sep 97 2:53 #40 By: Pedro Lima To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ Hi, WD> Me neither ... this is just something technical again which, in this WD> particular case, is quite easy. Ok, then let's get on with it. :-) Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 11 Sep 97 10:04 #41 By: Ward Dossche To: Pedro Lima Re: The IP-document comittee St: ------------------------------------------------------------------------------ Pedro, > All in all, I think we need to start considering defining some standard > procedures for the situations of non-modem nodes. Ideally, these would > be covered by Policy, but for the time being it's possible to write a > non-official document about it (much in the way gatewaying between zones > and other nets is "standardized"). Very good! Do I take you have hereby volunteered to at least participate in that document drafting-group? \x/ard --- DB 1.58/001874 * Origin: His master's voice (2:292/854.1) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 11 Sep 97 10:12 #42 By: Ward Dossche To: Jan Ceuleers Re: Name-server St: ------------------------------------------------------------------------------ Jan, > Quite obviously, the name server for zone 2 is z2.fidonet.org. Nope, it is : ns.z2.fidonet.org If you want to check the who caboozle pls access ftp://ftp.z2.fidonet.org/pub/fidonet/dns.z2/fidoz2.dns \x/ard --- DB 1.58/001874 * Origin: His master's voice (2:292/854.1) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 11 Sep 97 12:06 #43 By: Anton Kuznetsov To: All Re: Quiet St: ------------------------------------------------------------------------------ Buenas dias, amigo All All ! Replying to a message of Ward Dossche to Jan Ceuleers [07 Sep 97, 09:40]: >> It's awfully quiet in here. Looks like the discussion is continuing in >> several places instead of being concentrated here as per Ward's >> suggestion. WD> Give or take 2 weeks for people to become aware and get this stuff WD> distributed. Just look at SEEN+BY and you see where it goes for the WD> moment. AFAIK that's all yet: 292/854 ÃÄ251/20 ÃÄ292/857 ÃÄ362/21 Ã-465/50 ÀÄ201/329 ÀÄ201/330 Anthony. Donetsk, Ukraine --- E-mail: tony@tony.donetsk.ua * Origin: Everything you say may be used against you. (2:465/50) SEEN-BY: 50/0 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 450/97 460/111 461/35 463/94 SEEN-BY: 463/690 465/11 50 70 85 87 467/67 480/33 2432/200 4625/1 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 11 Sep 97 5:28 #44 By: Pedro Lima To: Rune Johansen Re: IP-access St: ------------------------------------------------------------------------------ * Crossposted from: ENET.SYSOP Hi, RJ> The domain "dyndns.com" uses a very low timeout in their SOA record, RJ> forcing the DNS server caches to expire very quickly. I thought it was possible for a name server, not being a primary nor a secondary name server for the requested domain, to cache the IP address of that domain, but I must be wrong, then. In that case, a FQDN attributed to a variant IP address is indeed possible. RJ> The clients of the RJ> dyndns.com domain uses some kind of signalling to the dyndns.com RJ> server that they are currently available at this-and-that IP address. If the system knows that client X is now connected to a certain IP address, it can create the corresponding entries in the DNS database files and update their SOA serial numbers. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 11 Sep 97 5:48 #45 By: Pedro Lima To: Denis McMahon Re: IP-access St: ------------------------------------------------------------------------------ * Crossposted from: ENET.SYSOP Hello, PL> I'm curious. How? DM> hostname in systemname field perhaps I was asking how was it possible to have a FQDN pointing to a variant IP address. Take a look at Rune's message. DM> OK, modified suggestion for ip nodes in nodelist: DM> ,nodenumber,dns-name,location,sysop,0xx-ip-address,300,flags Hmm, this would lead to software having to consult up to two fields (otherwise one of them would be useless) to understand where it should try the connection. I see no advantage in it, it's quite likely that the phone number field supports FQDNs (it supports '-Unpublished-'), and if it's possible for a FQDN pointing to a variant IP address, then the FQDN can be used, no need for the IP address. But I also think we shouldn't allow addresses which are probably not online when someone tries to connect to them. This could be solved by the Txx flags, perhaps? DM> suggested values for xx are: DM> 00 .. 09 reserved for ISDN allocations DM> 10 - vmodem available at this address DM> 11 - ftp access available at this address DM> 12 .. 19 reserved for other ip methods DM> 20 .. 29 reserved for x25 methods No need for it, as long as we have flags to differentiate between the several network technologies and respective protocols. Or are we also worried of someone attempting to IP-connect to X.25? :-) DM> If we make the data available in the nodelist in a standard format, DM> then the software developers have the information they need to use the DM> data. If we spend 20 years arguing over how we put the data in the DM> nodelist, we never get the new connectivity!!!!!!!!! Agreed. We're all working on it, and let me tell you, this is going surprisingly fast, IMHO. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 12 Sep 97 13:34 #46 By: Mats Wallin To: Lothar Behet Re: IP-access St: ------------------------------------------------------------------------------ * Message copied from "SysOps [Z2]" [A copy of this message has been posted in IP_CONNECT. Please move the discussion there instead.] MW>> IMHO, a node that only have a dialup-connection to the MW>> internet should not be listed in the nodelist. LB> At least not as regular node (maybe pvt). I think that the requirements for Pvt status should apply equally to a TCP/IP nodes as it does for modem nodes, so the only reason why someone should receive a Pvt-TCP/IP node should be that they provide some type of service to the network as a whole. If they only want to a TCP/IP connections to pickup their mail, they can either be setup as a point, or if they already have a modem-node number, use that nodenumber. MW>> In most cases, you do not have a static IP-address MW>> when you use a dialup connection to your ISP. LB> But it is possible. True, and I guess that it would be possible to list them as non-CM nodes in that case, _if_ they are connected to their ISP during ZMH every day. Mats mw@defsol.se --- * Origin: Definite APXw (2:201/329.11) SEEN-BY: 201/300 329 330 203/614 212/8 235/100 254/60 280/100 292/854 SEEN-BY: 292/857 876 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 12 Sep 97 13:39 #47 By: Mats Wallin To: Rune Johansen Re: IP-access St: ------------------------------------------------------------------------------ * Message copied from "SysOps [Z2]" [Also posted in IP_CONNECT] RJ> Hmm, gotta check up on the availability of it at 24/888, or RJ> maybe at 20/11. If you can't find it either of these system, let me know, and I'll make sure to send it to 20/11. RJ> Example (I don't know what port VModem has got, but I'll use RJ> #505) if you have Vmodem and you have telnet cap. in vmodem: I must admit that I thought that VModem used port 23 for telnet sessions, and the VModem port for VModem sessions? Guess it is possible to change that, but are you sure you can use the same port number for different protocols? I would assumed that it wasn't possible, but I might be wrong. Mats mw@defsol.se --- * Origin: Stockholm, Sweden (2:201/329.11) SEEN-BY: 201/300 329 330 203/614 212/8 235/100 254/60 280/100 292/854 SEEN-BY: 292/857 876 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 12 Sep 97 7:05 #48 By: Jan Ceuleers To: All Re: IP-connectivity St: ------------------------------------------------------------------------------ --> Note: Forwarded (from: REGCON) by Jan Ceuleers using timEd. Originally from Jan Ceuleers (2:292/857.0) to David Moufarrege. Original dated: Sep 12 '97, 07:03 I quote David Moufarrege: JS> See my response to Jan: a non-existent country code could all too JS> easily be mistaken for a long distance carrier access code. What are JS> the odds that this mistake would save someone money? ;<) DM> Any long distance carrier access code starts like a DM> traditional US phone number, namely 1, followed by the DM> number 0. DM> Starting an IP with a 9 will would at most emulate DM> PBX-systems that require a 9 to get an outside line. Please see my earlier message about this. You are mixing up the NANP with the international dialling plan. A country code is what you dial after the 011 prefix. It's there that we need to avoid collisions, not at the first digit position (the one that breaks the first dialtone). Jan ___ timEd-A10a - Origin: Experimenter Board, Antwerp, Belgium (2:292/857) --- timEd-A10a * Origin: Experimenter Board, Antwerp, Belgium (2:292/857) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 12 Sep 97 20:45 #49 By: Guido Gybels To: Anton Kuznetsov Re: Quiet St: ------------------------------------------------------------------------------ Hello Anton. Thursday September 11 1997 12:06, Anton Kuznetsov wrote to All: AK> AFAIK that's all yet: AK> 292/854 AK> ÃÄ251/20 AK> ÃÄ292/857 AK> ÃÄ362/21 AK> Ã-465/50 AK> ÀÄ201/329 AK> ÀÄ201/330 AK> Anthony. Donetsk, Ukraine AK> -+- E-mail: tony@tony.donetsk.ua AK> + Origin: Everything you say may be used against you. (2:465/50) AK> SEEN+BY: 50/0 201/329 212/8 235/100 254/60 270/101 291/709 1928 AK> 292/812 818 AK> SEEN+BY: 292/854 857 862 865 876 900 335/533 341/70 362/21 450/97 AK> 460/111 AK> SEEN+BY: 461/35 463/94 690 465/11 50 70 85 87 467/67 480/33 2432/200 AK> 4625/1 Not complete, your list ! :-) Guido --- GoldED/2 2.50.Beta5+ * Origin: [ The Optimist ] Antwerp-Belgium (2:292/876) SEEN-BY: 201/329 212/8 235/100 254/60 270/101 280/100 291/709 1928 SEEN-BY: 292/812 818 854 857 862 865 876 900 335/533 341/70 362/21 SEEN-BY: 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 12 Sep 97 4:02 #50 By: Pedro Lima To: Ward Dossche Re: The IP-document comittee St: ------------------------------------------------------------------------------ Hi, WD> Do I take you have hereby volunteered to at least participate in that WD> document drafting-group? If such a group would indeed be formed, and personal circumstances allowing it, yes. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 10:11 #51 By: rafal wiosna To: Joaquim Homrighausen Re: IP-connectivity St: ------------------------------------------------------------------------------ þ (IP_PUBLIC, Wed Sep 03 1997) Joaquim Homrighausen -> Ward Dossche: JH> What's wrong with using a simple prefix character.. (e.g. JH> "#198.32.3.112" or "#198-32-3-112")? Or does that break MakeNL/other JH> nodelist tools? My copy of FD 2.20c would still try to dial ATDP0w0#198323112, wouldn't it? - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Kiwa‡ to my, ale nie nas (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 10:42 #52 By: rafal wiosna To: Ward Dossche Re: Why IP? St: ------------------------------------------------------------------------------ þ (IP_PUBLIC, Sun Sep 07 1997) Ward Dossche -> All: WD> What would be the technical argument against including an entry in the WD> next nodelist with an IP-node for test-purposes? WHY IP?! Do you write e-mails to someone@[his.hum.ble.ip] or someone@his.humble.doma.in? Canonical addresses are much better by being more flexible. If my computer is moved from my office to the server room with a network sharing diffrent subnet, the IP-based nodelist entry immediatelly becomes obsolete. With a refresh period of 604800 seconds [7 days] of our "Fido DNS" 8^) it's nearly impossible to provide ACTUAL data on internet-based hosts by publishing their IP number. Using cannonical names it's up to the net part to resolve it and connections to orfika.office.polbox.pl are still possible whenever my computer is in my room or in server room -- as long as I change DNS entry and kill -1 NS' named. - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: You're a chicken, McFly! (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 10:19 #53 By: rafal wiosna To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ þ (IP_PUBLIC, Sun Sep 07 1997) Ward Dossche -> Pedro Lima: WD> However, should it bother an NC if an IP-only node cannot apply for a WD> node-number? I think it's the node's problem to figure thatone out ... WD> it's the same now too. What is the reason for becoming yet one more of the internet-based nodes? In my opinion, there should be 2-3 NODES in the internet area per region acting as Fido<->net gates [and being DIALABLE] and the rest should be just points. The net-only node [one that doesn't gate anything and is not dialable from PSTN/ISDN side] is some kind of phantom system -- i.e. in the internet you don't care from where you get your MAIL since you can CONNECT at any time paying nothing and there's no BUSY signal. It should be in some kind of order by means of properly routing mail but there is no need to have 100+ nodes there. Why do you set your Fido-internet system? To become a node or to read/write mail?... To tell you the truth you could use tin to read/write echomail as I do sometimes. One dialable net-node is unaccpetable since if it fails or it's bridge router gets a flu [or something] everybody's dead also. It's common sense to have more than one bridge node. - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Lieutenant Colonel Kilgore: You either surf or you fight (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 10:43 #54 By: rafal wiosna To: Alberto Pasquale Re: IP addressing St: ------------------------------------------------------------------------------ þ (IP_PUBLIC, Sat Aug 23 1997) Alberto Pasquale -> Ward Dossche: AP> ;S ,bbs.os2net.net,300,IP,VM - AP> ;S ,bbs.os2net.net,300,VM AP> ;S ,bbs.os2net.net,300,IP - Does MAKENL.EXE get upset when processing such entries? If yer, what do they use? - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Wst†p do OBZK - Ochotniczej Brygady Zaludniania Kraju! (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 10:43 #55 By: rafal wiosna To: Mats Wallin Re: IP-connectivity St: ------------------------------------------------------------------------------ þ (IP_PUBLIC, Mon Sep 08 1997) Mats Wallin -> Pedro Lima: MW> system. This means that there somewhere in the region has to be a MW> system with both IP and modem access. Well... It's implied. What's the reason for a net that has no gate to PSTN/ISDN world?... - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: If you're really hot, I bet I can cool you down [(c) B&B (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 10:43 #56 By: rafal wiosna To: Mats Wallin Re: Test of an IP-number in the nodelist St: ------------------------------------------------------------------------------ þ (IP_PUBLIC, Mon Sep 08 1997) Mats Wallin -> Ward Dossche: MW> As far as I can tell, MakeNL does not accept that as a phonenumber. (I MW> just tried it here, and it complained about it). Maybe it's time to improve MAKENL or start using another program? - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Fido gnije -- Rafa’ pali, Marek pije (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 10:37 #57 By: rafal wiosna To: Ward Dossche Re: Test of an IP-number in the nodelist St: ------------------------------------------------------------------------------ þ (IP_PUBLIC, Tue Sep 09 1997) Ward Dossche -> Mats Wallin: WD> That is a totally different matter, but quite appropriate. No updates on WD> MAKENL have been forthcoming for about 3,5 years I think. I have one young and enthusiastic programmer in my net. He did very nice netmail tracker and nodediff compiler. Do you want a MAKENLIP? 8^) - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: To piwo ukszta’towa’o to pi‘kne cia’o (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 12:20 #58 By: Ward Dossche To: rafal wiosna Re: Why IP? St: ------------------------------------------------------------------------------ Rafal, > WHY IP?! > Do you write e-mails to someone@[his.hum.ble.ip] or > someone@his.humble.doma.in? > Canonical addresses are much better by being more flexible. You have just started a complete new discussion on another issue. Please proceed. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 12:30 #59 By: Ward Dossche To: rafal wiosna Re: IP-connectivity St: ------------------------------------------------------------------------------ Rafal, > What is the reason for becoming yet one more of the > internet-based nodes? In my opinion, there should be 2-3 NODES in the > internet area per region acting as Fido<->net gates [and being DIALABLE] > and the rest should be just points. What you are telling here is that there should be a government telling the people what is good for them while denying them access to the things what they think themselves are good for them. Sound familiar? The technology is there and the people that want it are there, I see no reason to deny anyone to choose the technology he/she wants. > It should be in some kind of order by means of properly > routing mail but there is no need to have 100+ nodes there. I'll repeat ... the technology is there and I see no need why to refuse it. Sticking to the PSTN-only dreamworld is a killer for Fidonet, the introduction of IP is a breather! \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 12:31 #60 By: Ward Dossche To: rafal wiosna Re: IP addressing St: ------------------------------------------------------------------------------ > þ (IP_PUBLIC, Sat Aug 23 1997) Alberto Pasquale -> Ward Dossche: This was the forward of a netmail by Alberto. I don't think he's linked. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 12:34 #61 By: maciej grzeszczuk To: Pedro Lima Re: IP-access St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Thu, 11 Sep 97 01:51:12 +0200 Pedro Lima napisal byl w fido.ip_connect: > [IP-nodes not listed as Pvt] > [Existant software already dealing with address in phone field] what about nodes that allow both standard, and tcp connections? i reccomend using U,IFC or U,VMD flags that would indicate, that this system allows tcp connections by ifcico or vmodem, and DNS entry with ip address of that node. that's the easiest way. for software that cannot use DNS maybe we should reconsider using some kind of U,ADRmyhost.mydomain flags? > [Maintaining user-friendliness of system name field] > [Default ports not listed] what do you mean default? every protocol (telnet,vmodem,ifcico,binkd) has it's own 'default' port. > [Non-default ports listed in the flags] > [Differentiation between direct and indirect connectivity] if a system is not reachable by emsi (only e-mail transfers of echomail) it shouldn't be listed as a node... 1102 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: Vale Fax-Modem (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 SEEN-BY: 480/88 91 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 21:22 #62 By: rafal wiosna To: all Re: what about... St: ------------------------------------------------------------------------------ þ (IP_PUBLIC, Sat Sep 13 1997) maciej grzeszczuk -> Pedro Lima: mg> some kind of U,ADRmyhost.mydomain flags? IMHO it should be only "U,NET,__address__,..." where __address__ is either canonical name or an IP. I belive both gethostbyname("orfika.office.polbox.pl") and gethostbyname("195.116.6.19") returns a nice long with IP address (as long as IPv4 is considered). [OK, it could be "U,IP,__address__", it's one byte shorter...]. Or even better. We could put CNAME/IP in the _NAME_ field of nodelist entry. It'd shorten the nodelist a bit. This way only "U,NET" flag would be required [or better yet, mailer could assume that the node's net-able whenever either TCP or VMODEM flag is present]. This way software would know that it should resolve the systemname field to satisfy socket functions. And the flags could be like this: U,VMODEM,TCP,PORT,xxxxx... Port could be added to systemname field in the form of ":numbr", like in URLs. We can also put "0" in the phone field of a node entry if node's undialable. MAKENL.EXE does not complain. I strongly belive that it doesn't matter if you dial 0w00 from Poland or 011 from UK or any other prefix from any other country you get BUSY signal [or at least a free "invalid number" automated female announcement 8^)]. Anyone could confirm that? Ward? Mats? Zone1 people? mg> what do you mean default? every protocol (telnet,vmodem,ifcico,binkd) mg> has it's own 'default' port. It's default but it's not mandatory. I agree with Pedro: it'd be more flexible to provide non-standard port number information in the nodelist entry when a node uses one. mg> if a system is not reachable by emsi (only e-mail transfers of echomail) mg> it shouldn't be listed as a node... You're right. What kind of node is that? 8^) Proposals for nodelist entries: ,99,Im_dialable,Somewhere,John_Doe,12-34-567890,9600,...,U,NET,host.domain, VMODEM,TCP ,99,Im_not_dialable,Somewhere,John_Doe,0,300,U,NET,host.domain,VMODEM,TCP ,99,Im_not_dialable,Somewhere,John_Doe,0,300,U,NET,123.45.6.78,VMODEM,TCP ,99,host.domain,Somewhere,John_Doe,12-34-567890,9600,...,U,NET,VMODEM,TCP ,99,host.domain,Somewhere,John_Doe,0,300,U,NET,VMODEM,TCP ,99,123.45.6.78:7666,Somewhere,John_Doe,0,300,U,NET,VMODEM,TCP ,99,host.domain,Somewhere,John_Doe,0,300,U,TCP - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Seks w domu i zagrodzie (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 21:31 #63 By: rafal wiosna To: Ward Dossche Re: Test of an IP-number in the nodelist St: ------------------------------------------------------------------------------ þ (IP_PUBLIC, Sat Sep 13 1997) Ward Dossche -> rafal wiosna: >> Maybe it's time to improve MAKENL or start using another program? WD> Quite correct! No problem. 8^) - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Kumpli w §adnym wypadku nie wolno t‘pi‡ (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 21:54 #64 By: rafal wiosna To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ þ (IP_PUBLIC, Sat Sep 13 1997) Ward Dossche -> rafal wiosna: WD> What you are telling here is that there should be a government WD> telling the people what is good for them while denying them access to WD> the things what they think themselves are good for them. Sound WD> familiar? Well, it's elections time in Poland and maybe I'm getting a little paranoid from watching/hearing all this political crap. 8^) No. It's just common sense not to exaggerate the number of non-dialable internet nodes. On one side there's a network with 1 node/gate and many points and on the other end everyone's node and no points exist. It wouldn't harm to have, let's say, 10 nodes in the IP-area but there should be also points there... I'm puting 3 tablespoons of sugar in my tea not because they're rationed but because 5-spoon tea is worse than 3-spoon one. [WARNING: DON'T TRY THIS AT HOME! IT IS A CONTROLLED EXPERIMENT! 8^)] WD> The technology is there and the people that want it are there, I see WD> no reason to deny anyone to choose the technology he/she wants. Well... Being a point is not a dishonour. The sole purpose for being a node is to provide way/help with mail routing and to distribute traffic, right? So, if there's no BUSY and the connections are free it doesn't matter if there are 100 of nodes or 2 nodes in region... It's like Amway, they can't just exist, they have to sell something but the idea is to build the network, not to sell. 8^) Let's not become FidoAmway! 8^) Of course, I'm not saying that you should CTRL-Y 99% of IP-nodes in the region segments you get. It's just a suggestion for NCs. >> It should be in some kind of order by means of properly routing mail >> but there is no need to have 100+ nodes there. WD> I'll repeat ... the technology is there and I see no need why to WD> refuse it. It's some kind of politics voodoo I'm not good at. You know my point of view but I'm willing to negotiate. Just put down the gun, please?... 8^) - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Biuro Podr¢§y poleca: Pokusa co mnie wzi‘’a w USA (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 21:52 #65 By: rafal wiosna To: Ward Dossche Re: IP addressing St: ------------------------------------------------------------------------------ þ (IP_PUBLIC, Sat Sep 13 1997) Ward Dossche -> rafal wiosna: >> þ (IP_PUBLIC, Sat Aug 23 1997) Alberto Pasquale -> Ward Dossche: WD> This was the forward of a netmail by Alberto. I don't think he's linked. Should I forward it back to ENET.SYSOP or what?... - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Po co komu zlewozmywak? (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 12 Sep 97 22:39 #66 By: Johnny Carlsson To: Mats Wallin Re: IP-access St: ------------------------------------------------------------------------------ Replying to a message of Mats Wallin to Rune Johansen: MW> * Message copied from "SysOps [Z2]" MW> I must admit that I thought that VModem used port 23 for telnet sessions, MW> and the VModem port for VModem sessions? Guess it is possible to change MW> that, but are you sure you can use the same port number for different MW> protocols? I would assumed that it wasn't possible, but I might be wrong. From the Vmodem.doc The Virtual Modem Protocol (VMP) that is implemented by Vmodem uses TCP/IP Sockets. The default port number used by Vmodem is 3141 (the first 4 digits of pi, un-rounded). However, this default port number can be overridden by using the SERVICES file in the TCP/IP ETC directory. If the user defines a "Well-Known Port" called "vmodem" (lower case) in the SERVICES file, then that port will be used instead of 3141. The port number 3141 has been assigned to the Virtual Modem Protocol (VMP) by the Internet Assigned Numbers Authority (IANA). The name associated with port 3141 is "vmodem" (without the quotes). Likewise, the Telnet Server in Vmodem can be directed to use any port number by adding a "Well-Known Port" called VMOTelnet to your SERVICES file. The Telnet Server defaults to port 23 (the standard Telnet port). M.v.h Jny Sysop of Dive's BBS HUB 2:201/300 --- FleetStreet 1.17+ * Origin: Dive's BBS * 08-590 81576 * OS/2 BBS * (2:201/300) SEEN-BY: 201/0 300 324 329 330 203/614 212/8 235/100 254/60 280/100 SEEN-BY: 292/854 857 876 335/533 341/70 362/21 380/106 465/50 480/33 SEEN-BY: 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 13 Sep 97 16:26 #67 By: Carlos Fernandez Sanz To: Jan Ceuleers Re: IP-access St: ------------------------------------------------------------------------------ JC> Please tell me how this bit about dynamic IP addresses works. JC> Is it possible for an ISP to allocate a DNS entry to a dialup JC> customer who has a dynamically-assigned IP address? Theorically it is. Problem arises with caching, however this can be (more or less) avoided by using non-constant DNS names. For example, say that I am asigned cfs.nova.es each time I connect to my ISP (which is in fact true, but I have a fixed IP address). The calling mailer tries to resolve something like 9709131618.cfs.nova.es The number being the date and the time. Of course this has two problems, a) The must a DNS server resolving any address *.cfs.nova.es to my IP address. Since that DNS would be in cfs.nova.es, it is possible to do that locally. b) While 9709131618.cfs.nova.es would not be cached, cfs.nova.es would be. I donït know what happens if a program attempts to resolve an address and finds no DNS. I suppose the correct thing to do is to move upwards and resolve the address of the DNS first, but Iïm not sure (I donït have the RFCs collection here). There is of course a different option, which consists on running our own DNS on a fixed address, and provide the tools so each node can update its IP address on that DNS server. I think is solution to be easier, since there are more than enough nodes 24 hours in internet to run such a service for us. I would write whatever software it takes to do that... --- * Origin: Probando modem con pegatinas en hebreo (2:341/70.1) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/12 70 122 136 343/106 140 345/801 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 14 Sep 97 2:22 #68 By: Denis McMahon To: Pablo Saratxaga Re: IP-access St: ------------------------------------------------------------------------------ * Original message posted in: ENET.SYSOP. * Crossposted in: IP_CONNECT. Pablo Saratxaga wrote to Alto Speckhardt: PS> From: Pablo Saratxaga PS> The reason I propose to use the DNS is that it will decreas *a lot* PS> the required size in nodelist OK, so with your proposal all that is needed is a flag for each type of ip connect a system can accept, and the system should be identified using a dns lookup on the f*.n*.z*.fidonet.org address, yes? In that case, we still need a flag for each type of connect available, ie ftp, vmodem session, uucp etc. Otherwise, how do you know whether you should send your pkt file to the ip node using vmodem, uucp or anon ftp transmission? Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 14 Sep 97 2:54 #69 By: Denis McMahon To: Pedro Lima Re: IP-access St: ------------------------------------------------------------------------------ * Original message posted in: ENET.SYSOP. * Crossposted in: IP_CONNECT. Pedro Lima wrote to Denis McMahon: DM> hostname in systemname field perhaps PL> I was asking how was it possible to have a FQDN pointing to a PL> variant IP address. Take a look at Rune's message. DM> OK, modified suggestion for ip nodes in nodelist: DM> ,nodenumber,dns-name,location,sysop,0xx-ip-address,300,flags PL> Hmm, this would lead to software having to consult up to two fields PL> (otherwise one of them would be useless) to understand where it PL> should try the connection. I see no advantage in it, it's quite PL> likely that the phone number field supports FQDNs (it supports PL> '-Unpublished-') I think you'll find that makenl does some syntax checking on the telno field, and only accepts "-Unpublished-" or numerical telno data. I may be wrong. Nope, I'm not wrong. I just tried it: These are the ones I tried: ,96,test_1,nowhere,no-one,1-host-sub-domain-org,300,UFLAGS ,97,test_2,nowhere,no-one,013-251-18-49-167,300,UFLAGS Pvt,98,test_3,nowehere,no-one,host.sub.domain.org,300,UFLAGS ,99,test_4,nowhere,no-one,0host-sub-domain-org,300,UFLAGS 96 was rejected as invalid telno 97 was OK 98 telno was changed to "-Unpublished-" 99 was rejected as invalid telno Now, makenl is probably the single most important piece of software in fidonet today, and Ben Baker isn't listed in the Nodelist anymore. Unless someone is going to re-write makenl, or has Ben's source code and is in a position to develop the software, we've got little choice but to work within the constraints of that program! PL> , and if it's possible for a FQDN pointing to a PL> variant IP address, then the FQDN can be used, no need for the IP PL> address. PL> But I also think we shouldn't allow addresses which are probably PL> not online when someone tries to connect to them. This could be PL> solved by the Txx flags, perhaps? DM> suggested values for xx are: DM> 00 .. 09 reserved for ISDN allocations DM> 10 - vmodem available at this address DM> 11 - ftp access available at this address DM> 12 .. 19 reserved for other ip methods DM> 20 .. 29 reserved for x25 methods PL> No need for it, as long as we have flags to differentiate between PL> the several network technologies and respective protocols. Or are PL> we also worried of someone attempting to IP-connect to X.25? :-) Look again at the whole proposal, and think about how telno translation is applied to the nodelist. This method allows translation to a gateway telno using the first three digits of the telno in the nodelist. It's a lot easier to implement in current mailers than selecting a gateway according to the flags. Here's an example translation for qnode: ========== Example Dialcost 10 50 0 00 Local numbers 44-1705- / 5 ; National numbers 44-1 01 10 ; IP gateways 010- 01abc-defghi 10 ; vmodem gateway 011- 01jkl-mnopqr 10 ; ftp gateway End ========= End With this translation, any packets for a vmodem address would be sent to a gateway that could forward to vmodem ip nodes, and any packets for ftp address would be sent to a gateway that could use anonymous ftp to send to ftp ip nodes. The nodelist entries are compatible with makenl. If the node was entered in the nodelist with a domain name instead, then the address field would be 0xx-0-0-0-0, the number translation would still work, and the gateway would be responsible for the dns lookup. Gateways might need an extra bit of software to generate an ip list to use in a format like: nodenumber,dns-name,ip-address,connection-method This could be generated by extraction from the nodelist, and could then be used by the ip side of the gateway to send the packets / files on to the destination node. Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 14 Sep 97 14:04 #70 By: Ward Dossche To: rafal wiosna Re: IP-connectivity St: ------------------------------------------------------------------------------ Yow Rafal, > WD> The technology is there and the people that want it are there, I see > WD> no reason to deny anyone to choose the technology he/she wants. > Well... Being a point is not a dishonour. Maybe it is turning the world upside down. IP-capable points could become the major feed of the non-IP-capable boss-system. > It's some kind of politics voodoo I'm not good at. You know my > point of view but I'm willing to negotiate. Just put down the gun, > please?... 8^) I know too little of this technology to hold a gun, I'm merely trying to facilitate the discussion and the implementation. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 14 Sep 97 14:05 #71 By: Ward Dossche To: rafal wiosna Re: IP addressing St: ------------------------------------------------------------------------------ >>> þ (IP_PUBLIC, Sat Aug 23 1997) Alberto Pasquale -> Ward Dossche: > WD> This was the forward of a netmail by Alberto. I don't think he's > linked. > Should I forward it back to ENET.SYSOP or what?... You could ... but too much of the IP-discussion still goes on elsewhere ... it would be spreading the effort. Lets convice Alberto to link-in, I think he has a contribution to make. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 14 Sep 97 17:46 #72 By: rafal wiosna To: Denis McMahon Re: IP-access St: ------------------------------------------------------------------------------ þ (IP_PUBLIC, Sun Sep 14 1997) Denis McMahon -> Pablo Saratxaga: DM> OK, so with your proposal all that is needed is a flag for each type of DM> ip connect a system can accept, and the system should be identified using DM> a dns lookup on the f*.n*.z*.fidonet.org address, yes? That's probably not the best idea. Maintaining this huge .fidonet.org would be a hassle, transfering the whole domain could be tricky and keeping it up to date nearly impossible. Can you imagine the machine capable of holding and RESOLVING ~100,000 DNS entries [nodes+points; make it double for rDNS] and the load it gets when some query comes in?... I know it could be distributed but not every region has it's MX/NS... - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Lepiej 7 razy ze ˜nie§k† ni§ raz z siedmioma krasnoludka (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 14 Sep 97 14:33 #73 By: Joaquim Homrighausen To: rafal wiosna Re: IP-connectivity St: ------------------------------------------------------------------------------ * On Sat 13 Sep 97 10.11, wrote: JH>> What's wrong with using a simple prefix character.. (e.g. JH>> "#198.32.3.112" or "#198-32-3-112")? Or does that break JH>> MakeNL/other nodelist tools? rw> My copy of FD 2.20c would still try to dial rw> ATDP0w0#198323112, wouldn't it? Not if you as the first string in your dial translation table put: # ;Internet connection That oughta work.. %JoHo% joho@defsol.se * Origin: Definite Solutions ~/Stockholm, Sweden (2:201/330.1) SEEN-BY: 201/300 329 330 203/614 212/8 235/100 254/60 280/100 292/854 SEEN-BY: 292/857 876 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 14 Sep 97 22:10 #74 By: rafal wiosna To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ þ (RAF, Sun Sep 14 1997) Ward Dossche -> rafal wiosna: >> Well... Being a point is not a dishonour. WD> Maybe it is turning the world upside down. IP-capable points could become WD> the major feed of the non-IP-capable boss-system. It's very likely. Usually the readers feed newsservers with articles, not the administrators... 8^) - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: This is generic origin line, no added features (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 14 Sep 97 15:50 #75 By: Denis McMahon To: Jan Ceuleers Re: IP-access St: ------------------------------------------------------------------------------ * Original message posted in: NETMAIL. * Crossposted in: IP_CONNECT. * Reply to a message in REGCON.EUR. Jan Ceuleers wrote to Jan Vermeulen: JC> I quote Jan Vermeulen: JV>> ... nodes that are allocated on the fly a different IP number each JV>> time they connect to their ISP jc> Please tell me how this bit about dynamic IP addresses works. JV> I do not know HOW it works, I do know THAT it works. JC> You haven't addressed the question at the center of my argument. JC> Yes, I know that dynamic IP addresses themselves do indeed work. My JC> ISP assigns me one each time I log on. What's more important is JC> that it's not possible for dynamic IP addresses to be assigned DNS JC> entries. This is why the IP address _must_ be included in the JC> nodelist, at least for nodes that have dynamic IP addresses. See my modified proposal, where a dynamic ip address is used the system name field contains the host.sub-domain.domain information needed for dns lookup, then as long as the ip-node is connected via their isp during zmh then ftp, vmodem etc (whatever is supported) becomes available for callers. Maintaining the dns for the lookup function is a problem of the isp, who has to ensure that dns lookup gives a correct ip address when the host is online. Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 14 Sep 97 18:01 #76 By: Mario Mure' To: All Re: IP addressing St: ------------------------------------------------------------------------------ From: mure@denise.sol.net (Mario Mure') In a message dated 14 Sep 1997 about < IP addressing >, Ward Dossche wrote: WD> Lets convice Alberto to link-in, I think he has a contribution to make. I know for sure that Alberto is linked, as I'm feeding this echo to him :) Ciao ! /mario/ mure@sistemia.it --- BBBS/L v3.33 How * Origin: Speed Of Life II +39-931-833-763 (2:335/534) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 14 Sep 97 20:30 #77 By: Pedro Lima To: maciej grzeszczuk Re: IP-access St: ------------------------------------------------------------------------------ Hi, mg> what about nodes that allow both standard, and tcp connections? mg> i reccomend using U,IFC or U,VMD flags that would indicate, that this mg> system allows tcp connections by ifcico or vmodem, and DNS entry with mg> ip address of that node. that's the easiest way. It would be solved by an AKA. However, one doesn't exclude the other, i.e., it's also possible to use the FQDN associated with the Fido address. In that case, only the flags would be listed. Let me be clear about this, either way I think that appropriate flags should be listed. mg> for software that cannot use DNS maybe we should reconsider using mg> some kind of U,ADRmyhost.mydomain flags? I'm really not in favor of using flags to specify an Internet address, for all the reasons stated before. > [Default ports not listed] mg> what do you mean default? every protocol (telnet,vmodem,ifcico,binkd) mg> has it's own 'default' port. Precisely. Each IP-related flag has a default port associated, so there's no need to specify it. For example, it would only be needed to list 'U,VMD' to know that the port is 3141, or 'U,ITLN' to know the port is 23. These flags are obviously just examples, we still have to address this subject in more detail. > [Differentiation between direct and indirect connectivity] mg> if a system is not reachable by emsi (only e-mail transfers of mg> echomail) it shouldn't be listed as a node... For me, the real difference lies in whether it's possible or not to connect directly to the node. EMSI is only one of many possible choices. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 14 Sep 97 20:34 #78 By: Pedro Lima To: rafal wiosna Re: Why IP? St: ------------------------------------------------------------------------------ rw> WHY IP?! Why not admitting both possibilities? Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 14 Sep 97 20:36 #79 By: Pedro Lima To: Rune Johansen Re: IP-access St: ------------------------------------------------------------------------------ * Crossposted from: ENET.SYSOP Hello, RJ> They _do_ cache it, but they are only allowed to cache it with regard RJ> to the expiry time in the SOA record of the entry's domain. Ah, then I wasn't completely mistaken, they're working similarly to a secondary name server. This also allows for a FQDN to be attributed to a variant IP. There's always the risk of the cache not being updated at the time someone tries to connect, but it's minimal. Also, many of those connections would work with authentication procedures (like session passwords), so I think it's perfectly feasible. RJ> Yes, they have a certain client software that tells the DNS server RJ> where they are now. I took a peek at http://www.dyndns.com, and even downloaded the ELF binary for Linux, but I couldn't find the source. Nonetheless, this should actually be a simple program. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 15 Sep 97 16:32 #80 By: Anton Kuznetsov To: All Re: IP addressing St: ------------------------------------------------------------------------------ Buenas dias, amigo All All ! Replying to a message of Alberto Pasquale to Ward Dossche [23 Aug 97, 08:50]: AP> May I bring your attention to what has already been "standardized" in AP> OS2Net ? AP> ;S Internet flags: AP> ;S AP> ;S Flag Meaning AP> ;S IP Internet Protocol. Node accepts Telnet sessions. AP> ;S VM Vmodem capability, Host accepts Vmodem connection. What about BinkD capability [BSO FTN-mailer via TCP/IP] ?.. Anthony. Donetsk, Ukraine --- E-mail: tony@tony.donetsk.ua * Origin: Everything you say may be used against you. (2:465/50) SEEN-BY: 50/0 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 450/97 460/111 461/35 463/94 SEEN-BY: 463/690 465/11 50 70 85 87 467/67 480/33 2432/200 4625/1 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 15 Sep 97 10:26 #81 By: Mats Wallin To: Johnny Carlsson Re: IP-access St: ------------------------------------------------------------------------------ JC> From the Vmodem.doc True, but if you attempt to use the same port number for both services, how would VModem know which service you're requesting? I have always thought that it was the portnumber that told the server how to proceed. There is nothing special a client program should send during a Telnet session to indicate that it is a Telnet session. Mats mw@defsol.se --- * Origin: Definite APXw (2:201/329.11) SEEN-BY: 20/11 201/300 329 330 203/614 212/8 235/100 254/60 280/100 SEEN-BY: 292/854 857 876 335/533 341/70 362/21 380/106 465/50 480/33 SEEN-BY: 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 15 Sep 97 10:32 #82 By: Mats Wallin To: rafal wiosna Re: IP-connectivity St: ------------------------------------------------------------------------------ rw> What is the reason for becoming yet one more of the rw> internet-based nodes? In my opinion, there should be 2-3 rw> NODES in the internet area per region acting as Fido<->net rw> gates [and being DIALABLE] and the rest should be just rw> points. One of the reasons why they should not be points, is that a point is not a member of FidoNet, they are only a user of another system. Just because you're using Internet to connect to your hub/host doesn't mean that you're not interested in being a member of FidoNet, which for instance makes it possible to vote in elections. The Internet is just another carrier of data, and there is no reason why one shouldn't be allowed to be a full FidoNet member just because they connect via the Internet. Mats mw@defsol.se --- * Origin: Definite Solutions, Stockholm, Sweden (2:201/329.11) SEEN-BY: 20/11 201/300 329 330 203/614 212/8 235/100 254/60 280/100 SEEN-BY: 292/854 857 876 335/533 341/70 362/21 380/106 465/50 480/33 SEEN-BY: 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 15 Sep 97 10:46 #83 By: Mats Wallin To: Denis McMahon Re: IP-access St: ------------------------------------------------------------------------------ PS>> The reason I propose to use the DNS is that it will decreas PS>> *a lot* the required size in nodelist DM> OK, so with your proposal all that is needed is a flag for DM> each type of ip connect a system can accept, and the system DM> should be identified using a dns lookup on the DM> f*.n*.z*.fidonet.org address, yes? The problem with that, IMHO, is that it's not possible to use the programs of today together with a (fossil-) driver. Only programs that have been written for this purpose would be possible to use, while it would be possible to use most of the current mailers if the IP/domain address were inserted into the phone number field. I think that is a rather important fact that one should consider. We can't expect every mailer to be updated to support this, but it would technically be possible to provide different drivers for most operating systems that could make sure that most of the mailers available today could work with the IP/domain address in the phone number field. DM> In that case, we still need a flag for each type of connect DM> available, ie ftp, vmodem session, uucp etc. Since FTS-1 is a basic requirement for a member of FidoNet, I thought we only were talking about sessions that could establish FTS-1, FTS-6 or EMSI sessions. Mats mw@defsol.se --- * Origin: Definite APXw (2:201/329.11) SEEN-BY: 20/11 201/300 329 330 203/614 212/8 235/100 254/60 280/100 SEEN-BY: 292/854 857 876 335/533 341/70 362/21 380/106 465/50 480/33 SEEN-BY: 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 15 Sep 97 8:59 #84 By: rafal wiosna To: Joaquim Homrighausen Re: IP-connectivity St: ------------------------------------------------------------------------------ þ (RAF, Sun Sep 14 1997) Joaquim Homrighausen -> rafal wiosna: JH> Not if you as the first string in your dial translation table put: ># ;Internet connection JH> That oughta work.. So it won't dial at all? - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Uh, get out of my car and into my dreams, baby [(c) B&B] (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 15 Sep 97 14:46 #85 By: maciej grzeszczuk To: rafal wiosna Re: what about... St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Sat, 13 Sep 97 20:22:04 +0200 rafal wiosna napisal byl w fido.ip_connect: > IMHO it should be only "U,NET,__address__,..." where __address__ is > either canonical name or an IP. I belive both > gethostbyname("orfika.office.polbox.pl") and gethostbyname("195.116.6.19") > returns a nice long with IP address (as long as IPv4 is considered). [OK, it > could be "U,IP,__address__", it's one byte shorter...]. you got my idea, however we name that flag... IP seems to be the best, as it's the shortest one. > Or even better. We could put CNAME/IP in the _NAME_ field of > nodelist > entry. It'd shorten the nodelist a bit. This way only "U,NET" flag would be > required [or better yet, mailer could assume that the node's net-able > whenever > either TCP or VMODEM flag is present]. This way software would know that it > should resolve the systemname field to satisfy socket functions. i don't agree with it. the name should remain as it is now. if we need address, we should provide it after a separate flag (IP,ADR,NET, whatever...). > And the flags could be like this: U,VMODEM,TCP,PORT,xxxxx... Port > could > be added to systemname field in the form of ":numbr", like in URLs. vmodem, ifcico, binkd and so on has already their own port numbers, so there is no need to add one after vmodem or ifcico flag. > We can also put "0" in the phone field of a node entry if node's > undialable. MAKENL.EXE does not complain. I strongly belive that it doesn't i prefer tcp-only nodes would be pvt ones - as there is no way to reach them via standard modem connection. their phone number is then -unpublished-. pvt nodes makes no dishonor to it's operator - he's a member with the same rights as non-pvt one. > It's default but it's not mandatory. I agree with Pedro: it'd be > more > flexible to provide non-standard port number information in the nodelist > entry > when a node uses one. but not in case ifcico resides on it's standard port. > mg> if a system is not reachable by emsi (only e-mail transfers of > echomail) > mg> it shouldn't be listed as a node... > You're right. What kind of node is that? 8^) > Proposals for nodelist entries: > ,99,Im_dialable,Somewhere,John_Doe,12-34-567890,9600,...,U,NET,host.domain, > VMODEM,TCP > ,99,Im_not_dialable,Somewhere,John_Doe,0,300,U,NET,host.domain,VMODEM,TCP > ,99,Im_not_dialable,Somewhere,John_Doe,0,300,U,NET,123.45.6.78,VMODEM,TCP > ,99,host.domain,Somewhere,John_Doe,12-34-567890,9600,...,U,NET,VMODEM,TCP > ,99,host.domain,Somewhere,John_Doe,0,300,U,NET,VMODEM,TCP > ,99,123.45.6.78:7666,Somewhere,John_Doe,0,300,U,NET,VMODEM,TCP > ,99,host.domain,Somewhere,John_Doe,0,300,U,TCP what do you mean TCP? all that vmodem, ifcico, binkd use tcp for transport. my proposals are: if undialable Pvt,99,hostname,somewhere,John_Doe,-unpublished-,300,U,IP,host.domain,VMD,IFC, BNK if dialable ,99,hostname,somewhere,John_Doe,48-22-31337,9600,V32B,V42B,U,IP,host.domain,VM D,IFC,BNK where VMD stands for vmodem connections (with it's default port), IFC for ifcico with it's default port, and BNK for binkd. if a node provides service on other than standard port, then shoulk use flag PORT after the flag that stands for specified service, fe: ,U,IP,host.zdupy.pl,IFC,PORT,60177,BNK,PORT,31337 1107 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: ja pana znam, panie... yyy... mmm.... (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 SEEN-BY: 480/88 91 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 15 Sep 97 15:07 #86 By: maciej grzeszczuk To: All Re: IP-access St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Sun, 14 Sep 97 16:46:20 +0200 rafal wiosna napisal byl w fido.ip_connect: > RESOLVING ~100,000 DNS entries [nodes+points; make it double for rDNS] and > the > load it gets when some query comes in?... I know it could be distributed > but not > every region has it's MX/NS... talking about separate NS'es for every region capable - i think Wim should delegate NS record for region that can maintain it's own DNS - we could then maintain our MX records, A records for EMSI via TCP purposes, and so on. publishing all zone files wouldn't be impossible as fidonet.z2.fidonet.org would be secondary dns for all that subdomains and keep zone files retrieved from primary one. 1108 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: w pupe chrupac - (c) by ____ (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 SEEN-BY: 480/88 91 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 15 Sep 97 9:27 #87 By: Denis McMahon To: Jan Ceuleers Re: IP-connectivity St: ------------------------------------------------------------------------------ * Original message posted in: NETMAIL. * Crossposted in: IP_CONNECT. * Reply to a message in REGCON. Jan Ceuleers wrote to David Moufarrege: JC> I quote David Moufarrege: JS> See my response to Jan: a non-existent country code could all too JS> easily be mistaken for a long distance carrier access code. What are JS> the odds that this mistake would save someone money? ;<) DM> Any long distance carrier access code starts like a traditional US DM> phone number, namely 1, followed by the number 0. Starting an IP DM> with a 9 will would at most emulate PBX-systems that require a 9 to DM> get an outside line. JC> Please see my earlier message about this. You are mixing up the JC> NANP with the international dialling plan. A country code is what JC> you dial after the 011 prefix. It's there that we need to avoid JC> collisions, not at the first digit position (the one that breaks JC> the first dialtone). No, you are misunderstanding dial translation as applied to the nodelist. At the moment, a telno in the nodelist consists of: "valid-country-code"-"national-code"-"national-number" Nodelist compilers translate "valid-country-code" by either replacing it with "national-prefix" (if own-country-code) or by prefixing it with "international-access-code" if (not own-country-code). Additional translations may be done for "ld-access-code" and national-code = own-national-code. At this translation level (the translation applied to national code) the proposed "invalid-national-code" used to identify non analogue modem systems can be translated to a gateway telno (see another recent message in ip_connect), extracted by third party software for ip connection use (if it is an ip connect identifier) or used for isdn connect if it is an isdn identifier. Instead of trying to invent problems where none exist, why don't you look for a solution instead? Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 15 Sep 97 10:52 #88 By: Steve Woodmore To: Denis McMahon Re: IP-access St: ------------------------------------------------------------------------------ Hi Denis McMahon, hope you are having a nice day 14-Sep-97 02:22:47, Denis McMahon wrote to Pablo Saratxaga Subject: IP-access DM> In that case, we still need a flag for each type of connect DM> available, ie ftp, vmodem session, uucp etc. Read my proposal. -=> Steve Woodmore <=- --- Terminate 5.00/Pro * Origin: Terminate has most advertising origin lines ;-) (2:440/410.1) SEEN-BY: 2/200 201/329 212/8 235/100 251/25 254/60 61 62 261 410 620 SEEN-BY: 280/100 292/854 857 876 335/533 341/70 362/21 380/106 440/410 SEEN-BY: 442/403 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 15 Sep 97 10:53 #89 By: Steve Woodmore To: All Re: Hmm. St: ------------------------------------------------------------------------------ Hi All , hope you are having a nice day Ip-Access, DNS look-up, Name servers, etc. You are still forgetting that most fidonet nodes that use the internet to shift mail around are NOT online all of the time. I only ever connect to the internet when I am sending mail out, so looking up my IP number would be a totally pointless exercise unless you knew exactly what time I was online. However putting my IP number or my domain into a nodelist would still ensure that any Emails would reach me (they are held at my ISP until I collect them), but we need a way of letting people know that they just cannot look people like me up and get an IP session with my system whenever they want. This then comes back to having special flags in the nodelist. -=> Steve Woodmore <=- --- Terminate 5.00/Pro * Origin: I Like Snickers, In fact I am off Topic (2:440/410.1) SEEN-BY: 2/200 201/329 212/8 235/100 251/25 254/60 61 62 261 410 620 SEEN-BY: 280/100 292/854 857 876 335/533 341/70 362/21 380/106 440/410 SEEN-BY: 442/403 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 15 Sep 97 7:38 #90 By: Pedro Lima To: Denis McMahon Re: IP-access St: ------------------------------------------------------------------------------ Hello, DM> Now, makenl is probably the single most important piece of software in DM> fidonet today, and Ben Baker isn't listed in the Nodelist anymore. DM> Unless someone is going to re-write makenl, or has Ben's source code DM> and is in a position to develop the software, we've got little choice DM> but to work within the constraints of that program! IP adresses and FQDNs are already being included in the phone number field of IP nodelists in some Z2 regions. So it should be possible, even if it means trading MakeNL for another software. I don't know however what software is being used or if it has problems in generating 'normal' nodelist entries. DM> This method allows translation to a gateway telno using the first DM> three digits of the telno in the nodelist. It's a lot easier to DM> implement in current mailers than selecting a gateway according to the DM> flags. I'm not so sure. I've only used FrontDoor so far (and the TBBS mailer whose name I luckily can't recall now), but I never used the flags possibilities in the argument of most commands in ROUTE.FD, so I'm not aware of eventual limitations. Perhaps JoHo can help on this, although I know FD is only one of several possible mailers? DM> ; National numbers DM> 44-1 01 10 DM> ; IP gateways DM> 010- 01abc-defghi 10 ; vmodem gateway DM> With this translation, any packets for a vmodem address would be sent DM> to a gateway that could forward to vmodem ip nodes If I'm understanding correctly, this would mean that something listed as 010-123-123-123-123 would generate an 'ATD01abc-defghi123-123-123-123'. This might work if dialing in tone, but I reckon it could give problems if dialing in pulse. Also, this is a routing problem, not a phone translation one. By your system, and if I'm not mistaken, most software would need a complete nodelist compilation even if a temporary change is made. DM> The nodelist entries are compatible with makenl. True. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 16 Sep 97 6:15 #91 By: rafal wiosna To: Pedro Lima Re: Why IP? St: ------------------------------------------------------------------------------ þ (RAF, Sun Sep 14 1997) Pedro Lima -> rafal wiosna: rw>> WHY IP?! PL> Why not admitting both possibilities? Yep, DNS and IP should be treated equally since the first one is a synonym of the second but I prefer using CNAMEs due to flexibility. - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: You cannot ALT-CTRL-DEL a mainframe (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 16 Sep 97 14:24 #92 By: maciej grzeszczuk To: Pedro Lima Re: IP-access St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Sun, 14 Sep 97 19:30:21 +0200 Pedro Lima napisal byl w fido.ip_connect: > It would be solved by an AKA. However, one doesn't exclude the other, but aka isn't needed - as the phone field stays unmodified: if a node has phone number it cointains the valid number, if not it's -unpublished-. > i.e., it's also possible to use the FQDN associated with the Fido > address. In that case, only the flags would be listed. Let me be clear > about this, either way I think that appropriate flags should be listed. agree. > mg> for software that cannot use DNS maybe we should reconsider using > mg> some kind of U,ADRmyhost.mydomain flags? > I'm really not in favor of using flags to specify an Internet address, > for all the reasons stated before. so the dns entry. but we shouldn't touch any of the existing fields. > Precisely. Each IP-related flag has a default port associated, so > there's no need to specify it. For example, it would only be needed to > list 'U,VMD' to know that the port is 3141, or 'U,ITLN' to know the port > is 23. These flags are obviously just examples, we still have to address > this subject in more detail. sure. > mg> if a system is not reachable by emsi (only e-mail transfers of > mg> echomail) it shouldn't be listed as a node... > For me, the real difference lies in whether it's possible or not to > connect directly to the node. EMSI is only one of many possible choices. the only way to reach immediatelly fidonet mailer is emsi. smtp or fts isn't a way. 1109 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: ty jestes glupi arecki! zaraz sie na ciebie obnaz (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 SEEN-BY: 480/88 91 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 16 Sep 97 14:30 #93 By: maciej grzeszczuk To: Mats Wallin Re: IP-access St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Mon, 15 Sep 97 09:46:03 +0200 Mats Wallin napisal byl w fido.ip_connect: > I think that is a rather important fact that one should consider. We can't > expect every mailer to be updated to support this, but it would > technically be > possible to provide different drivers for most operating systems that > could make > sure that most of the mailers available today could work with the IP/domain > address in the phone number field. we shouldn't use ip/domain name address in phone number field. mailers that don't support other way to retrieve address from nodelist should use some kind of subst.lst file that contains addressess for specified nodes. > Since FTS-1 is a basic requirement for a member of FidoNet, I thought we > only > were talking about sessions that could establish FTS-1, FTS-6 or EMSI > sessions. right. 1110 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: niezle ssie... (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 SEEN-BY: 480/88 91 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 16 Sep 97 17:11 #94 By: maciej grzeszczuk To: Steve Woodmore Re: Hmm. St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Mon, 15 Sep 97 09:53:56 +0200 Steve Woodmore napisal byl w fido.ip_connect: > However putting my IP number or my domain into a nodelist would still but you don't have separate ip - it could change or be used by someone else. what then? we shouldn't include ip or domain.name into a nodelist when specified ip can be used by other systems (fe. of the same isp). > This then comes back to having special flags in the nodelist. systems that don't allow direct dialup access via modem should be pvt nodes. systems that don't allow direct access should be points, as it's impossible to immediatelly direct access them by emsi. 1115 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: rave (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 SEEN-BY: 480/88 91 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 16 Sep 97 19:49 #95 By: rafal wiosna To: maciej grzeszczuk Re: what about... St: ------------------------------------------------------------------------------ þ (IP_CONNECT, Mon Sep 15 1997) maciej grzeszczuk -> rafal wiosna: >> And the flags could be like this: U,VMODEM,TCP,PORT,xxxxx... Port >> could be added to systemname field in the form of ":numbr", like in URLs. mg> vmodem, ifcico, binkd and so on has already their own port numbers, mg> so there is no need to add one after vmodem or ifcico flag. Of course there isn't. U,PORT,xxxx flag is optional. mg> i prefer tcp-only nodes would be pvt ones - as there is no way to reach mg> them via standard modem connection. their phone number is then mg> -unpublished-. pvt nodes makes no dishonor to it's operator - he's a mg> member with the same rights as non-pvt one. What is the difference between a Pvt node or a point in this case? >> It's default but it's not mandatory. I agree with Pedro: it'd be >> more flexible to provide non-standard port number information in the >> nodelist entry when a node uses one. mg> but not in case ifcico resides on it's standard port. Sure, when it's standard no additional flags are needed. mg> what do you mean TCP? all that vmodem, ifcico, binkd use tcp for mg> transport. Nope. TCP means ifcico, VMODEM is self-explanatory. - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Samiec twoj wr¢g! Samiec twoj wr¢g!! (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 16 Sep 97 19:51 #96 By: rafal wiosna To: Anton Kuznetsov Re: IP addressing St: ------------------------------------------------------------------------------ þ (IP_CONNECT, Mon Sep 15 1997) Anton Kuznetsov -> All: AP>> ;S IP Internet Protocol. Node accepts Telnet sessions. AP>> ;S VM Vmodem capability, Host accepts Vmodem connection. AK> What about BinkD capability [BSO FTN-mailer via TCP/IP] ?.. BD? - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Official 'U Boryny' member (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 16 Sep 97 19:52 #97 By: rafal wiosna To: Mats Wallin Re: IP-connectivity St: ------------------------------------------------------------------------------ þ (IP_CONNECT, Mon Sep 15 1997) Mats Wallin -> rafal wiosna: MW> One of the reasons why they should not be points, is that a point is not MW> a member of FidoNet, they are only a user of another system. So what? The IP-world differs from the ISDN/PSTN worlds only by means of transport and accesibility [no BUSY, no costs]. I think there should be points and nodes in both "worlds". With one clause: the IP world needs definitely less nodes than PSTN/ISDN. MW> The Internet is just another carrier of data, and there is no reason MW> why one shouldn't be allowed to be a full FidoNet member just because MW> they connect via the Internet. I'm not saying IP systems should be points. I only try to convince you that not EVERYONE should became a node there since one node can serve much more systems than in Fido. - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: NOWO˜•: Urz†dzenie, kt¢re robi wszystko w ka§d† stron‘. (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 16 Sep 97 19:55 #98 By: rafal wiosna To: Pedro Lima Re: IP-access St: ------------------------------------------------------------------------------ þ (IP_CONNECT, Mon Sep 15 1997) Pedro Lima -> Denis McMahon: PL> IP adresses and FQDNs are already being included in the phone number PL> field of IP nodelists in some Z2 regions. So it should be possible, Such nodelist[s] is[are] assembled by hand and CRC is then calculated. Definitely not an option for our [TADA.WAV] beloved ZC. 8^) - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Do zatocania jeden krok (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 16 Sep 97 19:56 #99 By: rafal wiosna To: Steve Woodmore Re: Hmm. St: ------------------------------------------------------------------------------ þ (IP_CONNECT, Mon Sep 15 1997) Steve Woodmore -> All: SW> You are still forgetting that most fidonet nodes that use the internet to SW> shift mail around are NOT online all of the time. Maybe it's the thing that will part the IP world -- CM IP-systems, available via IP [not necessary via PSTN/ISDN] should become nodes and non-CM IP-systems should remain points? But OTOH why bother dialing to the internet for an ARCmail then when you're non-CM but IP? It'd be simpler just to call PSTN/ISDN node then... 8^) - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Bazgrzesz jak kura joystickiem (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 13:34 #100 By: Jan Vermeulen To: Ward Dossche Re: Test of an IP-number in the nodelist / Proposal St: Local Sent ------------------------------------------------------------------------------ * Crossposted in IP_CONNECT Quoting Ward Dossche on Tue 16 Sep 97 22:05 to Jan Vermeulen: wd> Does anyone have a technical problem with it that as a test this wd> line appears in the European version of the world-nodelist next wd> Friday? wd> wd> ,3000,IP-test,somewhere,Lothar_Behet,000-194-231-142-17,300,VM wd> wd> I think a prefix of "000-" will technically bother no-one, correct? Dutch Telecom sends the 'number not obtainable' signal. This means no cost, but it could impair the operation of my system being off hook while others try to call it. -=<[ JV ]>=- ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 16 Sep 97 21:54 #101 By: Ward Dossche To: All Re: Alberto's proposal St: ------------------------------------------------------------------------------ > ;S ,192.168.1.25:669,300,IP - port 669 accepts telnet sessions > ;S ,192.168.1.25:669,300,VM - port 669 accepts Vmodem sessions > ;S ,bbs.os2net.net,300,IP,VM - accepts telnet and Vmodem sessions > ;S ,bbs.os2net.net,300,VM - accepts Vmodem sessions only > ;S ,bbs.os2net.net,300,IP - accepts telnet sessions only I have the funny feeling that all of these entries will make MAKENL barf. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 16 Sep 97 21:56 #102 By: Ward Dossche To: All Re: Speed St: ------------------------------------------------------------------------------ Alberto proposes speed-flags ... > ;S B28 28800 kbit/s max speed > ;S B256 256 kbit/s max speed > ;S B2000 2 Mbit/s max speed Can anyone tell me what the purpose of this is? \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 7 Sep 97 23:10 #103 By: Lothar Behet To: Ward Dossche Re: Test of an IP-number in the nodelist St: ------------------------------------------------------------------------------ *** Antwort auf eine Nachricht in mymail-lb (MyMail-lb). Hello Ward! On 07 Sep 97 wrote Ward Dossche to Lothar Behet: WD> Colleagues, WD> What would be the technical argument against including an entry in the WD> next nodelist with an IP-node for test-purposes? WD> The nodenumber would be 2:2/3000 and have the following outlook. WD> ,3000,IP-test,somewhere_unimportant,Lothar_Behet,0000-194.231.142.17,3 WD> 00,U,IP I would appreciate a notice about Vmodem in that entry, otherwise it will fail :) WD> I would say as a fake-country-code this "0000-" would work pretty WD> well as it can nowhere imo be dialled as a valid-number. Is there somebody else assigned as victim (check for password protected sessions, etc.)? WD> Please, all participants, consider this as "brainstorming", not as an WD> accomplished fact. Sure :) Regards Lothar --- GoldED/2 3.00.Alpha5+ --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 16 Sep 97 22:02 #104 By: Ward Dossche To: All Re: Test of an IP-number in the nodelist / Proposal St: ------------------------------------------------------------------------------ Does anyone have a technical problem with it that as a test this line appears in the European version of the world-nodelist next Friday? ,3000,IP-test,somewhere,Lothar_Behet,000-194-231-142-17,300,VM I think a prefix of "000-" will technically bother no-one, correct? \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 21:38 #105 By: Ward Dossche To: All Re: Availability St: ------------------------------------------------------------------------------ This conf is now also available on Southern Star and PAOnline. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 21:40 #106 By: Ward Dossche To: rafal wiosna Re: IP-access St: ------------------------------------------------------------------------------ > PL> IP adresses and FQDNs are already being included in the phone number > PL> field of IP nodelists in some Z2 regions. So it should be possible, > Such nodelist[s] is[are] assembled by hand and CRC is then > calculated. Definitely not an option for our [TADA.WAV] beloved ZC. 8^) I don't see the point ... at least 1/3 of Z2 assembles their region-files by hand and do _not_ include a CRC. This has never bothered me. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 21:44 #107 By: Ward Dossche To: rafal wiosna Re: IP-connectivity St: ------------------------------------------------------------------------------ > I'm not saying IP systems should be points. I only try to > convince you that not EVERYONE should became a node there since one node > can serve much more systems than in Fido. So are you saying that soon you will give up _your_ nodenumber, or that you think it's the guy next-door who should give up _his_ nodenumber? I see no restriction why someone shouldn't be allowed to operate an IP-node. Is there an ego-problem involved maybe? \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 9:55 #108 By: Joaquim Homrighausen To: rafal wiosna Re: IP-connectivity St: ------------------------------------------------------------------------------ JH>> Not if you as the first string in your dial translation JH>> table put: JH>> # ;Internet connection JH>> That oughta work.. rw> So it won't dial at all? Uh, no.. so that it will first of all strip the '#' (if you don't want it, otherwise just put ##), and second (more importantly) not prefix the number with the international dial sequence. %JoHo% joho@defsol.se * Origin: Definite Solutions ~/Stockholm, Sweden (2:201/330.1) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/27 212/8 235/100 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 341/70 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 16 Sep 97 23:38 #109 By: maciej grzeszczuk To: rafal wiosna Re: what about... St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Tue, 16 Sep 97 17:49:37 +0200 rafal wiosna napisal byl w fido.ip_connect: > mg> i prefer tcp-only nodes would be pvt ones - as there is no way to reach > mg> them via standard modem connection. their phone number is then > mg> -unpublished-. pvt nodes makes no dishonor to it's operator - he's a > mg> member with the same rights as non-pvt one. > What is the difference between a Pvt node or a point in this case? someone told that the point isn't a member of fidonet. pvt node is. pvt nodes should be nodes that are available cm via tcp only. systems that are not available via emsi (tcp or standard modem connection) should stay with point address. > mg> what do you mean TCP? all that vmodem, ifcico, binkd use tcp for > mg> transport. > Nope. TCP means ifcico, VMODEM is self-explanatory. i suggest IFC (IF) stands for ifcico, as it's clear and couldn't be misunderstood. 66 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: kanapeczke? (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 SEEN-BY: 480/88 91 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 12:36 #110 By: Pedro Lima To: rafal wiosna Re: Why IP? St: ------------------------------------------------------------------------------ rw> Yep, DNS and IP should be treated equally Ok. :-) Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 18 Sep 97 1:14 #111 By: Pedro Lima To: maciej grzeszczuk Re: IP-access St: ------------------------------------------------------------------------------ Hello, mg> but aka isn't needed - as the phone field stays unmodified: if a node mg> has phone number it cointains the valid number, if not it's mg> -unpublished-. The flag fields are limited in size, which may place problems if we're going to allow several network technologies. Also, I understand most of the existant software already works with the address in the phone field. Furthermore, I think using the phone field to contain the address is the natural way. mg> so the dns entry. but we shouldn't touch any of the existing fields. I see why not, as long as general compatibility is maintained. mg> the only way to reach immediatelly fidonet mailer is emsi. smtp or fts mg> isn't a way. Why not? Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 12:48 #112 By: Pedro Lima To: rafal wiosna Re: IP-access St: ------------------------------------------------------------------------------ rw> Such nodelist[s] is[are] assembled by hand and CRC is then rw> calculated. Definitely not an option for our [TADA.WAV] beloved ZC. rw> 8^) :-) I was under the impression that a specific software was being used in R50 for it? Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 18 Sep 97 9:38 #113 By: Anton Kuznetsov To: Rafal Wiosna Re: IP addressing St: ------------------------------------------------------------------------------ Buenas dias, amigo Rafal Wiosna ! Replying to a message of rafal wiosna to Anton Kuznetsov [16 Sep 97, 19:51]: AP>>> ;S IP Internet Protocol. Node accepts Telnet sessions. AP>>> ;S VM Vmodem capability, Host accepts Vmodem connection. AK>> What about BinkD capability [BSO FTN-mailer via TCP/IP] ?.. rw> BD? ************************************************************************* * The official list of R46 nodes' IP-connectivity. * ************************************************************************* Valid flags are: TCP - mandatory for any ip-node. VMP - Vmodem capability (default port 3141). TEL - Telnet capability (default port 23). IFC - ifcico capability (default port 60179). BND - BinkD capability. U,xxxxx - marks non-standard port number, relates to first capability flag in the line (only one non-standard port can be lested). U,"foo.bar" - marks corresponding domain address (temporarily). [ ... ] Anthony. Donetsk, Ukraine [The 465/70 team] [T-Mail team] --- E-mail: tony@tony.donetsk.ua * Origin: Everything you say may be used against you. (2:465/50) SEEN-BY: 50/0 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 450/97 460/111 461/35 463/94 SEEN-BY: 463/690 465/11 50 70 85 87 467/67 480/33 2432/200 4625/1 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 19:25 #114 By: Rune Johansen To: Mats Wallin Re: IP-access St: ------------------------------------------------------------------------------ >> Hmm, gotta check up on the availability of it at 24/888, or >> maybe at 20/11. >If you can't find it either of these system, let me know, and I'll make >sure t > send it to 20/11. Got it here now, through you :-)) >> Example (I don't know what port VModem has got, but I'll use >> #505) if you have Vmodem and you have telnet cap. in vmodem: > I must admit that I thought that VModem used port 23 for telnet sessions, > and > the VModem port for VModem sessions? Guess it is possible to change that, > but >are you sure you can use the same port number for different protocols? I >would > assumed that it wasn't possible, but I might be wrong. Hmm, I don't know if I have answered this one before but.. If the daemon listening to the port can differentiate between the setup messages the client is sending, you can have different protocols on the same port. If it cannot, you have to have different ports for different protocols. That's what the port number really are: Pointers to what application you are trying to reach. So, when I, in my TCP packet says: "Connect to port 23", it means that I want to access the application that is on port 23, and that is, according to the table of well-known port numbers (RFC1700 last time I looked), the telnet application. --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 341/70 362/21 380/106 SEEN-BY: 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 19:35 #115 By: Rune Johansen To: Mats Wallin Re: IP-access St: ------------------------------------------------------------------------------ > True, but if you attempt to use the same port number for both services, how > would VModem know which service you're requesting? I have always thought > that > it was the portnumber that told the server how to proceed. There is nothing > special a client program should send during a Telnet session to indicate > that > it is a Telnet session. Been looking a little on different rfc's now.. No, you can't have two different protocols on the same port number. It would be up to the server to distinguish a telnet protocol from a VModem protocol, by having some kind of "I want VModem" in the negotiation after the TCP sessions has been established, and falling back to telnet protocol if that does not come in. I don't know the VModem _protocol_, so I cannot say wether it can or not. --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 341/70 362/21 380/106 SEEN-BY: 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 19:39 #116 By: Rune Johansen To: rafal wiosna Re: Why IP? St: ------------------------------------------------------------------------------ > Canonical addresses are much better by being more flexible. If my > computer is moved from my office to the server room with a network sharing > diffrent subnet, the IP-based nodelist entry immediatelly becomes obsolete. > With a refresh period of 604800 seconds [7 days] of our "Fido DNS" 8^) it's >nearly impossible to provide ACTUAL data on internet-based hosts by >publishing > their IP number. Using cannonical names it's up to the net part to resolve > it Even though the NS you looked at has got a 7 day refresh, does not mean that you can have a delegated part of the DNS on your server, with much less refresh time. --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 341/70 362/21 380/106 SEEN-BY: 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 19:50 #117 By: Rune Johansen To: Pedro Lima Re: IP-access St: ------------------------------------------------------------------------------ > Ah, then I wasn't completely mistaken, they're working similarly to a > secondary name server. This also allows for a FQDN to be attributed to a > variant IP. There's always the risk of the cache not being updated at > the time someone tries to connect, but it's minimal. Also, many of those > connections would work with authentication procedures (like session > passwords), so I think it's perfectly feasible. Yes. And then people think "We'll just set a refresh time of 1 at the fidonet.org subdomain". No. What you would do is like this: f20.n210.z2.fidonet.org CNAME rune.dyndns.com So, when you ask for my IP address, you get a "Canonical Name" record thrown back, and have to ask the dyndns.com server instead. Since the refresh time of the fidonet NS is not short, you will have this CNAME info cached, but hav to ask dyndns each time (since it has got a short timeout). > I took a peek at http://www.dyndns.com, and even downloaded the ELF > binary for Linux, but I couldn't find the source. Nonetheless, this > should actually be a simple program. It surely would be simple. I don't know why they haven't put out the source, but I would suspect that it is because of their (so far) unique business position, and also the fear of people spoofing the dyndns messages to make it point elsewhere... There are a lot of bad people in the world... --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 341/70 362/21 380/106 SEEN-BY: 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 19:58 #118 By: Rune Johansen To: Steve Woodmore Re: Hmm. St: ------------------------------------------------------------------------------ > Hi All , hope you are having a nice day > Ip-Access, DNS look-up, Name servers, etc. > You are still forgetting that most fidonet nodes that use the internet to > shift mail around are NOT online all of the time. True, that's why we have to differentiate between two types of internet-as-carrier systems. One group is the systems that can be accessed directly at some point (either 24x7 or via some kind of dynamically assigned domain name to IP mapping). The other group is the one that receives and probably sends their mail via other kind of applications, like email-attached files. > I only ever connect to the internet when I am sending mail out, so looking > up my IP number would be a totally pointless exercise unless you knew > exactly what time I was online. True. If you have fixed times you are online, the Txx flag could be used. If you on the other hand is not online at fixed times, there would be much that should snap together, to make a sending node and you being online at the same time. So, for your kind of system, it would be better to use the "other" method of receiving and sending mail; attachment-by-email or some other store-and-forward kind of application. > This then comes back to having special flags in the nodelist. Most of the discussion here is based upon the existance of the first kind of nodes. If we just can find a way that could be politically accepted, and that doesn't break too much software, then we could look into a way of telling nodes how store-and-forward-reliant nodes can be directly addressed in the nodelist. --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 341/70 362/21 380/106 SEEN-BY: 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 20:21 #119 By: Rune Johansen To: All Re: Observations and questions St: ------------------------------------------------------------------------------ Hi all. I have just read (and replied to) a few of the messages in this echo. To me it looks like there are some points that need to be investigated by the people that has the software available, and to differentiate on a few things: MakeNL: Some people report that MakeNL spits on using non-standard phone numbers, like "bbs.fqdn.com". Others say that they have nodelists using those addresses. To me it looks like this is a configuration issue in the MakeNL program. Can someone that has this working please tell all of us how it is configured? We have to consider the fact that MakeNL is not being used in a Fidonet basis by other people than those _producing_ nodelist segments. They are locallized in a hierarchy, with a ZC on the top. It should not be too hard for a ZC to tell it's uplinks for the segments HOW to configure their MakeNL to accept such strings in the phone number field. That goes for other types of phone numbers too. These uplinks can then again tell their uplinks etc. Nodelist compilers: It is important to se the difference between nodelist MAKERS and nodelist COMPILERS. The compilers is being used to let a mailer get a nodelist entry in a understandable format. Since the nodelist compilers are able to fiddle with the actual phone number you are to call, it can also be talked into making IP or DNS addresses undialable, by configuration, if you have a PSTN only mailer. Other things: I am not fond of making these things reliant on FIDONET or the fidonet.org domain name. Sure, we can say "We are fidonet, forget about the rest", but I don't think that's any fair at all. There ARE a LOT of Fidonet Technology based networks out there, and we should not KILL them by saying "All domain name requests should be done against a built-up address on the fidonet.org domain". That's why we HAVE to allow for IP addresses AND fqdn symbolic addresses in the nodelist. The nodelist is the "publication" that glues us together has been said for many years. It seems to me that if it is not being allowed for a change, and that some people actually has to do something to their software, it will be the publication that breaks us up. In my network (2:210) I have a node (2:210/30) that has got 44 nodes only capable of IP, and one node with a phone directly attached, due to the fact that it had to, or not be in Fidonet as a node. It is reachable at telnet://fix.no. I have been exchanging mail with it via IP for about two years now, via private insertions in the nodelist, supporting the userflag "IP". We have to allow for exceptions to the "diallable by PSTN" rule to make Fidonet Technology move forward. I'll come with more viewpoints later.. :-) --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 341/70 362/21 380/106 SEEN-BY: 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 23:47 #120 By: rafal wiosna To: Ward Dossche Re: IP-access St: ------------------------------------------------------------------------------ þ (RAF, Wed Sep 17 1997) Ward Dossche -> rafal wiosna: WD> I don't see the point ... at least 1/3 of Z2 assembles their region-files WD> by hand and do _not_ include a CRC. This has never bothered me. But it's not the proper way to do it, is it?... - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Pho c0? (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 23:48 #121 By: rafal wiosna To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ þ (RAF, Wed Sep 17 1997) Ward Dossche -> rafal wiosna: WD> So are you saying that soon you will give up _your_ nodenumber, or that WD> you think it's the guy next-door who should give up _his_ nodenumber? I would if I didn't serve 20+ systems. It's only eating my money. For example -- by dialing international. 8^) WD> I see no restriction why someone shouldn't be allowed to operate an WD> IP-node. Is there an ego-problem involved maybe? Ok, let's suppose your way. Should we follow the USUAL procedure of applying for a node? Should it be done by NC or someone actually connected to the IP-world? - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Przera§ony pilot poczu’, §e ma obci‘ty ogon (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 17 Sep 97 23:50 #122 By: rafal wiosna To: Joaquim Homrighausen Re: IP-connectivity St: ------------------------------------------------------------------------------ þ (RAF, Wed Sep 17 1997) Joaquim Homrighausen -> rafal wiosna: JH>>> Not if you as the first string in your dial translation JH>>> table put: JH> Uh, no.. so that it will first of all strip the '#' (if you don't JH> want it, otherwise just put ##), and second (more JH> importantly) not prefix the number with the international dial JH> sequence. So it will still dial some old lady who's phone number is the same as the first seven digits of node's IP?... - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: O! Przyjecha’ do nas Pan Obraz Kontrolny! (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 18 Sep 97 14:11 #123 By: maciej grzeszczuk To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Wed, 17 Sep 97 20:44:59 +0200 Ward Dossche napisal byl w fido.ip_connect: > I see no restriction why someone shouldn't be allowed to operate an > IP-node. Is > there an ego-problem involved maybe? ip node is ok, if it's really ip node - i meant if it's accessible by tcp all the time. smtp nodes and ftp ones shouldn't be treated as nodes, as there is no way to reach them directly. do you know what i mean? 1118 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: grupa narkotyczno-zboczeniowa (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 SEEN-BY: 480/88 91 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 19 Sep 97 2:07 #124 By: Ward Dossche To: All Re: It's in ... St: ------------------------------------------------------------------------------ Dear all, > ,3000,IP-test,somewhere,Lothar_Behet,000-194-231-142-17,300,VM 2:2/3000 made it in this week's world-nodelist in zone-2. The first IP-node ever. To not possibly disturb/annoy the other zones this entry will not appear elsewhere. The fake-country-prefix is "000", "VM" is a non-approved-flag to signal VModem. This entry is for test-purposes only and eventually will not continue to be listed in this location of the nodelist. If during the test-phase other nodes want to be listed as IP-node in the top-section of the nodelist then these people will have to netmail me. What I need for now is a static IP-number and a proposal for a flag signaling the capability if it isn't VModem. Please report on how it works, or maybe doesn't. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 18 Sep 97 18:17 #125 By: Jan Ceuleers To: Denis McMahon Re: IP-connectivity St: ------------------------------------------------------------------------------ I quote Denis McMahon: DM> No, you are misunderstanding dial translation as applied to DM> the nodelist. I don't think I am, but I'm willing to be convinced. DM> At the moment, a telno in the nodelist consists of: DM> DM> "valid-country-code"-"national-code"-"national-number" DM> DM> Nodelist compilers translate "valid-country-code" by either DM> replacing it with "national-prefix" (if own-country-code) or DM> by prefixing it with "international-access-code" if (not DM> own-country-code). DM> DM> Additional translations may be done for "ld-access-code" and DM> national-code = own-national-code. Agreed. DM> At this translation level (the translation applied to DM> national code) the proposed "invalid-national-code" used to DM> identify non analogue modem systems can be translated to a I need to interrupt you here. What is proposed is not an "invalid-natio "invalid-national-code" but an "invalid-country-code". With the translations you described above, a nodelist compiler (or whatever else reads the nodelist) will compare the first few digits of the listed number of a non-modem node with own-country-code, establish that they differ and hence prepend the international access code (011 in North America, 00 in most of Europe etc.). Therefore, a node that is not explicitly configured to be able to call IP nodes will attempt to originate calls to a non-existent number, which is fine. His fortune is being protected. DM> Instead of trying to invent problems where none exist, why DM> don't you look for a solution instead? I'm sorry, but I can't see how you arrive at the conclusion that I'm somehow not being constructive. I have proposed a solution: list IPv4 addresses in the nodelist in the telephone number field, by prepending them with 01, like so: ,999,Region_29_Igate,Antwerpen,Jan_Ceuleers,01-138-203-224-167,9600,etc Your reaction was that this cannot work since it clashes with carrier selection prefixes in North America, to which I replied that IMHO they do not (since carrier selection prefixes are of the form 10xxx(x), not of the form 011-0xxx(x). Example: AT&T's carrier selection prefix is 10ATT or 10288). I had one remaining doubt as to the feasibility of my proposal due to a remark of yours: you hinted that 011-0 puts the caller through to the international operator. Is that indeed the case? If it is, does the call to the operator go through immediately or only after a 10-second-or-so timeout (or possibly after pressing #)? Jan --- timEd-A10a * Origin: Experimenter Board, Antwerp, Belgium (2:292/857) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 19 Sep 97 0:23 #126 By: Denis McMahon To: Pedro Lima Re: IP-access St: ------------------------------------------------------------------------------ Pedro Lima wrote to Denis McMahon: DM> ; National numbers DM> 44-1 01 10 DM> ; IP gateways DM> 010- 01abc-defghi 10 ; vmodem gateway DM> With this translation, any packets for a vmodem address would be sent DM> to a gateway that could forward to vmodem ip nodes PL> If I'm understanding correctly, this would mean that something PL> listed as 010-123-123-123-123 would generate an PL> 'ATD01abc-defghi123-123-123-123'. This might work if dialing in PL> tone, but I reckon it could give problems if dialing in pulse. Hmm, a valid technical reason to bounce my proposal. :-( I'm glad it's been found now, the next question is whether to explore the telno field route further or not? PL> Also, this is a routing problem, not a phone translation one. By PL> your system, and if I'm not mistaken, most software would need a PL> complete nodelist compilation even if a temporary change is made. Most software needs a complete nodelist recompilation if any nodes details change. It is to be expected that gateway telno details will remain valid for the duration of each weeks nodelist, ditto for ip addresses and host/sub-domain/domain data. Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 19 Sep 97 0:35 #127 By: Denis McMahon To: Michiel van der Vlist Re: IP-connectivity St: ------------------------------------------------------------------------------ * Original message posted in: ENET.SYSOP. * Crossposted in: IP_CONNECT. Michiel van der Vlist wrote to Denis McMahon: > I think you'll find it's an ITU diktat that country codes > may not begin with 0. MvdV> Maybe. But that is no guarantee, it will never be used for MvdV> anything at all. It could be used for special access codes or MvdV> whatever. It is generally speaking a bad idea to use anything MvdV> marked as "reserved" for ones own purposes. However, you are ignoring the way nodelist telno translation works. Nodelist telno translation works on the basis that the first section of the telno is the country code. The country code is then either removed and replaced (for own country), or prefixed with idd code (other country). I am merely proposing a third level of translation on the country code field, namely non-analogue-connection/use-gateway, where gateway may be own-system in relevant non-analogue mode (if capable) or another system, identified by a telno. MvdV> How about using one of the 'digits' A-D as the special prefix? MvdV> They are (or should be) acceptable for all dialing software, MvdV> after all they _are_ dialable by any DTMF capable modem. Makenl will not allow characters in the telno field, this means that to use a telno based translation we must use numeric data. MvdV> Since the 'digits' A-D do not appear on an ordinary phone, it is MvdV> extremely unlikely that their use fore a special prefix to denote MvdV> IP nodes will ever interfere with normal operations of the PSTN. Since the digits a to d are not accepted in the telno field by makenl, it's impractical to use them for this purpose. However, if a workable means of implementing numeric non telno type addressing in the telno field is possible, it does seem that it will need to be prefixed other than with a valid country code. Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 341/70 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 18 Sep 97 5:57 (Rcvd: Mon 22 Sep 97 21:16) #128 By: Pedro Lima To: Jan Vermeulen Re: Test of an IP-number in the nodelist / Proposal St: Rcvd Replied ------------------------------------------------------------------------------ wd> ,3000,IP-test,somewhere,Lothar_Behet,000-194-231-142-17,300,VM JV> Dutch Telecom sends the 'number not obtainable' signal. This means JV> no cost, but it could impair the operation of my system being off hook JV> while others try to call it. Can't you cancel the attempt from the start by configuring the 'VM' flag? Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 19 Sep 97 10:43 #129 By: rafal wiosna To: Pedro Lima Re: IP-access St: ------------------------------------------------------------------------------ þ (RAF, Wed Sep 17 1997) Pedro Lima -> rafal wiosna: >: -) I was under the impression that a specific software was being used PL> in R50 for it? PUTNLCRC.EXE? 8^) - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Frankie says: RELAX don't do it! (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 19 Sep 97 10:44 #130 By: rafal wiosna To: Rune Johansen Re: Why IP? St: ------------------------------------------------------------------------------ þ (RAF, Wed Sep 17 1997) Rune Johansen -> rafal wiosna: RJ> Even though the NS you looked at has got a 7 day refresh, does not RJ> mean that you can have a delegated part of the DNS on your server, RJ> with much less refresh time. Yes, I can. But even when I have R48 DNS set on my computer at work [or even on our main NS machine] the refresh plus PROPAGATION rates are rather unacceptable. And it's a hell lot of work for someone to put all then [new] points in there. I think the name-resolving problem should be DISTRIBUTED [CNAME in the nodelist], not CENTRALIZED [*.fidonet.org entry] even by sacryfying nodelist size. - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Zwalczaj przeludnienie -- zabij bociana! (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 19 Sep 97 12:17 #131 By: Mats Wallin To: rafal wiosna Re: IP-connectivity St: ------------------------------------------------------------------------------ rw> So what? The IP-world differs from the ISDN/PSTN rw> worlds only by means of transport and accesibility [no BUSY, rw> no costs]. I think there should be points and nodes in both rw> "worlds". With one clause: the IP world needs definitely less rw> nodes than PSTN/ISDN. Of course there should be points in both worlds. But points is something that in FidoNet policy is completely different from nodes. Since a point is not a member of FidoNet, he does not have any rights, he can not candidate for any *C hats, many systems do not allowe points to request files, etc. Anyone that fulfill the basic requirement of a FidoNet node should be allowed to become a node, if he so wishes. Those that prefer to be a point, or does not fulfill the requirements can become a points. I see no reason whatsover why we should try to lower the amount of nodes in FidoNet, IMHO, it should be the opposite. The more nodes the better. rw> I'm not saying IP systems should be points. I only rw> try to convince you that not EVERYONE should became a node rw> there since one node can serve much more systems than in rw> Fido. But we don't have the nodes in FidoNet to serve other systems. They do of course serve their BBS users and points, if they have any, but that is not a requirement. Mats mw@defsol.se --- * Origin: Definite Solutions, Stockholm, Sweden (2:201/329.11) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/27 212/8 235/100 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 341/70 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 19 Sep 97 13:24 #132 By: Mats Wallin To: maciej grzeszczuk Re: IP-access St: ------------------------------------------------------------------------------ mg> we shouldn't use ip/domain name address in phone number mg> field. mailers that don't support other way to retrieve mg> address from nodelist should use some kind of subst.lst file mg> that contains addressess for specified nodes. Using a subst.lst file, or anything similiar, prevents all current mailers from being used over the Internet (well, with the exception of those that alreayd supports it of course). Including the IP/DNS address in the phone number field would make it possible for anyone to use their mailer via the Internet (if the mailer uses some kind of driver for the serial port, which can be replaced with a driver for the Internet). Preventing all of todays mailers from using Internet as a carrier is not a step in the correct direction, and would IMHO definitely prevent or delay the use of Internet as a carrier. Mats mw@defsol.se --- * Origin: Definite Solutions, Stockholm, Sweden (2:201/329.11) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/27 212/8 235/100 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 341/70 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 19 Sep 97 5:06 #133 By: Pedro Lima To: Rune Johansen Re: IP-access St: ------------------------------------------------------------------------------ Hi, RJ> So, when you ask for my IP address, you get a "Canonical Name" record RJ> thrown back, and have to ask the dyndns.com server instead. Yep. Unless ns.z?.fidonet.org offered a similar capability, which would actually be nice. But as you say, there may be serious security problems involved. RJ> I would suspect that it is because of their (so far) unique RJ> business position, It may be that business has something to do with it, but I'm not so sure it's a unique position, take a look at 'http://www.nws.net'. (hope I got the URL right) RJ> and also the fear of people spoofing the RJ> dyndns messages to make it point elsewhere... There are a lot of bad RJ> people in the world... It's quite likely that's the main reason. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 19 Sep 97 4:59 #134 By: Pedro Lima To: Denis McMahon Re: IP-access St: ------------------------------------------------------------------------------ Hello, DM> Hmm, a valid technical reason to bounce my proposal. :-( I'm sorry. DM> Most software needs a complete nodelist recompilation if any nodes DM> details change. Does a route configuration change implies a nodelist recompilation with most software? If I'm reading my mailer's docs correctly, I might be able to route mail for nodes carrying a certain flag to a certain system without needing to recompile, but I don't know if most software behaves that way. DM> It is to be expected that gateway telno details will remain valid for DM> the duration of each weeks nodelist, ditto for ip addresses and DM> host/sub-domain/domain data. Naturally, like with any other node it's expected that the information in the nodelist is correct and up-to-date. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 19 Sep 97 8:14 #135 By: Steve Woodmore To: Rune Johansen Re: Hmm. St: ------------------------------------------------------------------------------ * Replying to a message in : SAVEAREA Hi Rune Johansen, hope you are having a nice day 17-Sep-97 19:58:54, Rune Johansen wrote to Steve Woodmore Subject: Hmm. RJ> Most of the discussion here is based upon the existence of the RJ> first kind of nodes. If we just can find a way that could be RJ> politically accepted, and that doesn't break too much software, RJ> then we could look into a way of telling nodes how RJ> store-and-forward-reliant nodes can be directly addressed in the RJ> nodelist. This is what I am trying to avoid. MOST nodes that have internet access do so the same way that I do, its only a minority who have full time accounts. I really do not think Fido should be making the same mistakes as it always has, stop thinking about the minority, and concentrate on the majority first. -=> Steve Woodmore <=- --- Terminate 5.00/Pro * Origin: Terminate has most advertising origin lines ;-) (2:440/410.1) SEEN-BY: 2/200 201/329 212/8 235/100 251/25 254/60 61 62 261 410 620 SEEN-BY: 280/100 292/854 857 876 335/533 341/70 362/21 380/106 440/410 SEEN-BY: 442/403 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 19 Sep 97 8:39 #136 By: Steve Woodmore To: maciej grzeszczuk Re: IP-connectivity St: ------------------------------------------------------------------------------ Hi maciej grzeszczuk, hope you are having a nice day 18-Sep-97 14:11:03, maciej grzeszczuk wrote to Ward Dossche Subject: Re: IP-connectivity mg> ip node is ok, if it's really ip node - i meant if it's accessible mg> by tcp all the time. smtp nodes and ftp ones shouldn't be treated mg> as nodes, as there is no way to reach them directly. do you know mg> what i mean? Incorrect, My system is online 24hrs/day via PSTN. You can also send internet email to my IP address 24 hrs/day What you cannot do is to connect to my system Via the internet, unless I am online. -=> Steve Woodmore <=- --- Terminate 5.00/Pro * Origin: Terminate point system, the easiest in the world! (2:440/410.1) SEEN-BY: 2/200 201/329 212/8 235/100 251/25 254/60 61 62 261 410 620 SEEN-BY: 280/100 292/854 857 876 335/533 341/70 362/21 380/106 440/410 SEEN-BY: 442/403 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 19 Sep 97 8:45 #137 By: Steve Woodmore To: Ward Dossche Re: It's in ... St: ------------------------------------------------------------------------------ Hi Ward Dossche, hope you are having a nice day 19-Sep-97 02:07:14, Ward Dossche wrote to All Subject: It's in ... WD> nodelist then these people will have to netmail me. What I need WD> for now is a static IP-number and a proposal for a flag signaling WD> the capability if it isn't VModem. At your suggestion I did exactly that 7 weeks ago. -=> Steve Woodmore <=- --- Terminate 5.00/Pro * Origin: Terminate point system, the easiest in the world! (2:440/410.1) SEEN-BY: 2/200 201/329 212/8 235/100 251/25 254/60 61 62 261 410 620 SEEN-BY: 280/100 292/854 857 876 335/533 341/70 362/21 380/106 440/410 SEEN-BY: 442/403 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 19 Sep 97 23:34 #138 By: rafal wiosna To: Mats Wallin Re: IP-connectivity St: ------------------------------------------------------------------------------ þ (RAF, Fri Sep 19 1997) Mats Wallin -> rafal wiosna: MW> But points is something that in FidoNet policy is completely MW> different from nodes. Since a point is not a member of FidoNet, he MW> does not have any rights, he can not candidate for any *C hats, many So what? Is that or the mail the main reason you Fido?... MW> systems do not allowe points to request files, etc. Unlisted, that is. MW> Anyone that fulfill the basic requirement of a FidoNet node should be MW> allowed to become a node, if he so wishes. Those that prefer to be a If that's true no IP-only node should be allowed. They can't be reached during zone's ZMH. MW> I see no reason whatsover why we should try to lower the amount of MW> nodes in FidoNet, IMHO, it should be the opposite. The more nodes the MW> better. I do not try to force anything. Just to state my point of view -- this technology could reduce the number of nodes needed as long as technical needs are considered. By all means I won't block any IP-only node in my net [480]. Well, that is untill the standard for IP-only nodes nodelist-entries is official, at least in Z2. MW> But we don't have the nodes in FidoNet to serve other systems. They do of MW> course serve their BBS users and points, if they have any, but that is MW> not a requirement. So, I ask again: what's the point of becoming such a leaf node then? Only to be able to vote? - Rafa’ "jejku jak extra" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: * The Night Santa Went Crazy * (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 #139 By: Jan Vermeulen To: Denis McMahon Re: IP-connectivity St: Local Sent ------------------------------------------------------------------------------ * Crossposted ENET.SYSOP Quoting Denis McMahon on Fri 19 Sep 97 0:28 to Michiel van der Vlist: dm> Nodelist telno translation works on the basis that the first section of dm> the telno is the country code. The country code is then either removed dm> and replaced (for own country), or prefixed with idd code (other dm> country). dm> I am merely proposing a third level of translation on the country code dm> field, namely non-analogue-connection/use-gateway, where gateway may be dm> own-system in relevant non-analogue mode (if capable) or another system, dm> identified by a telno. There is a drawback to that solution, however ingenious it is, which will not affect ip-access nodes, but may affect telco oriented nodes to a greater effect than you'd imagine. Consider the case of a node that has some important functions to perform as a hub, a host, a mail mover, or a file mover that for some reason attempts to deliver a crash mail to 2:2/3000 (or any other node configured that way) and that has not implemented your third level of translation as either its sysop may not know about it or the software used does not allow this or other means of making the destination undialable. Quite naturally, that node will never obtain a connection as some automated telco device will try to tell it that the destination is not obtainable. Unfortunately, the node's modem knows zilch about such signals (as is the case for the majority of those engins) and will send back a NO CARRIER line to the mailer, after it has been sitting on the line for a certain time (between 30 and 90 seconds, let's assume an average of 60 seconds). After a configured delay (say 60 seconds) the mailer will quite happily execute a retry, until a configured maximum of retries, if any, has been reached. All this will restart as soon as the node has received an incoming call. The bottom line is that on average this node will have been unobtainable for incoming calls during a larger part of the day due to no fault of its own. This is the main reason why I preconize using the system name field for the IP address. The additional benefits are that (1) those who have dynamic IP address can list their site name and (2) those who have a telco line -- hopefully the majority, for connectivity's sake! -- can list both in the same nodelist line. -=<[ JV ]>=- * Origin: The Poor Man's Workstation (2:280/100) ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 20 Sep 97 11:40 #140 By: Pedro Lima To: rafal wiosna Re: IP-access St: ------------------------------------------------------------------------------ rw> PUTNLCRC.EXE? 8^) May be... :-) PL --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 341/70 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 21:52 #141 By: Ward Dossche To: All Re: MAKENL ! ? St: ------------------------------------------------------------------------------ Colleagues, I've already had some remarks re: the hyphens in the IP-number of the test-node (only in the world-nodelist of zone-2 !!). People request me to put dots there, this is what makenl says : ************************* QUOTE ************************* Begin processing NODELIST -- 21:26, Sunday, September 21, 1997 ,3000,IP-test,somewhere,Lothar_Behet,000-194.231.142.17,300,VM D:\MAKENL\CTL-FILE\TEST.CTL -- Invalid phone number -- "000-194.231.142.17" ************************ UNQUOTE ************************ Which is fully compliant with FTS-0005. So, 2 things are needed: 1) an update of FTS-0005, and I welcome the input of the newt FTSC in this; 2) nodelist-software that will produce nodelists/-diffs the way MAKENL does but a bit more flexible. Anyone ambitious to rewrite MAKENL? \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 20 Sep 97 1:09 #142 By: Rune Johansen To: All Re: Flags in nodelist St: ------------------------------------------------------------------------------ Hi all. I've been "thinking" a litte here :-)) We are in the middle of an discussion about how we should mangle the phone number field, so that ordinary PSTN nodes would be restricted from dialling a IP/DNS address. We are assuming that noone other than IP nodes are reading the flags of a nodelist entry (yeah, we are assuming that there are only ignoarnt nerds out there). Question 1: Does there exist a way to make a node undiallable? Answer 1: Yes, by making it a Pvt node, with "-Unpublished-" phone number. Question 2: If that's the only way, how can we make sure that a IP-only node can be inserted in the nodelist? Answer 2: By NOT putting the address in the phone number field. Question 3: Where should we put it then? Answer 3: Not in the node name field, as it would disrupt the name of the node, so we MUST put it in the flags field. Question 4: How long can a flag be then? And how should it be denoted? Answer 4: I don't know how long a flag can be, neither how long the total part of the flaglist can be. Somebody else must supply this information. So, by the reasoning over, we must put the address in the (for the time being) userflag fields. It is actually easier to change the IP-capable software of those using it today, than to make the rest of Fidonet actually READ the flags in their nodelist compiler, and take into consideration that they won't be able to dial all of them. After all, even though people in europe or america can live with (and don't get any cost by) dialling a foreignprefix+000 as a default conversion rule, we don't know how it would affect nodes in, let's say Uganda. On the other hand: If we can say that the flag IP denotes that the phone number field is a IP/DNS address, it would be complient with existsing software. If people don't care about what happens in the society they are a member of (read: Fidonet) and dont implement calling restrictions to nodes that have got this flag (and ISDN-only nodes), they are up a well-known creek (if they dial these numbers unconditionally). One thing is what people want to be able to (call anyone crash), another thing is what they actually do (route mail to non-agreed nodes). Example: I have been in Fidonet since 1994. Have I ever gotten a crash mail from nodes outside zone 2? No. Have I gotten routed netmails? Lots. The crash netmails I do have gotten, are from people that checks if they actually can crash to my system, or the have flag checkers that make sure their protocols are comaptible with mine. So, if I put up a node on my system that is IP-capable, it won't hurt anyone a bit, except a few extra bytes in a already decreasing nodelist. My suggestion is therefor: Go with IP flag, and let ignorant people that don't pay attention to extended connectivity sail their own sea, up shit creek. Rune (Netmails to me can be crashed by IP/Telnet/EMSI via 2:210/27 at 195.204.30.254/runejpc.cinet.no) --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 0:30 #143 By: Rune Johansen To: rafal wiosna Re: Why IP? St: ------------------------------------------------------------------------------ >> Even though the NS you looked at has got a 7 day refresh, does not >> mean that you can have a delegated part of the DNS on your server, >> with much less refresh time. > Yes, I can. But even when I have R48 DNS set on my computer at work [or > even on our main NS machine] the refresh plus PROPAGATION rates are rather > unacceptable. I see now I had a typo there. I meant: "...does not mean you can't" :-)) --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 0:49 #144 By: Rune Johansen To: Steve Woodmore Re: Hmm. St: ------------------------------------------------------------------------------ >> Most of the discussion here is based upon the existence of the >> first kind of nodes. If we just can find a way that could be >> politically accepted, and that doesn't break too much software, >> then we could look into a way of telling nodes how >> store-and-forward-reliant nodes can be directly addressed in the >> nodelist. > This is what I am trying to avoid. > MOST nodes that have internet access do so the same way that I do, its only > a minority who have full time accounts. > I really do not think Fido should be making the same mistakes as it always > has, stop thinking about the minority, and concentrate on the majority > first. OK, let's look at what the email type of fidonet-distribution does: It needs a email address to send the attachments to. If you have only one email address that you can receive at, you have to differ those mails from your regular email, supposedly by using some kind of special subject. Both of these should be denoted in the nodelist, to make people able to send "crashmail" via the email-attachment method. Then, a kind of flag that denotes the method of receiving mail being used. Example: I have one email account that I can receive mail to, and that I can fetch via the POP3 protocol. Let's say that the account is "rune@some.net". Since I already receive mail on that account, I want the mail to have a subject of "FIDONET MAIL", to be able to make my receiving software differing the fidonet mail bundles from my personal internet-type mail. I use a software that is capable of recognizing MIME-encoded attachments, and other types of attachments, encoded with both base64 and uuencode (to make it simpler). How should this information (rune@some.net and "FIDONET MAIL") be denoted in the nodelist? As you can see, this kind of information can be both quite different for the nodes using it, and it can be difficult to let people get this information automatically. It can also be used for a denial-of-service attack, by just swamping some poor guy with FIDONET MAIL emails, with bundles upon bundles of fake mail, without the possibility to trace it. With the direct method (mailer<->mailer) you have ordinary session password, and a possibility to block incoming bundles _before_ they are transferred. The discussion that are taking place is also used to find the best place to put the information in the nodelist, without breaking too much software. When we find a place, it can be used for the information that you would need to put in the nodelist, to denote your kind of distribution system, or some other kind of system (like FTP). It _is_ easier to get the direct mailers up and running, than it is to get fido-by-email up an running (with respect to the nodelist information). Rune --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 1:19 #145 By: Rune Johansen To: All Re: Proposal? St: ------------------------------------------------------------------------------ Proposal for flags to denote various kinds of mail transport via the internet: First, we have to differ between two kinds of transport: - Direct IP connection, using proprietary protocols like: - Telnet (uses ordinary mailer handshake, like FTS-1, FTS-6 and/or EMSI) - VModem ( --""--" ) - IFCico (uses proprietary handshake and transfer method) - BinkP(D?) (uses proprietary handshake and transfer method) - ICQ (paging/transfer system) - Indirect connections, using a application that uses the internet transport mechanisms like: - Email - FTP The first group I would like to call DIRECT connections. The second group INDIRECT connections. For the DIRECT connections, I would propose using (for now) a userflag of "IP". Then, to denote the protocol used for initial connection, these ones can be used: - Telnet: ITEL - VModem: VMO - IFCico: IFC - BinkP: BNK - ICQ: ICQ If the system supports the protocol on a non-standard port, it should be denoted after the protocol flag, with a colon: - Telnet on port 1024: ITEL:1024 - VModem on port 3142: VMO:3142 - IFCico on port 100: IFC:100 - BinkP on port 10234: BNK:10234 So, for a system that support telnet on standard port, Vmodem on port 100, and IFCico on port 102, the user-flags would look like this: ...,U,IP,ITEL,VMO:100,IFC:102 For the INDIRECT connections, I would propose using (for now) a userflag of "IOT". Then, to denote the method used for sending the files, these ones can be used: - Email: EM - FTP: FTP If the receiving system needs a special keyword or something to distinguish the files from other things, it should be denoted after the keyword: - Email need to be sent with FIDONET in the subject: EM:FIDONET So, for a system that support email with no special indication, the userflags would be: ...,U,IOT,EM I am aware of that the placement of the receiving address is not by the time being finally placed, but I would assume that we will use the "phone" field of the nodelist entry to denote this. The phone field is being used for letting your application know where it should "connect", by looking at this field. For DIRECT connections this would be a network address (IP number, DNS address, X25 id, ICQ number, whatever). For INDIRECT connections it would be the receivers address (email address, FTP server IP/DNS, whatever). Example for DIRECT connection: ...,SysOp,bbs.some.net,9600,CM,U,IP,ITEL,VMO ...,SysOp,12.34.56.78,9600,CM,U,IP,ITEL:1024,VMO,IFC,BNK:234 Example for INDIRECT connection: ...,SysOp,rune@some.net,9600,CM,U,IOT,EM ...,SysOp,ftp.some.net,9600,CM,U,IOT,FTP --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 20 Sep 97 23:28 #146 By: maciej grzeszczuk To: Pedro Lima Re: IP-access St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Thu, 18 Sep 97 00:14:28 +0200 Pedro Lima napisal byl w fido.ip_connect: > The flag fields are limited in size, which may place problems if we're > going to allow several network technologies. Also, I understand most of > the existant software already works with the address in the phone field. > Furthermore, I think using the phone field to contain the address is the > natural way. but what with systems using both method - standard modem and tcp connections carrying one aka? fe. at 2:480/70 modem answers on number +48-22-8313212, and ifcico listens at 195.117.131.17:60179. i don't want to use another useless aka just for my ip address carried in phone field, as we could use specified flags for it. modification of software wouldn't be a problem, as we can easily specify subst file for every mailer (with the node number and the correct inet address) if it cannot use flag carrying address... > mg> the only way to reach immediatelly fidonet mailer is emsi. smtp or fts > mg> isn't a way. > Why not? as it's not direct, you know what i mean? you are fetching/processing your mail fe. twice a day. you have no static ip address, no tcp port that mailer is listening... 137 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: zadna dupa, zadna lala nie zastapi terminala (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 20 Sep 97 23:32 #147 By: maciej grzeszczuk To: Rune Johansen Re: Why IP? St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Wed, 17 Sep 97 18:39:02 +0200 Rune Johansen napisal byl w fido.ip_connect: > Even though the NS you looked at has got a 7 day refresh, does not mean that > you can have a delegated part of the DNS on your server, with much less > refresh > time. when i asked z2.fidonet.org dns maintainer about delegating IN NS record for our region as we have number of tcp nodes and this would help us maintaining A and MX records i received no answer. wonder why? 138 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: bede mial dwa bilety na jacksona, za ^Bzupelne fr (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 0:44 #148 By: maciej grzeszczuk To: Steve Woodmore Re: IP-connectivity St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Fri, 19 Sep 97 07:39:10 +0200 Steve Woodmore napisal byl w fido.ip_connect: > mg> ip node is ok, if it's really ip node - i meant if it's accessible > mg> by tcp all the time. smtp nodes and ftp ones shouldn't be treated > mg> as nodes, as there is no way to reach them directly. do you know > mg> what i mean? > Incorrect, My system is online 24hrs/day via PSTN. oh yes, yours is accesible via pstn. i was talking about systems that are not, and only accepting smtp 'connections'. 144 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: jestem los, po prosu los (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 #149 By: Jan Vermeulen To: Pedro Lima Re: Test of an IP-number in the nodelist / Proposal St: Local Sent ------------------------------------------------------------------------------ wd>> ,3000,IP-test,somewhere,Lothar_Behet,000-194-231-142-17,300,VM JV>> Dutch Telecom sends the 'number not obtainable' signal. This means JV>> no cost, but it could impair the operation of my system being off hook JV>> while others try to call it. pl> Can't you cancel the attempt from the start by configuring the 'VM' pl> flag? I would be able to find a way to do that (I think), but could every Jane or John Doe who's never heard of IP-acces? -=<[ JV ]>=- * Origin: The Poor Man's Workstation (2:280/100) ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 0:42 #150 By: Denis McMahon To: Sean Rima Re: IP-connectivity St: ------------------------------------------------------------------------------ * Original message posted in: ENET.SYSOP. * Crossposted in: IP_CONNECT. Sean Rima wrote to Pedro Lima: SR> Try this, In the phone field enter 192*193*25*125. All the Mailers SR> I have tried have have usually replied that there is no phone SR> translation available, except one, which is Argus which picks it up SR> as an IP address. I am not sure if IFmail will do the same but if SR> someone has IFmail then maybe if it works it could be the solution. Here's a control file that I'm using to test makenl: ; ; This is the control file for test ip inserts ; make network 999 outfile net999xy process mon master c:\MAKENL\MASTER outpath c:\MAKENL\FILES mailfiles c:\makenl\net\in badfiles c:\MAKENL\BAD notify self notify receipt notify errors CRASH clean submit 2:99/0 CRASH netaddress 2:999/0 messages c:\makenl\net\out data Host,999,ip_host,here,sysop,44-1234-567890,9600,CM,XA,H16,V34,V32T,VFC,U,X2C ,1,test_1,nowhere,no-one,1-host-sub-domain-org,300,UFLAGS ,2,test_2,nowhere,no-one,013-251-18-49-167,300,UFLAGS Pvt,3,test_3,nowehere,no-one,host.sub.domain.org,300,UFLAGS ,4,test_4,nowhere,no-one,0host-sub-domain-org,300,UFLAGS ,5,test_5,nowhere,no-one,123*234*231*67,300,UFLAGS Here's the error messages makenl generates: Begin processing NET999XY -- 0:39, Monday, September 22, 1997 ,1,test_1,nowhere,no-one,1-host-sub-domain-org,300,UFLAGS MAKENL.CTL -- Invalid phone number -- "1-host-sub-domain-org" ,4,test_4,nowhere,no-one,0host-sub-domain-org,300,UFLAGS MAKENL.CTL -- Invalid phone number -- "0host-sub-domain-org" ,5,test_5,nowhere,no-one,123*234*231*67,300,UFLAGS MAKENL.CTL -- Invalid phone number -- "123*234*231*67" Here's the makenl segment that comes out: ;A Host Nodelist for Friday, September 26, 1997 -- Day number 269 : 00000 ; Host,999,ip_host,here,sysop,44-1234-567890,9600,CM,XA,H16,V34,V32T,VFC,U,X2C ;E ,1,test_1,nowhere,no-one,1-host-sub-domain-org,300,UFLAGS ;E -- Invalid phone number -- "1-host-sub-domain-org" ,2,test_2,nowhere,no-one,013-251-18-49-167,300,UFLAGS Pvt,3,test_3,nowehere,no-one,-Unpublished-,300,UFLAGS ;E ,4,test_4,nowhere,no-one,0host-sub-domain-org,300,UFLAGS ;E -- Invalid phone number -- "0host-sub-domain-org" ;E ,5,test_5,nowhere,no-one,123*234*231*67,300,UFLAGS ;E -- Invalid phone number -- "123*234*231*67" Nice idea, but it failed. Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 0:47 #151 By: Denis McMahon To: Cliff Harrold Re: It's in ... St: ------------------------------------------------------------------------------ * Original message posted in: ENET.SYSOP. * Crossposted in: IP_CONNECT. Cliff Harrold wrote to Ward Dossche: CH> Friday September 19 1997 02:07, Ward Dossche wrote to All: WD> The fake-country-prefix is "000", "VM" is a non-approved-flag to signal WD> VModem. CH> 000 is an international ISDN data call from the UK! So? Translate the 000- prefix into something else in your dial translation! The reason for using an invalid country code is so that you can strip it out or replace it with a gateway telno in the dial translation table. Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 12:04 #152 By: Marco d'Itri To: All Re: sdasd St: ------------------------------------------------------------------------------ dads -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:332/206.10@fidonet) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 12:27 #153 By: Marco d'Itri To: maciej grzeszczuk Re: IP-access St: ------------------------------------------------------------------------------ krap@zwieracz.psych.uw.edu.pl wrote: >if a system is not reachable by emsi (only e-mail transfers of echomail) >it shouldn't be listed as a node... Then we shouldn't list binkd nodes and ftp nodes too. They aren't reachable by EMSI. Please provide a rationale for what you wrote. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 12:27 #154 By: Marco d'Itri To: rafal wiosna Re: IP-access St: ------------------------------------------------------------------------------ rafal.wiosna@f33.n480.z2.fidonet.org wrote: > Such nodelist[s] is[are] assembled by hand and CRC is then > calculated. >Definitely not an option for our [TADA.WAV] beloved ZC. 8^) And so what? We just need to rewrite MakeNL. I know that an italian node is already doing that because there is no MakeNL for linux. If he will not finish I will write MakeNL in perl myself (because it's a shame that one must use an old and non free software to run fidonet). -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 12:27 #155 By: Marco d'Itri To: Ward Dossche Re: Why IP? St: Replied ------------------------------------------------------------------------------ Ward.Dossche@f854.n292.z2.fidonet.org wrote: >You have just started a complete new discussion on another issue. Please >proceed. IP are A Bad Thing(tm) because: - in a couple of years we will have IPv6 nodes, and then we will not be able anymore to directly specify IP addresses because IPv6 addresses are very long and much more volatile than IPv4 ones - I can change the IP of my node and then reload the zone of my domain, but if the nodelist has my IP I will be unreachable for some days - with domains I can support dynamic IP addresses - I can see no reason to use them anyway. We would probably have to rewrite MakeNL (and maybe make it free software), so this time we should do it right from start. Someone wrote about dynamically assigned domains, he could has been right six months ago, but the new BIND supports an extension of the DNS protocol designed to reduce caching problems. AFAIK the dynamic DNS is already used in mobile-IP situations, but ISPs don't have any incentive to provide them to their clients. There is also a non standard system provided by Monolyth (ml.org, dyn.ml.org?). -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 12:27 #156 By: Marco d'Itri To: maciej grzeszczuk Re: Hmm. St: ------------------------------------------------------------------------------ krap@zwieracz.psych.uw.edu.pl wrote: >systems that don't allow direct dialup access via modem should be pvt nodes. >systems that don't allow direct access should be points, as it's impossible >to immediatelly direct access them by emsi. His system allows direct access and EMSI. Via modem. The policy does not mandates that EMSI must be the only one method for exchanging fido mail. This is why we should allow normal PSTN or ISDN nodes to exchange their mail in an email tunnel or via FTP, and we should have nodelist entries for them (but not in the global nodelist). -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 12:27 #157 By: Marco d'Itri To: rafal wiosna Re: Hmm. St: ------------------------------------------------------------------------------ rafal.wiosna@f33.n480.z2.fidonet.org wrote: >IP-systems should remain points? But OTOH why bother dialing to the >internet for >an ARCmail then when you're non-CM but IP? It'd be simpler just to call >PSTN/ISDN node then... 8^) My hub (this is not my main address) does not receive international areas, I had to open two other points email-tunnelled to get them. And not all users can reach nodes at local toll. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 12:27 #158 By: Marco d'Itri To: Steve Woodmore Re: Hmm. St: ------------------------------------------------------------------------------ Steve.Woodmore@p1.f410.n440.z2.fidonet.org wrote: >You are still forgetting that most fidonet nodes that use the internet to >shift mail around are NOT online all of the time. The extended nodelist tries to solve that, and you would not need another node number. (std. discl.: before starting flaming please note that I'm not suggesting that extended entries should go in the global nodelist.) >This then comes back to having special flags in the nodelist. Yes, technically it's a flag :-), you would have in the IP-list something like: ,U,#,EMAIL,ward-bbs@isp.com,UUE,TRNX -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 12:27 #159 By: Marco d'Itri To: maciej grzeszczuk Re: what about... St: ------------------------------------------------------------------------------ krap@zwieracz.psych.uw.edu.pl wrote: >you got my idea, however we name that flag... IP seems to be the best, as >it's >the shortest one. Please explain why we can't put the domain in the phone field. It's easier to modify IP mailers to parse letters that writing the code to recognise that the flag after another flag is a domain. >vmodem, ifcico, binkd and so on has already their own port numbers, so there >is no need to add one after vmodem or ifcico flag. No one wrote that the port must be always specified. And please remember that on unix systems (and on every multiuser system) the telnet daemon already binds to the port 23, and the used port is 60177. >what do you mean TCP? all that vmodem, ifcico, binkd use tcp for transport. TCP also collides with the Txy time flag. >,U,IP,host.zdupy.pl,IFC,PORT,60177,BNK,PORT,31337 This is difficult to parse. In the extended nodelist we can have multiple entries for a node and we can specify the port in the usual internet way, but on normal nodelist we should use (only when needed) "TELN:60177" and so on. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 12:28 #160 By: Marco d'Itri To: Anton Kuznetsov Re: IP addressing St: ------------------------------------------------------------------------------ Anton.Kuznetsov@f50.n465.z2.fidonet.org wrote: >What about BinkD capability [BSO FTN-mailer via TCP/IP] ?.. R50 uses BND. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 12:28 #161 By: Marco d'Itri To: All Re: Extended nodelist format St: ------------------------------------------------------------------------------ This is my draft of the extended nodelist format specifications. Any comment is welcome (even about spelling!). Please note that I'm not suggesting anywhere that the global nodelist will have to contain extended data entries, or that IP nodes will not have their own node number, those are a political decisions that only *C should take. This is just a technical standard, so don't start flaming. The augmented BNF meta language is the one of RFC 822. (AFAIK this format is in use in R50, and I hope it will be in R33.) The extended nodelist format - an augmented BNF definition. nodelist = header eol [ crc eol ] *(line eol) header = comment_A nlist_desc [ nlist_version ] ;this string identifies an xtd nodelist " Nodelist for " date " -- Day number " daynumber [ " : " compat_crc ] nlist_desc = text nlist_version = " v 0.2" date = day ", " month dmonth ", " year daynumber = 1*3( digit ) compat_crc = 5( number ) ; This is a 16 bit XMODEM CRC crc = comment_0 "CRC: " crc_type " " crc_val crc_type = ( "CRC32" / "SHA" / "MD5" / ) crc_val = hexnumber line = ( dataline / comment ) dataline = keyword "," node_num "," node_name "," location "," sysop_name "," compat_phone "," compat_bps "," flags keyword = ( "Zone" / "Region" / "Host" / "Hub" / "Point" / "Pvt" / "Hold" / "Down" / "" ) node_num = ( fido_addr / fido_num ) fido_addr = number ":" number "/" number [ "." number ] fido_num = number ; 16 bit signed integers except the point num. node_name = text location = text sysop_name = text compat_phone = 1*( digit / "-" / "-Unpublished-" ) compat_bps = ( "300" / "1200" / "2400" / "9600" / "19200" / "38400" ) flags = #( flag ) [ ",U," #( flag ) ; extended flags [ 1#( ",#," xtd_data ) ] ] ; extended entries flag = 1*( / number ) xtd_data = ( xt_phone / xt_iskra / xt_inet / xt_email ) xt_phone = "PHONE," phone_num "," bps_rate "," flags xt_iskra = "ISKRA," phone_num "," bps_rate "," flags xt_inet = "INET," inet_addr "," bps_rate "," flags xt_email = "EMAIL," email_addr "," flags phone_num = 1*( digit / "-" / "*" / "#" / "." / "_" / "+" / ) bps_rate = 1*( digit / "K" / "M" ) inet_addr = 1*( / / number / "." / "-" ) [ ":" number ] email_addr = "> comment = 1*( comment_S / comment_U / comment_F / comment_A / comment_E / comment_0 ) comment_S = ";S " text comment_U = ";U " text comment_F = ";F " text comment_A = ";A " text comment_E = ";E " text comment_0 = "; " text day = ( "Monday" / "Tuesday" / "Wednesday" / "Thursday" / "Friday" / "Saturday" / "Sunday" ) month = ( "January" / "February" / "March" / "April" / "May" / "June" / "July" / "August" / "September" / "October" / "November" / "December" ) dmonth = 2( digit ) year = 4( digit ) number = 1*( digit ) ; decimal number hexnumber = 1*( digit / character ) ; hexadecimal number text7 = 1*( ) text = 1*( text7 / digit = <0-9> character = / eol = The extended nodelist format - content of the fields. The first line ("header") is the standard header and checksum line described in FTS-0005. After the network name there is a string that can be parsed (with a regexp like /\sv\s(\d+\.\d+)\sNodelist/) to get the version number of the used extended nodelist standard. This string should not interfere with the parsing made by normal nodelist processors. The version number will be incremented when a new release of the extended nodelist format will not be anymore backward compatible. If the local policy allows that, the checksum can be omitted, because the extended one works better. The second line ("crc") contains a cryptographic hash of the "line" parts of the nodelist, in the order they appears. After the string "CRC:" there are two words: the type of the hash computed and his hexadecimal value. If the type of hash used is cryptographically strong (i.e. SHA or the weaker MD5) the coordinator who publish the nodelist can sign it with PGP. All other lines are standard nodelist data lines as described by FTS-0005, with the exception that there is no limit on the lenght of the user flags. The extended data entries are stored after the ",U," user flags delimiter; each extended data entry is separed from other entries by the ",#," flag. Each extended data entry begin with the type flag that distinguish the kind of communication media: PHONE a normal PSTN or ISDN line ISKRA ??? INET a direct IP connection EMAIL an Internet mailbox The "PHONE" type identifies a normal phone line, like the ones of the classic nodelist, but which share the same node number of the classic data entry. The "ISKRA" type is not known by the writer of this document, but looks like some kind of telephonic service. The "INET" type identifies some kind of connection that takes place over an IP Internet connection. The address field contains an internet domain or a quad dotted IPv4 address between square brackets or an IPv6 address (in a format that will be determined in the future, maybe RFC 1924) and optionally a port number. The flags for the INET connection type are: BND Binkp protocol (24554) RAW fidonet handshake over the raw socket (60179) IFC RAW flag plus support for the ifcico protocol (60179) TELNET fidonet handshake over the telnet protocol (23) TELN fidonet handshake over the telnet protocol (60177) VMP vmodem protocol (3141) FTP FTP transfert between fake outbounds The extended nodelist is not powerful enough to describe all informations needed to set up a FTP link in a simple way. The *C structure should ask sysops to provide a file with a standard name (like /pub/README.fido) that will contains those informations. The "TCP" flag should not be used because it can't be differenced from the Txy time flag. If all services use the default port they can be put in one extended entry, otherwise there will be another entry with the nonstandard port number and the specific flag. The "EMAIL" type advertises an Internet mailbox that can receive arcmail or file attaches encoded in one of those formats: UUE uuencoded files, one per email message MIME MIME base 64 encoded files, one or more per email message TRNX the proprietaty protocol of the TransX program Other flags will be assigned when more powerful variants of UUE and MIME will be standardized. Nodelist processors must support 8 bit characters, but this document does not mandate any specific character set. The *C structure will determine the character set that will be used in national nodelists. For international nodelists UTF-8 should be used. Authors: the original concept of the extended nodelist is by Alex Konshin ; this document and some enanchements to the original format are by Marco d'Itri . This document is copyright (C) 1997 by Marco d'Itri. It can be distributed and used under the GPL. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 12:28 #162 By: Marco d'Itri To: Ward Dossche Re: Alberto's proposal St: ------------------------------------------------------------------------------ Ward.Dossche@f854.n292.z2.fidonet.org wrote: >I have the funny feeling that all of these entries will make MAKENL barf. And then we will use MakeNL 4.0. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 12:28 #163 By: Marco d'Itri To: Jan Ceuleers Re: IP-access St: ------------------------------------------------------------------------------ Jan.Ceuleers@f857.n292.z2.fidonet.org wrote: >My main problem is that due to the caching function I mentioned above, it >doesn't seem to be possible to accomodate nodes whose ISP dynamically >assigns >them an IP address as they log on. I don't think DNS can cope with that, >but I'd >be happy to be proved wrong. You should check latest RFCs about dynamic DNS and UPDATE. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 2:20 #164 By: Steve Woodmore To: maciej grzeszczuk Re: IP-connectivity St: ------------------------------------------------------------------------------ Hi maciej grzeszczuk, hope you are having a nice day 21-Sep-97 00:44:09, maciej grzeszczuk wrote to Steve Woodmore Subject: Re: IP-connectivity mg> oh yes, yours is accessible via PSTN. i was talking about systems mg> that are not, and only accepting smtp 'connections'. Then if they are not FTSC001 Compliant, they shouldn't be in fidonet -=> Steve Woodmore <=- --- Terminate 5.00/Pro * Origin: of Species, a good read (2:254/410.1) SEEN-BY: 2/200 201/329 212/8 235/100 251/25 254/60 61 62 261 410 620 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 440/410 465/50 SEEN-BY: 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 5:34 #165 By: Pedro Lima To: Steve Woodmore Re: Hmm. St: ------------------------------------------------------------------------------ Hi, SW> I really do not think Fido should be making the same mistakes as it SW> always has, stop thinking about the minority, and concentrate on the SW> majority first. I disagree that's the case here. I think we're getting close to a solution for including IP nodes in FidoNet. Naturally, it's easier for everybody if those nodes have permanent Internet connectivity, as IP isn't designed for a store-and-forward network such as FidoNet, but this is not a problem of FidoNet. At best, it's a problem for FidoNet. Now, nodes without fulltime IP connectivity may be listed in the same manner as permanent ones, as long as the Txx flag is used. The problem of the common dynamic IP addressing scheme may be solved through a name server with a low expire time for its caches. Some commercial services are available for it and it seems technically simple to translate a fxx.nxx.zxx.fidonet.org into the 'dynamic DNS' name if needed. IMHO, this allows for the majority to have an IP node in FidoNet, if that majority is indeed willing to be connected to the Internet during the declared operation time. Those who do not wish or see no need to be reachable through Internet as FidoNet nodes, may still profit, as if they have Internet access they can probably connect to an IP FidoNet node. I'm sorry, but as all of this is being currently discussed and considered, I don't understand where we're thinking solely about the minorities. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 5:35 #166 By: Pedro Lima To: Steve Woodmore Re: IP-connectivity St: ------------------------------------------------------------------------------ Hi, SW> You can also send internet email to my IP address 24 hrs/day SW> What you cannot do is to connect to my system Via the internet, unless SW> I am online. Then either the IP address isn't in fact yours, or you're purposely denying the possibility of a connection, since AFAIK, it's not possible to send e-mail to a disconnected IP address. :-) Perhaps you didn't mean an 'IP address', and what you were thinking of was a way of using e-mail as a transport mechanism. That has nothing to do with the possibility of a direct connection between two nodes, and let's not forget that's one of the main purposes of a nodelist in the first place. Listing an e-mail address would be the same as publishing a street address. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 10:21 #167 By: Mats Wallin To: rafal wiosna Re: IP-connectivity St: Replied ------------------------------------------------------------------------------ rw> So what? Is that or the mail the main reason you Fido?... The mail of course, but that does not prevent people from being interested in voting, standing as candidates, etc. (If that was not the case, then there wouldn't be a need for elections at all, would it?). MW>> systems do not allowe points to request files, etc. rw> Unlisted, that is. Correct. And how many pointlists from other regions are you using? What about systems that are not using any pointlists at all? (And there are systems that doesn't allow points to request at all, listed or not). rw> If that's true no IP-only node should be allowed. rw> They can't be reached during zone's ZMH. Sure they can. Anyone running an IP-node will be able to contact them during ZMH (if that's not possible, they should of course not have a nodenumber). Since ISDN-only nodes are allowed, and those cannot be contacted by nodes with an analogue modem, it's natural that IP-only nodes also should be allowed. And this does actually go back much further than ISDN-only nodes. 300-bps nodes are allowed (even though I doubt that there exists one in FidoNet today). However, a 300-bps node in zone 1 cannot call a 300-bps node in zone 2 anyway, since Bell 103 and V.21 is not compatible with each other. So we already have, and have had for a very long time, the case with devices that are not compatible. rw> So, I ask again: what's the point of becoming such a rw> leaf node then? Only to be able to vote? The reason will probably differen from SysOp to SysOp, but in my case, I would think that the most important thing is that it is much easier to communicate with others. It is much easier to write a message to a person that exists in the nodelist. Just type the name of the person and most programs should suggest the correct address automatically. If you don't have the person in the nodelist, you will have to remember and type the address manually. Maybe not a big problem, but I do find that a big advantage. Mats mw@defsol.se --- * Origin: Definite Solutions, Stockholm, Sweden (2:201/329.11) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/27 212/8 235/100 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 10:37 #168 By: Denis McMahon To: Jerry Schwartz Re: IP-connectivity St: ------------------------------------------------------------------------------ * Original message posted in: NETMAIL. * Crossposted in: IP_CONNECT. * Reply to a message in REGCON. Jerry Schwartz wrote to Jan Ceuleers: JS> [Sep 04, 97 - 07:11] Jan Ceuleers of 2:292/857 wrote to Jerry JS> Schwartz: JS> My concern is that the dialing translations in the nodelist JS> compiler might produce things that look like an IP address to a JS> mailer. No, you misread / misunderstood the proposal. The proposal was that use of a fake country code in the nodelist would allow dialling translation to substitute the telno of a suitable gateway for those nodes without ip capabilities themselves, and to provide ip addresses to those nodes that can use them. JS> You're suggesting using a country code of 0 to flag an IP JS> address, so that an IP address would look like a phone number of JS> the form 0-223-456-789-012 or some such, correct? (Again, I object JS> to using any format that doesn't conform to local phone numbers so JS> that NCs don't have to eliminate syntax checking on the nodelist JS> segment.) Fine, where's your proposal for how we accomodate ip and named addressing in the nodelist then? Or do you still believe that 300 baud FTS1 is the defining factor of Fidonet? Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 20:28 #169 By: Denis McMahon To: All Re: IP Protocols St: ------------------------------------------------------------------------------ Hello All! Looking at the next stage (ie after we agree how an IP address should appear in the nodelist), what protocols might an ip connected system support? I can think of three at the moment. 1) Using normal fido mailers communicating via vmodem fossil drivers. 2) Using ftp protocols to transfer pkt, arcmail and files. 3) Using uuencoded etc email to transfer arcmail and files. I believe that UUCP might also be used like ftp? What about using smtp / pop3 / nntp to transfer netmail / echoes? I would say that at that level, the systems are not using "fido protocols" but are using pure internet protocols. I also think that telnet connections are more likely to be virtual bbs, rather than a means of fido communication? Comments? Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 23 Sep 97 0:37 #170 By: Denis McMahon To: All Re: Another failed idea St: ------------------------------------------------------------------------------ Hello All! OK, I tried this line in makenl earlier: ,10,test_10,nowhere,no-one,44-1234-567890,9600,CM,XA,H16,V34, V32T,VFC,V110L,V110H,V120L,V120H,X75,X2S,IP123.221.65.9, DNSalpha.beta.gamma.delta,VM65534,FTP65535,GUUCP,NEC,NC This rep[resents a fully capable analogue/ISDN modem eg I-Modem with USR V.Everything analogue capability, X2 server and also ip and dns entries with Vmodem on port 65534 and ftp on 65535 etc ... The line is too long for makenl to process! Apart from the combination of ip and dns data being redundant, and port numbers not being required if they're at the default values, is there any other way to shorten the line? Is it possible that V110 and V120 can be defined as encapsulating both the L and H settings, so that eg instead of ",V110L,V110H," use ",V110," Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 13:57 #171 By: Marco d'Itri To: Denis McMahon Re: IP-connectivity St: ------------------------------------------------------------------------------ Denis.McMahon@f20.n251.z2.fidonet.org wrote: > SR> as an IP address. I am not sure if IFmail will do the same but if > SR> someone has IFmail then maybe if it works it could be the solution. Ifmail needs a flag on the command line. I asked Pablo to add the support for the extended nodelist format and for the IP in the phone field but I haven't received mail from him since some weeks. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 SEEN-BY: 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 14:04 #172 By: Marco d'Itri To: maciej grzeszczuk Re: IP-access St: ------------------------------------------------------------------------------ krap@zwieracz.psych.uw.edu.pl wrote: >but what with systems using both method - standard modem and tcp connections >carrying one aka? fe. at 2:480/70 modem answers on number +48-22-8313212, They will need another aka. Or they will have to use the extended nodelist format. ;-) Multi line nodes have the same problem. >as it's not direct, you know what i mean? you are fetching/processing your >mail fe. twice a day. you have no static ip address, no tcp port that >mailer is listening... If you use FTP [1] (as a server) you have the same capabilities of a normal node, the only difference is that you don't use FTS-1 or EMSI. [1] Even if I can't understand why someone whould like to use ftp when there is binkd that it's much simpler... -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 SEEN-BY: 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 14:26 #173 By: Marco d'Itri To: Rune Johansen Re: Hmm. St: ------------------------------------------------------------------------------ Rune.Johansen@f20.n210.z2.fidonet.org wrote: >It needs a email address to send the attachments to. If you have only one >email >address that you can receive at, you have to differ those mails from your >regular email, supposedly by using some kind of special subject. Both of >these Any decent OS can filter incoming mail any way the sysop likes. All tunnel programs includes a string somewhere in the headers or in the body. To catch transx nodes you would just have to use: :0 * Received:.*TRANSX | /path/to/transx/unix/interface [1] or whatever is the equivalent for your OS. This is a non-problem. >should be denoted in the nodelist, to make people able to send "crashmail" >via >the email-attachment method. Then, a kind of flag that denotes the method of We can expect a node will have to recognize no more than two or three programs until we will have a standard for email tunneling (I will post my draft to NET_DEV next week). >automatically. It can also be used for a denial-of-service attack, by just >swamping some poor guy with FIDONET MAIL emails, with bundles upon bundles >of >fake mail, without the possibility to trace it. Someone could also put a fake PSTN number in the nodelist, but I have never read about anyone concerned about that. >With the direct method (mailer<->mailer) you have ordinary session password, >and a possibility to block incoming bundles _before_ they are transferred. You can't block email if your provider don't allow that, but my standard supports passwords and encryption. >The discussion that are taking place is also used to find the best place >to put >the information in the nodelist, without breaking too much software. When we Looks like the extended nodelist proposal is a good one. The only software that would have problems is the one that truncates flags, and I suppose no one interested in IP fido will use something that broken[2]. Anyway we will probably never have smtp only nodes, so we could use as well some new nodelist format that will be distributed only to nodes which need it. [1] Yes, I'm writing one. If someone needs it feel free to ask at md@linux.it. [2] Ok, the standard does not allows long flags, but I don't think it's a good reason to not try to be forward compatible. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 SEEN-BY: 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 14:39 #174 By: Marco d'Itri To: Rune Johansen Re: Proposal? St: ------------------------------------------------------------------------------ Rune.Johansen@f20.n210.z2.fidonet.org wrote: > - Telnet (uses ordinary mailer handshake, like FTS-1, FTS-6 and/or EMSI) It uses the ordinary handshake over the telnet protocol. This is an important thing, because ifcico uses the ordinary handshake too, but they are not compatible. > - VModem ( --""--" ) I think VMP (VModem Protocol) is a whole different protocol. Does anyone knows something about it? > - IFCico (uses proprietary handshake and transfer method) Ifcico uses the same FTS-1 or EMSI handshake any PSTN node uses, but it does that over the raw socket. It also has an optional comunication protocol (like zmodem or hydra) designed for TCP connections, but it uses it only if the remote ends supports it. >The first group I would like to call DIRECT connections. The second group >INDIRECT connections. An FTP server is as direct as the other methods. I can't undestand why you don't put it with them. > - Telnet: ITEL The extended nodelist already uses TELN, please try to not invent a new flag if possible. > - VModem: VMO Many standard nodelists (eg: R50) uses VMP. > - IFCico: IFC > - BinkP: BNK Many standard nodelists (eg: R50) uses BNP (binkd is the name of the program, binkp is the name of the protocol). > - ICQ: ICQ What is ICQ? > - Email: EM There are a lot of ways to send email. My extended nodelist proposal states: -snip- The "EMAIL" type advertises an Internet mailbox that can receive arcmail or file attaches encoded in one of those formats: UUE uuencoded files, one per email message MIME MIME base 64 encoded files, one or more per email message TRNX the proprietaty protocol of the TransX program Other flags will be assigned when more powerful variants of UUE and MIME will be standardized. -snip- It also states, about FTP: -snip- The extended nodelist is not powerful enough to describe all informations needed to set up a FTP link in a simple way. The *C structure should ask sysops to provide a file with a standard name (like /pub/README.fido) that will contains those informations. -snip- >If the receiving system needs a special keyword or something to >distinguish the >files from other things, it should be denoted after the keyword: > - Email need to be sent with FIDONET in the subject: EM:FIDONET There is no need to specify those things, the protocol should already mandate some standard way to identify messages. >being finally placed, but I would assume that we will use the "phone" >field of >the nodelist entry to denote this. The phone field is being used for letting I agree. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 SEEN-BY: 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 14:40 #175 By: Marco d'Itri To: Ward Dossche Re: MAKENL ! ? St: ------------------------------------------------------------------------------ Ward.Dossche@f854.n292.z2.fidonet.org wrote: >Anyone ambitious to rewrite MAKENL? If no one will write a GPL'ed multiplatform version I will write a perl one. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 SEEN-BY: 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 21:03 #176 By: Ward Dossche To: Rune Johansen Re: Test-status St: Replied ------------------------------------------------------------------------------ Rune, > After all, > even though people in europe or america can live with (and don't get any > cost by) dialling a foreignprefix+000 as a default conversion rule, we > don't know how it would affect nodes in, let's say Uganda. In theory this is correct, I'd like to utter 2 buts though: But-1: What is in the nodelist now is merely a test, nothing more, nothing less. To avoid trouble outside zone-2 that/those listing(s) do(es)n't appear in another zone; But-2: I have a hunch triple-"0"-prefix will hurt no-one. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 21:08 #177 By: Ward Dossche To: maciej grzeszczuk Re: Why IP? St: ------------------------------------------------------------------------------ > when i asked z2.fidonet.org dns maintainer about delegating IN NS record > for our region as we have number of tcp nodes and this would help us > maintaining A and MX records i received no answer. wonder why? You forgot to say "please"? Or maybe the postmaster of z2.fidonet.org is concerned about the possible deterioration of the records. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 21:23 #178 By: Ward Dossche To: Marco d'Itri Re: what about... St: ------------------------------------------------------------------------------ Marco, > Please explain why we can't put the domain in the phone field. Current software (makenl) doesn't process it. And some hardcore fidonetters will yelp about FTS-0005. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 21:26 #179 By: Ward Dossche To: Marco d'Itri Re: Alberto's proposal St: Replied ------------------------------------------------------------------------------ Marco, >>I have the funny feeling that all of these entries will make MAKENL > barf. > And then we will use MakeNL 4.0. Please proceed and just do it, don't talk about it. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 21:30 (Rcvd: Wed 24 Sep 97 21:10) #180 By: Ward Dossche To: Jan Vermeulen Re: IP-connectivity St: Rcvd Replied ------------------------------------------------------------------------------ Jan, > * Crossposted ENET.SYSOP Thank you. I'm not participating in any IP-discussion outside this conference. > The bottom line is that on average this node will have been > unobtainable for incoming calls during a larger part of the day due to no > fault of its own. The bottom-line is that the sysop of this node will have to properly configure his system or remain unobtainable for a major portion of the day until eternity. The bottom-line also is, Jan, that what you describe is most unlikely to happen but that to exclude the unlikely event of its occurence you advocate drastic other approaches which are far more complicated to implement than the case here of virtually adding another country. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 5:56 #181 By: Pedro Lima To: maciej grzeszczuk Re: IP-access St: ------------------------------------------------------------------------------ Hi, mg> i don't want to use another useless aka just for my ip address mg> carried in phone field, as we could use specified flags for it. That's where we disagree. :-) The nodelist works mainly as a translation table between a FidoNet address and a phone address. What we're aiming at is to translate a FidoNet address into an IP address, or a X.25 address or... For me, it's only natural that the phone field is used to contain that "other technology" address. Another aspect is that the length of the flags part of a nodelist entry is limited. If a node offers several connection possibilities via IP (ifcico, VModem, BinkD, etc.), if we added an IP address or a FQDN that limit would be easily exceeded. And we're thinking only about IP. What would happen if we had also X.25 or whatever? mg> as it's not direct, you know what i mean? You mentioned fts. I'm assuming that would namely include FTS-1 and FTS-6, and these are protocols for a direct connection. But apart from that, I fully agree with you. :-) Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 5:57 #182 By: Pedro Lima To: Rune Johansen Re: Proposal? St: ------------------------------------------------------------------------------ Hi, RJ> Proposal for flags to denote various kinds of mail transport via the RJ> internet: I like it, as far as direct connections are concerned, although I'd personally prefer 'VM' or 'VMD', instead of 'VMO'. I disagree of indirect methods being listed in the master nodelist, but I have nothing against that information being listed somewhere else. Said otherwise, if you can't offer direct connectivity, you can't be a node. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 5:58 #183 By: Pedro Lima To: Marco d'Itri Re: Why IP? St: ------------------------------------------------------------------------------ Hi, Md> - in a couple of years we will have IPv6 nodes, and then we will not Md> be able anymore to directly specify IP addresses because IPv6 addresses Md> are very long and much more volatile than IPv4 ones I apologize for my ignorance about IPv6, but from the samples I've seen, it looks perfectly possible to include IPv6 addresses in the nodelist, even if they can be quite long. If I'm understanding correctly, the volatility you're talking about is because of the enormous pool of possible addresses, as people could then as easily have an IP adresss as leave it, not because someone configured an IP address for his system and sometime during the night it evaporated? :-) Md> - I can change the IP of my node and then reload the zone of my Md> domain, but if the nodelist has my IP I will be unreachable for some Md> days Md> - with domains I can support dynamic IP addresses Then you'd list your FQDN, not your IP address. What I fail to see is why stop the possibility of listing an IP address. Md> - I can see no reason to use them anyway. Up to some months ago, I could easily obtain an IP address. However, getting a FQDN would imply some kind of payment or at least, a degree of involvement from the part of the ISP people I didn't want to impose on them. So, if we can have both possibilities, why not? Md> We would probably have to Md> rewrite MakeNL (and maybe make it free software), so this time we Md> should do it right from start. That's an *excellent* idea! Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 21 Sep 97 11:12 #184 By: Ger Vloothuis To: Ward Dossche Re: Test of an IP-number in the nodelist / Proposal St: ------------------------------------------------------------------------------ Hi Ward, Ward Dossche wrote in a message to All: WD> Does anyone have a technical problem with it that as a test this WD> line appears in the European version of the world-nodelist next WD> Friday? WD> WD> ,3000,IP-test,somewhere,Lothar_Behet,000-194-231-142-17,300,VM WD> I think a prefix of "000-" will technically bother no-one, correct? WD> \x/ard Triple 0 is the emergency phone number in Australia. Something like 911 in the USA and 112 in GMS networks. Don't know if this would be a problem but the emergency line operators would sure be hindered by a lot of unintended modem calls. So that should be avoided. Regards Ger --- WtrGate v0.93.p1 Unreg * Origin: Teletechnique Melbourne (3:633/284) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 270/101 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/0 200 321 700 2433/0 225 1200 1201 1204 1800 SEEN-BY: 2448/10 6000 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 23 Sep 97 6:54 #185 By: Jan Ceuleers To: Ward Dossche Re: MAKENL ! ? St: ------------------------------------------------------------------------------ I quote Ward Dossche: WD> I've already had some remarks re: the hyphens in the WD> IP-number of the test-node (only in the world-nodelist of WD> zone-2 !!). People request me to put dots there, this is WD> what makenl says : [invalid phone number] Meanwhile, is it necessary to include leading zeroes, or is existing software able to use the dashes as separators? Jan --- timEd-A10a * Origin: Experimenter Board, Antwerp, Belgium (2:292/857) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 23 Sep 97 8:49 #186 By: Mats Wallin To: Marco d'Itri Re: Why IP? St: ------------------------------------------------------------------------------ Md> - in a couple of years we will have IPv6 nodes, and then we Md> will not be able anymore to directly specify IP addresses Md> because IPv6 addresses are very long and much more volatile Md> than IPv4 ones Why would IPv6 addresses be more volatile? I had thought that due to the increased number of possible IP-addresses, that it would be much easier to get a static IP-address when IPv6 is introduced. So shouldn't IPv6 make the IP-addresses less volatile? Md> - I can change the IP of my node and then reload the zone of Md> my domain, but if the nodelist has my IP I will be Md> unreachable for some days - with domains I can support Md> dynamic IP addresses I don't have anything against domains (except the fact that MAKENL does not accept them in the phone number field), but I really don't understand your problem here. If you move your system to another phonenumber, you will have exactly the same problem as you would have if you changed the IP-address. This is not a new problem, all you have to do is to plan the move a little bit, just as you today have to do when switching phone numbers. Mats mw@defsol.se --- * Origin: Stockholm, Sweden (2:201/329.11) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/27 212/8 235/100 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 23 Sep 97 17:04 #187 By: Rune Johansen To: maciej grzeszczuk Re: Why IP? St: ------------------------------------------------------------------------------ > when i asked z2.fidonet.org dns maintainer about delegating IN NS record > for our region as we have number of tcp nodes and this would help us > maintaining A and MX records i received no answer. wonder why? I do not know what reasons the z2 DNS keeper does not answer you. I only point out the possibility of the delegation method. --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 222/0 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 23 Sep 97 17:07 #188 By: Rune Johansen To: Steve Woodmore Re: IP-connectivity St: ------------------------------------------------------------------------------ >> oh yes, yours is accessible via PSTN. i was talking about systems >> that are not, and only accepting smtp 'connections'. > Then if they are not FTSC001 Compliant, they shouldn't be in fidonet Would you then imply that Fido2Int-capable-only nodes should not be in Fidonet? --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 222/0 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 22 Sep 97 22:27 #189 By: maciej grzeszczuk To: Rune Johansen Re: Flags in nodelist St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Sat, 20 Sep 97 00:09:58 +0200 Rune Johansen napisal byl w fido.ip_connect: > My suggestion is therefor: Go with IP flag, and let ignorant people that > don't > pay attention to extended connectivity sail their own sea, up shit creek. absolutelly right. 176 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: radykal = wesola kupa (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 24 Sep 97 #190 By: Jan Vermeulen To: Marco d'Itri Re: Why IP? St: Local Sent ------------------------------------------------------------------------------ Quoting Marco d'Itri on Sun 21 Sep 97 12:27 to Ward Dossche: md> IP are A Bad Thing(tm) because: &c. Are you saying that we're up a blind alley or did I misread you? If this is what you were saying, what is the alternative? -=<[ JV ]>=- * Origin: The Poor Man's Workstation (2:280/100) ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 24 Sep 97 #191 By: Jan Vermeulen To: Mats Wallin Re: IP-connectivity St: Local Sent ------------------------------------------------------------------------------ Quoting Mats Wallin on Mon 22 Sep 97 10:21 to rafal wiosna: mw> Since ISDN-only nodes are allowed, and those cannot be contacted by mw> nodes with an analogue modem, it's natural that IP-only nodes also mw> should be allowed. Wrong argument for a good cause, Mats: first of all there are far less ISDN-only *sysops* and secondly there is no need to repeat the same error over again, is there? -=<[ JV ]>=- * Origin: The Poor Man's Workstation (2:280/100) ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 24 Sep 97 #192 By: Jan Vermeulen To: Ward Dossche Re: Test-status St: Local Sent ------------------------------------------------------------------------------ Quoting Ward Dossche on Mon 22 Sep 97 21:03 to Rune Johansen: wd> But-2: I have a hunch triple-"0"-prefix will hurt no-one. Do you really think that hunches are a good foundation for working out a new standard, Ward? My hunch is that it will hit us the Murphy way, i.e. at the least convenient moment. -=<[ JV ]>=- * Origin: The Poor Man's Workstation (2:280/100) ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 24 Sep 97 #193 By: Jan Vermeulen To: Ward Dossche Re: Alberto's proposal St: Local Sent ------------------------------------------------------------------------------ Quoting Ward Dossche on Mon 22 Sep 97 21:26 to Marco d'Itri: >> And then we will use MakeNL 4.0. wd> Please proceed and just do it, don't talk about it. Meaning that the discussion on where to put the internet address is terminated? I've still not seen a founded rebuttal of Rune's and my own objections.... -=<[ JV ]>=- * Origin: The Poor Man's Workstation (2:280/100) ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 24 Sep 97 #194 By: Jan Vermeulen To: Ward Dossche Re: IP-connectivity St: Local Sent ------------------------------------------------------------------------------ Quoting Ward Dossche on Mon 22 Sep 97 21:30 to Jan Vermeulen: wd> ... you advocate drastic other approaches which are far more wd> complicated to implement than the case here of virtually adding wd> another country. I advocate the investigation of other approaches which, in the long run, may avoid a disaster for FidoNet. You have not given any technically sound grounds for not investigating these other approaches; expedience is not an argument. We are not an enterprise that needs to show an ever improving quarterly bottom line. Dixi. -=<[ JV ]>=- * Origin: The Poor Man's Workstation (2:280/100) ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 23 Sep 97 22:46 #195 By: Ward Dossche To: Jan Ceuleers Re: MAKENL ! ? St: ------------------------------------------------------------------------------ Jan, > Meanwhile, is it necessary to include leading zeroes, or is existing > software able to use the dashes as separators? I have netmail confirmations of people that have been able to have a fido-session with 2:2/3000 by using and translating the entry in the nodelist. I'm not sure though how it was done as the dashes needed translating. There are so far 3 guys that I know of which will shortly come forward with versions of new software to better generate nodediffs/lists including internet-style- and IP-addresses. \x/ardx --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 23 Sep 97 8:04 #196 By: Eckhard Mueller To: Ward Dossche Re: It's in ... St: ------------------------------------------------------------------------------ * Replying to a msg in REGCON.EUR (REGCON.EUR) Hello Ward! >> ,3000,IP-test,somewhere,Lothar_Behet,000-194-231-142-17,300,VM WD> Please report on how it works, or maybe doesn't. Although I have an internet access for a few months now, I'm not an expert in it, use it only seldom and so i continued to think that the new IP-nodes would be unreachable and of no use for me. So yesterday I wanted to prevent my mailers from accidentally dialing such an IP-number... But while changing the mailer's setup, I suddently thought: "VM... VModem... sounds easy. Should work with my OS/2, and internet access is available on the mailbox computer - shouldn't I try to set up a new mailer line using VModem?" Well, I tried, and a lot of hours later I really succeded :-) OK, I'm far away from being able to be callable via VModem, but I can reach other VM-lines in the nodelist: I crashed mail to 2:2/3000. That's much more than I expected, and it might make a lot of things possible that I cannot really imagine now. This success alone, and the (technical) fun I had seeing pure fido over internet, should be more than enough to justify your little IP-test in the nodelist: thank you very much. Some technical details: I did not find a way to convert 000-194-231-142-17 into 194.231.142.17 with my T-Mail. I can strip 000- from it, no problem, and I can remove the dashes, but I could not convert "-" to "." by means of the mailer, and VModem refused 194-231-142-17. Since I'm not using the original nodelist but a converted one, I changed my conversion utilities a bit, so the dashes are converted now before T-Mail's nodelist compiler reads the nodelist. I had some trouble with SIO (VModem is part of SIO), and I don't know whether I really solved it, although everything seems to work now. SIO was running here earlier (but deinstalled later), now I have a new motherboard and SIO didn't work correctly with my serial ports. No idea what is the problem, seems SIO recognizes my modem as the mouse(???), I was forced to keep IBMs original COM port device driver, but I have no idea how to tell SIO to keep its fingers away from COM1, COM2, and COM3. andy elkin wants to support the VModem protocal internally in the next version of T-Mail (2.603), I hope that this will result in more smooth operations here. I expected some trouble since my internet access does not involve a serial port but ISDN (via Teles-CAPI) instead, but this was no problem. Way of the transmitted bytes: X75 Provider---------Teles S0.16---CAPI---ISDNPM---VModem/SIO---TMail My mailer missed a CM flag for 2:2/3000, so I told him to ignore that... Since X75 is still a _user_ flag, VM should be one, too. Could you change the line to ...,300,CM,U,VM ? Transfer speed was around 2400 cps. Calling Lothar directly via X75 would be 3 times faster, but no local call. Bye Eckhard --- GoldED/2 2.50.Beta6+ --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 23 Sep 97 11:31 #197 By: Ask Bjoern Hansen To: Rune Johansen Re: Flags in nodelist St: ------------------------------------------------------------------------------ Hello Rune! Saturday September 20 1997 01:09, Rune Johansen wrote to All: RJ> My suggestion is therefor: Go with IP flag, and let ignorant people RJ> that don't pay attention to extended connectivity sail their own sea, RJ> up shit creek. My suggestion: Don't include IP numbers or hostnames in the nodelist. Just the flags. You would be reached at f27.n210.z2.fidonet.org; no need for any other addresses. Kind Regards, Ask --- GoldED/2 3.00.Alpha4+ * Origin: PC=Programmable Chaos! Call the Source! <+45-36305533> (2:235/224) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 270/101 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/0 200 321 700 2433/0 225 1200 1201 1204 1800 SEEN-BY: 2448/10 6000 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 8:47 #198 By: Rune Johansen To: Marco d'Itri Re: Hmm. St: ------------------------------------------------------------------------------ > Any decent OS can filter incoming mail any way the sysop likes. All tunnel > programs includes a string somewhere in the headers or in the body. > To catch transx nodes you would just have to use: > :0 > * Received:.*TRANSX > | /path/to/transx/unix/interface [1] > or whatever is the equivalent for your OS. This is a non-problem. You are assuming that the email provider is using "a decent OS" :-)) That is not always the situation, and you have to take into consideration that the sysop is not always in charge when it comes to configuring the mail server, or any utilities that you theoretically can use on a mail server. We have one large ISP here in Norway (over 100.000 customers) that don't allow you any OS access at all. You can get your mail via POP3, and use special software to put your web pages on disk, but you can't fiddle with .forward files, procmail settings or anything. So, the filtering would have to be sone at your own system. >> automatically. It can also be used for a denial-of-service attack, by just >>swamping some poor guy with FIDONET MAIL emails, with bundles upon bundles >>of >> fake mail, without the possibility to trace it. > Someone could also put a fake PSTN number in the nodelist, but I have never > read about anyone concerned about that. Of course they can. But, email faking is a more common thing in the internet world than a PSTN phone number faking is in the fidonet world. --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 222/0 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 8:58 #199 By: Rune Johansen To: Marco d'Itri Re: Proposal? St: ------------------------------------------------------------------------------ >> - Telnet (uses ordinary mailer handshake, like FTS-1, FTS-6 and/or EMSI) > It uses the ordinary handshake over the telnet protocol. > This is an important thing, because ifcico uses the ordinary handshake too, > but they are not compatible. >> - VModem ( --""--" ) > I think VMP (VModem Protocol) is a whole different protocol. Does anyone > knows something about it? VModem and Telnet are two protcols that can be treated the same way, with regard to a mailer, as they are giving you a "raw" connection to antoher mailer, just like a modem does. >> - IFCico (uses proprietary handshake and transfer method) > Ifcico uses the same FTS-1 or EMSI handshake any PSTN node uses, but it does > that over the raw socket. It also has an optional comunication protocol > (like zmodem or hydra) designed for TCP connections, but it uses it only if > the remote ends supports it. OK, then Iifcico should be treated as a protocol on the same level as Telnet and VModem. >> The first group I would like to call DIRECT connections. The second group >> INDIRECT connections. > An FTP server is as direct as the other methods. I can't undestand why > you don't put it with them. Because you don't necessarily have to have a mailer on the same machine as a FTP server. You cannot use a fido handshake with a FTP server. You put and get files to/from a FTP sever, basically the same way as you put and get files from a mail account. >> - Telnet: ITEL > The extended nodelist already uses TELN, please try to not invent a new > flag if possible. I am trying to avoid confusion with the Txx flag. >> - VModem: VMO > Many standard nodelists (eg: R50) uses VMP. OK, then VMP can be used. >> - IFCico: IFC >> - BinkP: BNK > Many standard nodelists (eg: R50) uses BNP (binkd is the name of the > program, binkp is the name of the protocol). OK, then BNP can be used. >> - ICQ: ICQ > What is ICQ? :)) A "pagin" program that many people use. "I seek U". I included it to show that other transport mechanisms than those being discussed here are to be considered, and the format so flexible that other protocols and transport mechanisms could be included, without discussing for more years. For instance X25 numbers, and other "exotic" things. >> - Email: EM > There are a lot of ways to send email. My extended nodelist proposal states: True, there are a lots of ways to do things, also within a special group of applications. It also shows that one cannot tink of all possibilities, and should make a proposal that is so flexible that new things can be added seamlessly. Like the ICQ system. You don't know anything about it, but some people would maybe prefer it over other, more "common" transports. >> files from other things, it should be denoted after the keyword: >> - Email need to be sent with FIDONET in the subject: EM:FIDONET > There is no need to specify those things, the protocol should already > mandate some standard way to identify messages. Yes, there is a need to specify it, if you have to go and get the things done yourself. --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 222/0 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 23 Sep 97 22:22 #200 By: maciej grzeszczuk To: Steve Woodmore Re: IP-connectivity St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Mon, 22 Sep 97 01:20:22 +0200 Steve Woodmore napisal byl w fido.ip_connect: > mg> oh yes, yours is accessible via PSTN. i was talking about systems > mg> that are not, and only accepting smtp 'connections'. > Then if they are not FTSC001 Compliant, they shouldn't be in fidonet absolutelly right. now tell the same to ward. 189 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: ...bo jest ..ujowa... (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 23 Sep 97 22:26 #201 By: maciej grzeszczuk To: Mats Wallin Re: IP-connectivity St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Mon, 22 Sep 97 09:21:08 +0200 Mats Wallin napisal byl w fido.ip_connect: > rw> If that's true no IP-only node should be allowed. > rw> They can't be reached during zone's ZMH. > Sure they can. Anyone running an IP-node will be able to contact them during > ZMH (if that's not possible, they should of course not have a nodenumber). sure you're right. ip node should be allowed. smtp "nodes" should NOT, as there is NO way to reach them by ZMH, or anytime... 190 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: waljoooooom! (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 24 Sep 97 14:22 #202 By: Mario Mure' To: Ward Dossche Re: MAKENL ! ? St: ------------------------------------------------------------------------------ In a message dated < 21 Sep 97 21:52:30 >, addressed to All about < MAKENL ! ? >, Ward Dossche wrote: WD> Anyone ambitious to rewrite MAKENL? Andrea Baldoni (2:332/521) *IS* rewriting (from scratch) MAKENL for Linux and portable under other platforms too. Code will be free and released under GPL license. AFAIK, Andrea is looking for people willing to port his code under other platforms, so feel free to contact him directly. He is also e-mail reachable as abaldoni@xcal.net Ciao ! /mario/ mure@sistemia.it --- DMreply v2.0 * Origin: Djemaa El Fna (2:335/533@fidonet) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 24 Sep 97 14:25 #203 By: Marco d'Itri To: Ward Dossche Re: It's in ... St: ------------------------------------------------------------------------------ Ward.Dossche@f854.n292.z2.fidonet.org wrote: >If during the test-phase other nodes want to be listed as IP-node in the >top-section of the nodelist then these people will have to netmail me. Please add (my sysop will forward that message to you): ,,Eagle's_Nest,Montesilvano_Spiaggia_PE_I,Stefano_Brandimarte,000-x.x.x.x,\ 300,CM,U,IP,TELN,IFC,BND (The IP is the one of tella.alicom.org, I don't remember it now.) >What I need for now is a static IP-number and a proposal for a flag >signaling >the capability if it isn't VModem. Please note that OS/2 vmodem systems supports two protocols: I think we should adopt some of the flags used in R50 and R40: ; TCP - mandatory for any ip-node We can't use it because collides with the Txy time flag. We should use IP. ; VMP - Vmodem proprietary protocol (default port 3141) ; TEL - ftn handshake over telnet protocol (default port 23) We should use TELN. ; RAW - (my proposal, used by the xtd nl) ftn handshake over the raw socket ; (default port 60179) ; IFC - RAW + the ifcico TCP comm. protocol (default port 60179) ; BND - binkp protocol (default port 24554) Maybe this one should be BNP? ; U,xxxxx - marks non-standard port number, relates to _first_ ; capability flag in the line (only one non-standard port ; can be lested). IMHO we should not specify non-standard ports this way because it supports just one service and it's difficult to parse by programs. So far the best proposal (except the extended nodelist one) is to add something like TEL:60177. ; U,"foo.bar" - marks corresponding domain address (temporarily) I don't think it's useful, the domain should go in the phone field, if mailers can't use it, it's pointless to add it. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 SEEN-BY: 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 24 Sep 97 14:30 #204 By: Marco d'Itri To: Rune Johansen Re: Observations and questions St: ------------------------------------------------------------------------------ Rune.Johansen@f20.n210.z2.fidonet.org wrote: >Some people report that MakeNL spits on using non-standard phone numbers, >like Maybe has come the time to write a new MakeNL? >understandable format. Since the nodelist compilers are able to fiddle >with the >actual phone number you are to call, it can also be talked into making IP >or DNS >addresses undialable, by configuration, if you have a PSTN only mailer. I agree, this is the way to proceed, we should not add fake country codes. >don't think that's any fair at all. There ARE a LOT of Fidonet Technology >based >networks out there, and we should not KILL them by saying "All domain name >requests should be done against a built-up address on the fidonet.org >domain". This is true, if mailers would have to resolve the FQDN in fidonet.org, then why put an address (IP or FQDN) in the nodelist? >That's why we HAVE to allow for IP addresses AND fqdn symbolic addresses >in the >nodelist. There is no need to have both, the FQDN is enough. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 SEEN-BY: 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 24 Sep 97 14:34 #205 By: Marco d'Itri To: Rune Johansen Re: Hmm. St: ------------------------------------------------------------------------------ Rune.Johansen@f20.n210.z2.fidonet.org wrote: >> You are still forgetting that most fidonet nodes that use the internet to >> shift mail around are NOT online all of the time. IMHO those nodes should use their PSTN node number. >nodes. If we just can find a way that could be politically accepted, and >that >doesn't break too much software, then we could look into a way of telling >nodes >how store-and-forward-reliant nodes can be directly addressed in the >nodelist. The extended nodelist format supports email nodes without breaking any software, but I don't know if extended entries should go in the global nodelist or only in region managed or even unofficial IP lists. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 SEEN-BY: 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 24 Sep 97 14:40 #206 By: Marco d'Itri To: maciej grzeszczuk Re: IP-connectivity St: ------------------------------------------------------------------------------ krap@zwieracz.psych.uw.edu.pl wrote: >ip node is ok, if it's really ip node - i meant if it's accessible by tcp >all >the time. smtp nodes and ftp ones shouldn't be treated as nodes, as there is >no way to reach them directly. do you know what i mean? I can reach my sysop's FTP server any time. Maybe you thought about FTP /clients/. Email is a cheap way (in countries like Italy where connections to your boss on the other backbone could be really slow) to exchange mail with an internet hub, this is why we should define a way for those nodes to announce this capability. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 SEEN-BY: 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 24 Sep 97 14:42 #207 By: Marco d'Itri To: Jan Ceuleers Re: IP-connectivity St: ------------------------------------------------------------------------------ Jan.Ceuleers@f857.n292.z2.fidonet.org wrote: >,999,Region_29_Igate,Antwerpen,Jan_Ceuleers,01-138-203-224-167,9600,etc >Your reaction was that this cannot work since it clashes with carrier >selection >prefixes in North America, to which I replied that IMHO they do not (since I don't know about North America, but for sure it clashes with italian LD prefixes. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 SEEN-BY: 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 24 Sep 97 14:44 #208 By: Marco d'Itri To: Rune Johansen Re: IP-access St: ------------------------------------------------------------------------------ Rune.Johansen@f20.n210.z2.fidonet.org wrote: >protocols on the same port number. It would be up to the server to >distinguish a >telnet protocol from a VModem protocol, by having some kind of "I want >VModem" >in the negotiation after the TCP sessions has been established, and >falling back The protocol used by the VModem driver on port 23 is just the FTN handshake so it's really easy so write a patched telnet daemon that will recognize FTN "logins" and run the mailer. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 SEEN-BY: 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 24 Sep 97 14:48 #209 By: Marco d'Itri To: Pedro Lima Re: IP-access St: ------------------------------------------------------------------------------ Pedro.Lima@f21.n362.z2.fidonet.org wrote: >The flag fields are limited in size, which may place problems if we're >going to allow several network technologies. Also, I understand most of I suppose that the software with limitations will just ignore the characters it can't store. IP compliant nodes will have to use software that does not truncate flags, IMHO it's not an impossible requirement. >the existant software already works with the address in the phone field. >Furthermore, I think using the phone field to contain the address is the >natural way. I agree, there is no real reason to put it in other places. -- ciao, Marco --- ifmail v.2.11-tx8.5 * Origin: Md's Linux Box - On the Internet: md@linux.it (2:335/610.21) SEEN-BY: 33/101 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 534 600 610 625 362/21 380/106 465/50 480/33 2432/200 SEEN-BY: 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 24 Sep 97 4:58 (Rcvd: Sun 7 Dec 97 14:31) #210 By: Pedro Lima To: Jan Vermeulen Re: Test of an IP-number in the nodelist / Proposal St: Rcvd ------------------------------------------------------------------------------ Hi, JV> I would be able to find a way to do that (I think), but could JV> every Jane or John Doe who's never heard of IP-acces? I really can't tell, all I can say is that it's not common that unexperienced nodes are at a very busy system (but it may happen), that the problem occurs in only very specific conditions and is not generally that serious anyway (if it is serious, then the sysop will try to find out what's happening and solve the problem) and that it has nothing to do with any knowledge of IP-access, but with the configuration possibilities of the mailer used. FWIW, the same argument could be used for a modem trying to call an ISDN-only node, yet I have no knowledge of general problems in that area. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Wed 24 Sep 97 4:46 #211 By: Pedro Lima To: Ger Vloothuis Re: Test of an IP-number in the nodelist / Proposal St: ------------------------------------------------------------------------------ GV> Triple 0 is the emergency phone number in Australia. Something like GV> 911 in the USA and 112 in GMS networks. Don't know if this would be a GV> problem What's your international prefix? If '0' or '00', then indeed a problem may occur if no specific translation (e.g. translate to one's own phone number) or flag filtering is configured. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 17:03 #212 By: Ward Dossche To: All Re: How to not get IP-implemented St: ------------------------------------------------------------------------------ Four sysops are discussing the implementation of IP-connectivity in the nodelist. These sysops are: * Everybody; * Somebody; * Anybody; * Nobody. Now Everybody was sure that Somebody would do it. Anybody could have done it, but Nobody did it. Somebody got angry about that because it was Everybody's job. Everybody thought Anybody could do it, but Nobody realised that Everybody wouldn't do it. It may end up with Everybody blaming Somebody when Nobody does what Anybody could have done. .... Sound familiar to anyone? --- DB 1.58/001874 * Origin: His master's voice (2:292/854.1) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 20:26 #213 By: Ward Dossche To: Ask Bjoern Hansen Re: Flags in nodelist St: ------------------------------------------------------------------------------ Ask, > My suggestion: Don't include IP numbers or hostnames in the nodelist. > Just the flags. You would be reached at f27.n210.z2.fidonet.org; no need > for any other addresses. I dissagree ... your proposal can only work in the true context of what we're discussing here if f27.n210.z2.fidonet.org points to a definite system in the tables and not to a default-address. If f27.n210.z2.fidonet.org, or any other technically capable system for that matter, is not listed in the internet-tables you can forget about direct-addressing between 2 mailers ... which is the issue being discussed here. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 21:52 #214 By: Ward Dossche To: All Re: IP-nodes ... St: ------------------------------------------------------------------------------ At midnight the following will be added to the nodelist ... ; ********************************************************** ; *** IP-entries *** ; ********************************************************** ; ,3000,IP-test,somewhere,Lothar_Behet,000-194-231-142-17,300,CM,U,VM ; ,3001,Eagle's_Nest,Montesilvano_Spiaggia_PE_I,Stefano_Brandimarte,000-195-103 121-1,300,CM,U,IP,TELN,IFC,BND ; --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 21:54 #215 By: Ward Dossche To: Jan Ceuleers Re: MAKENL ! ? St: ------------------------------------------------------------------------------ Jan, > Meanwhile, is it necessary to include leading zeroes, or is existing > software able to use the dashes as separators? The makenl-manual and FTS-0005 are quite strikt ... only dashes and digits. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 26 Sep 97 8:21 #216 By: Rune Johansen To: Ask Bjoern Hansen Re: Flags in nodelist St: ------------------------------------------------------------------------------ >My suggestion: Don't include IP numbers or hostnames in the nodelist. Just >the >flags. You would be reached at f27.n210.z2.fidonet.org; no need for any >other > addresses. No, no, no, no. We must not be tying this to Fidonet and the fidonet.org DNS hierarchy. For that, is several reasons: What if this technique is to be used in other fidonet technology networks? What if fidonet.org disappears? What if I don't have DNS resolver, but only a hosts-list? --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 222/0 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 22:16 #217 By: Ward Dossche To: Rune Johansen Re: Flags in nodelist St: ------------------------------------------------------------------------------ Rune, > BATH: 210/20 27 201/329 292/854 FYI, 212/8 takes this conference directly _here_. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 1:49 #218 By: maciej grzeszczuk To: Ward Dossche Re: Why IP? St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Mon, 22 Sep 97 20:08:19 +0200 Ward Dossche napisal byl w fido.ip_connect: > > when i asked z2.fidonet.org dns maintainer about delegating IN NS record > > for our region as we have number of tcp nodes and this would help us > > maintaining A and MX records i received no answer. wonder why? > You forgot to say "please"? i did say please. even more than once... > Or maybe the postmaster of z2.fidonet.org is concerned about the possible > deterioration of the records. i can't imagine what deterioration do you mean? maintaining our record we'd update mx and a records for tcp nodes regularly and we wouldn't have to bother .z2.fidonet.org postmaster... 238 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: ty jestes glupi arecki! zaraz sie na ciebie obnaz (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 1:50 #219 By: maciej grzeszczuk To: Rune Johansen Re: Why IP? St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Tue, 23 Sep 97 16:04:38 +0200 Rune Johansen napisal byl w fido.ip_connect: > I do not know what reasons the z2 DNS keeper does not answer you. I only > point > out the possibility of the delegation method. and i'm complaining about all the delegation stuff... it's not so easy... 239 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: kanapeczke? (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 1:53 #220 By: maciej grzeszczuk To: Pedro Lima Re: IP-access St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Mon, 22 Sep 97 04:56:02 +0200 Pedro Lima napisal byl w fido.ip_connect: > mg> i don't want to use another useless aka just for my ip address > mg> carried in phone field, as we could use specified flags for it. > That's where we disagree. :-) [...] > FidoNet address into an IP address, or a X.25 address or... For me, it's > only natural that the phone field is used to contain that "other > technology" address. but what with nodes like 2:48/70 - that have both standard modem connection and tcp (vmodem) access? the phone field is only one, so it can carry only phone number, or only ip/domain name address... adding separate aka just for duplicate all information, with one differs - the phone number field - is a bad idea... 240 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: jestes zerem! (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 1:31 #221 By: Denis McMahon To: Pablo Saratxaga Re: IP-access St: ------------------------------------------------------------------------------ * Original message posted in: ENET.SYSOP. * Crossposted in: IP_CONNECT. Pablo Saratxaga wrote to Denis McMahon: DM> OK, so with your proposal all that is needed is a flag for each DM> type of ip connect a system can accept, and the system should be DM> identified using a dns lookup on the f*.n*.z*.fidonet.org DM> address, yes? PS> Yes. OK, so whoever maintains the fidonet.org dns table has to keep it updated for all ip capable nodes in fidonet. Even if this is delegated to the zx.fidonet.org level the administration will be excessive for a single person. Note that in many cases the primary dns entry will not be in the format fa.nb.zc.fidonet.org but will be of the format host.isp.domain, so a lot of CNAME statements will be required to map individual node addresses to their real host names. DM> In that case, we still need a flag for each type of connect DM> available, ie ftp, vmodem session, uucp etc. PS> Yes of course, just as for modem or isdn flags. PS> I also think that if IP connectivity becomes popular and officially PS> supported (eg: listed in the nodelist) then two or three types will PS> remain and others will be deprecated. DM> Otherwise, how do you know whether you should send your pkt file DM> to the ip node using vmodem, uucp or anon ftp transmission? PS> I think that transmission of packets by uucp or ftp will be PS> deprecated as it is complicated, I think you will find that transfer of compressed type 2 and type 2+ fts-001 packet files by ftp has been in use for some time as a means of echomail distribution. I also believe it is how ZEC2 is (or was) importing the Z1 backbone to his system. This is just as valid a means of IP connection as a vmodem session between mailers running over the internet, it's just that the ftp method provides an ip path between node a outbound area and node b inbound area, whereas using vmodem uses ip between the mailers. PS> Vmodem, IFC, binkd (and maybe TransX, I don't remember what it is) PS> will be the winners. Note that the ftp method is already estblished and hs been for some time, the only thing that's missing is it's integration into the nodelist. Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 1:33 #222 By: Denis McMahon To: Rune Johansen Re: IP-access St: ------------------------------------------------------------------------------ * Original message posted in: ENET.SYSOP. * Crossposted in: IP_CONNECT. Rune Johansen wrote to Denis McMahon: >> There's no breaking if the Txx standard; the objective is to use the >> same standard twice in one single entry, so you must be able to >> distinguish between them, must you not? > Does any software actually use the Txx flag? If not, then there is > no reason not to change it! RJ> My software honors the Txx flag, as it should. There is no point in RJ> calling a system that is not online outside the hours indicated by RJ> that flag. OK, in that case the current use of the Txy flag should not be changed. Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 1:37 (Rcvd: Sun 7 Dec 97 14:31) #223 By: Denis McMahon To: Jan Vermeulen Re: IP-connectivity St: Rcvd ------------------------------------------------------------------------------ * Original message posted in: ENET.SYSOP. * Crossposted in: IP_CONNECT. Jan Vermeulen wrote to Denis McMahon: dm> Nodelist telno translation works on the basis that the first section of dm> the telno is the country code. The country code is then either removed dm> and replaced (for own country), or prefixed with idd code (other dm> country). dm> I am merely proposing a third level of translation on the country dm> code field, namely non-analogue-connection/use-gateway, where dm> gateway may be own-system in relevant non-analogue mode (if dm> capable) or another system, identified by a telno. JV> There is a drawback to that solution, however ingenious it is, JV> which will not affect ip-access nodes, but may affect telco JV> oriented nodes to a greater effect than you'd imagine. If you have no gateway telno to translate to, then translate to own-number. Busy will be returned and thus you will not have a system spending large amounts of time listening to a "number misdialled" announcement. However, the hope is that gateway nodes will be defined and the telephone numbers of these nodes can be substituted for the ip systems in the dial translation, so you will send the mail to a gateway node in a packet addressed for the destination node, and the gateway node will handle the onward transmission over ip. Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 1:38 #224 By: Denis McMahon To: All Re: Is anyone else connected? St: ------------------------------------------------------------------------------ Hello All! I am trying to move discussion into here, and I know at least one message made it to somewhere else, but is anyone else reading this? At the moment I have many messages from myself in this area and nothing from others. Regards Denis --- timEd/386 1.10 * Origin: Pickaxe (2:251/20) SEEN-BY: 201/329 212/8 235/100 251/20 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 15:20 #225 By: Pedro Lima To: Marco d'Itri Re: IP-access St: ------------------------------------------------------------------------------ Hi, Md> I suppose that the software with limitations will just ignore the Md> characters it can't store. Maybe, but it would be the characters, not the flags... The real behavior would have to be tested to allow for conclusions. As the limitation is defined by a standard (FTS-5?), this could turn out to be a serious problem. Md> IP compliant nodes will have to use Md> software that does not truncate flags, IMHO it's not an impossible Md> requirement. Eventually, it's possible for a fossil-based browser to be IP-compliant, through RLFOSSIL, which means that most software now being used could be used for it, together with its limitations... All things considered, I think it would be best to try not to exceed the flag fields length. I'm however in favour of changing the nodelist format, and your previous suggestion could be a possible solution. Since that would imply replacing the existant software, we can control the possibilities and limitations. The process of migration between the old and new format would have to be well studied, and it'll certainly require a lot of time (if we started now, I doubt that the migration would be complete before the next millenium ), because of the several implications. It's perhaps the major change one can do in FidoNet. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 26 Sep 97 12:40 (Rcvd: Sun 7 Dec 97 14:32) #226 By: Mats Wallin To: Jan Vermeulen Re: IP-connectivity St: Rcvd ------------------------------------------------------------------------------ JV> Wrong argument for a good cause, Mats: first of all there JV> are far less ISDN-only *sysops* "less ISDN-only sysops" than what? And btw, sysops is completely uninteresting, we're talking about entries in the nodelist, i.e. nodes. I must admit that I haven't checked it lately, but I was under the impression that there were a lot of ISDN-only nodes in the nodelist. Doing a quick grep on the nodelist reveals that there are 2024 lines containing ,300, and X75, and there are also some that are ISDN only using V110 or V120, but I didn't count them. If these people also have an analog line beside this ISDN only line is not relevant, unless I also have a ISDN line, I can't call anyone of these 2024 nodes. JV> and secondly there is no need JV> to repeat the same error over again, is there? I don't see it as an error. It's natural development of FidoNet, starting to use new technologies. Mats mw@defsol.se --- * Origin: Definite APXw (2:201/329.11) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/27 212/8 235/100 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 26 Sep 97 12:51 #227 By: Mats Wallin To: maciej grzeszczuk Re: IP-access St: ------------------------------------------------------------------------------ mg> but what with nodes like 2:48/70 - that have both standard mg> modem connection and tcp (vmodem) access? the phone field is mg> only one, so it can carry only phone number, or only mg> ip/domain name address... adding separate aka just for mg> duplicate all information, with one differs - the phone mg> number field - is a bad idea... It might be, that is unfortunatley the way this is done in FidoNet today. If you want to have two phone numbers for your system, you need two entries in the nodelist. I think it's natural to do it the same way for a TCP/IP node, one entry in the nodelist for each different address. Mats mw@defsol.se --- * Origin: Definite APXw (2:201/329.11) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/27 212/8 235/100 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 26 Sep 97 23:59 #228 By: Ward Dossche To: All Re: A bit of steering ... St: ------------------------------------------------------------------------------ Myladies and Gents, A slight intervention to steer the activities of this group in the right direction. First of all during the past few weeks I have learned a lot from interacting with some of you people. I was quite impressed by the people who actually only want to add some flags in the nodelist and have all the rest done by table look-up. Technically this is feasably, but in a Fidonet-context it is bad medecine. Why? Because including a node would become rather complicated. After having dealt with the fido-structures a future IP-node must equally have its IP-address or similar in the tables maintained by the postmaster at z2.fidonet.org. This means 2 interventions where one would do when using the phone-field. It also means that from a fido-environment we will be deciding the behaviour of people in an internet-environment, we will have to write a policy-document regulating the entry of IP- or likewise-addresses in the internet-tables ... This is complicating the issue uselessly and making it top-heavy. I think we are now looking into 3 things: 1) integrating the full-IP-address in the nodelist; 2) integrating the domain-name in the nodelist; 3) both. The next question is "where"? The phone-field is the obvious sollution but then we need to do something about MAKENL and FTS-0005. Then flags are needed. Which-ones? And finally, a draft-FTS needs to be written and presented to the FTSC. Have I missed something? \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 26 Sep 97 11:08 #229 By: Pedro Lima To: maciej grzeszczuk Re: IP-access St: ------------------------------------------------------------------------------ mg> separate aka just for duplicate all information, with one differs - the mg> phone number field - is a bad idea... Naturally, all the capability flags are different as well. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 26 Sep 97 11:24 #230 By: Pedro Lima To: Denis McMahon Re: Is anyone else connected? St: ------------------------------------------------------------------------------ DM> is anyone else reading this? Yes, and the echo has been busy. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 26 Sep 97 2:24 #231 By: Ask Bjoern Hansen To: Denis McMahon Re: IP Protocols St: ------------------------------------------------------------------------------ Hello Denis! Monday September 22 1997 20:28, Denis McMahon wrote to All: DM> 1) Using normal fido mailers communicating via vmodem fossil drivers. DM> 2) Using ftp protocols to transfer pkt, arcmail and files. DM> 3) Using uuencoded etc email to transfer arcmail and files. DM> I believe that UUCP might also be used like ftp? DM> What about using smtp / pop3 / nntp to transfer netmail / echoes? I DM> would say that at that level, the systems are not using "fido DM> protocols" but are using pure internet protocols. How are you transfering emails [3)]? I'm using pop3 and smtp or imap. DM> I also think that telnet connections are more likely to be virtual DM> bbs, rather than a means of fido communication? Telnet could be used like vmodem. Kind Regards, Ask --- GoldED/2 3.00.Alpha4+ * Origin: PC=Programmable Chaos! Call the Source! <+45-36305533> (2:235/224) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 270/101 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2410/200 2432/0 200 321 700 2433/0 225 1200 1201 1204 SEEN-BY: 2433/1800 2448/10 6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 26 Sep 97 2:26 #232 By: Ask Bjoern Hansen To: Marco d'Itri Re: IP-connectivity St: ------------------------------------------------------------------------------ Hello Marco! Monday September 22 1997 13:57, Marco d'Itri wrote to Denis McMahon: >> SR> as an IP address. I am not sure if IFmail will do the same but >> SR> if someone has IFmail then maybe if it works it could be the >> SR> solution. > Ifmail needs a flag on the command line. I asked Pablo to add the > support for the extended nodelist format and for the IP in the phone > field but I haven't received mail from him since some weeks. The original ifmail author will add it when there's some standard way of finding the ip number of a node. And the program come as source, so anybody could make a patch. :-) Kind Regards, ask --- GoldED/2 3.00.Alpha4+ * Origin: Real power can't be given - it *must* be taken! (2:235/224) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 270/101 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2410/200 2432/0 200 321 700 2433/0 225 1200 1201 1204 SEEN-BY: 2433/1800 2448/10 6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 8:58 #233 By: Steve Woodmore To: maciej grzeszczuk Re: IP-connectivity St: ------------------------------------------------------------------------------ * Replying to a message in : SAVEAREA Hi maciej grzeszczuk, hope you are having a nice day 21-Sep-97 00:44:09, maciej grzeszczuk wrote to Steve Woodmore Subject: Re: IP-connectivity mg> oh yes, yours is accessible via PSTN. i was talking about systems mg> that are not, and only accepting smtp 'connections'. Surely the bare minimum requirement for a node number is to be reachable during ZMH?. All I am thinking of id what if I want to reach an SMTP node with an urgent netmail, and I don't have an internet account? we either have to disallow them from the nodelist unless they can be FTS001 compliant, or have special gateways which are guaranteed to run all day every day. Personally I think we should not even consider SMTP/IP nodes only as a viable option, they really do raise more problems than they solve. -=> Steve Woodmore <=- --- Terminate 5.00/Pro * Origin: inal thought is a strange concept in fidonet (2:254/410.1) SEEN-BY: 2/200 201/329 212/8 235/100 251/25 254/60 61 62 261 410 620 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 442/403 465/50 SEEN-BY: 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 22:04 #234 By: Ward Dossche To: Steve Woodmore Re: IP-connectivity St: ------------------------------------------------------------------------------ > Personally I think we should not even consider SMTP/IP nodes only as a > viable option, they really do raise more problems than they solve. We are not discussing the solving of problems which have not been raised. The true subject is 'how to embrace new transmission-technologies'. The hidden agenda is 'how to nuke FTS-0001'. A usefull dangerous statement for idiots who have no intention to be constructive nor to foster advancement. What we're also doing here is paving the way to hell for the FTSC, but they haven't caught on yet. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 22:10 #235 By: Jere Salonen To: Denis McMahon Re: Is anyone else connected? St: ------------------------------------------------------------------------------ > Hello All! >I am trying to move discussion into here, and I know at least one message >made > it to somewhere else, but is anyone else reading this? At the moment I have > many messages from myself in this area and nothing from others. I am. > Regards > Denis --- BBBS/2 v3.42 ToMmIk-6v * Origin: BCG-Box 4 (2:222/0) SEEN-BY: 20/11 22/10 201/300 307 329 330 203/614 210/20 27 212/8 222/0 SEEN-BY: 235/100 254/60 280/100 292/854 857 876 335/533 362/21 380/106 SEEN-BY: 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 1:52 #236 By: Lech Szychowski To: Jan Ceuleers Re: IP-access St: ------------------------------------------------------------------------------ > My main problem is that due to the caching function I mentioned above, > it doesn't seem to be possible to accomodate nodes whose ISP dynamically > assigns them an IP address as they log on. I don't think DNS can cope > with that, but I'd be happy to be proved wrong. My first thought is setting TTL to near-zero seems to be one possible solution. Although I have to admit that I don't remember what RFCs say about minimal TTL (if this issue is dealt with at all). And I hate what this means forthe packet traffic and namesver load; delegating authority to regions or even networks might be necessary. OTOH I doubt if there is much sense in allowing for dial-up nodes (at least when it comes to them not being the connection initiating party). Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 2:03 #237 By: Lech Szychowski To: Pedro Lima Re: IP-connectivity St: ------------------------------------------------------------------------------ > I'm not worried about breaching Policy where a clear need exists for it, Certainly. Policy4 never addressed issues we are discussing right now because at the time it was created such issues were well beyond the scope of interest and/or imagination. Time passes fast, technology advances even faster and we just have to either write an amendment to P4 or... (I wouldn't dare mentioning the alternative option). Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 2:11 #238 By: Lech Szychowski To: Pedro Lima Re: IP-access St: ------------------------------------------------------------------------------ > I thought it was possible for a name server, not being a primary nor a > secondary name server for the requested domain, to cache the IP address > of that domain, but I must be wrong, then. In that case, a FQDN You are right. But no DNS record should be cached for a period of time longer than its TTL says. > If the system knows that client X is now connected to a certain IP > address, it can create the corresponding entries in the DNS database > files and update their SOA serial numbers. Right again - and again a "but" :) But the problem is so far almost no DNS servers support triggered update scheme; classical DNS servers use polling approach. New bind has "notify" capabilities built in, but it's still pretty rarely used (being still a beta, I believe). Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 2:24 #239 By: Lech Szychowski To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ >> Well... Being a point is not a dishonour. > Maybe it is turning the world upside down. IP-capable points could > become the major feed of the non-IP-capable boss-system. What about using a very simple approach: if you don't feed any mail to anyone (ie you are a "leaf" in a routing/distribution tree) then there is (almost) no reason for you to be a node. Please keep in mind that most of the dial-up connected systems would do fine as points (nodelist problems saved herewith) of a boss that is a leased line / permanent connection system. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 2:37 #240 By: Lech Szychowski To: Mats Wallin Re: IP-connectivity St: ------------------------------------------------------------------------------ > Just because you're using Internet to connect to your hub/host doesn't > mean that you're not interested in being a member of FidoNet, which for > instance makes it possible to vote in elections. True enough. But let's put it this way: a non-permanent IP/FQDN assignment implies dialup connection which implies system equipped with a modem and having access to PSTN line which in turn implies if system operator really wants to be a node he can very well become a traditional technology based one. Lets face it: DNS in its present shape does not provide much room for permanent assignment of FQDN to a dynamically changing IP. There might be some changes coming, even soon (the demand for such features becomes immense). But this only makes eveything easier - if we arrange everything for "permanent IP<->FQDN" nodes to be able to join Fido, the very same solution will work for dynamically assigned IPs as soon as the "dynamic DNS" problem is solved. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 2:43 #241 By: Lech Szychowski To: maciej grzeszczuk Re: IP-access St: ------------------------------------------------------------------------------ > the only way to reach immediatelly fidonet mailer is emsi. smtp or fts > isn't a way. For the time being that's AFAIK true. Which does not mean we cannot witness some FTS-1 capable software appear soon (although I have to admit I doubt it). Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 2:45 #242 By: Lech Szychowski To: rafal wiosna Re: what about... St: ------------------------------------------------------------------------------ > mg> -unpublished-. pvt nodes makes no dishonor to it's operator - he's a > mg> member with the same rights as non-pvt one. > What is the difference between a Pvt node or a point in this case? Substantial: being a Fido member or not. > Nope. TCP means ifcico, VMODEM is self-explanatory. TCP should not be used this way. It's a short-sighted idea to use a generic name to denote a specific product/entity. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 2:53 #243 By: Lech Szychowski To: rafal wiosna Re: Hmm. St: ------------------------------------------------------------------------------ > nodes and non-CM IP-systems should remain points? But OTOH why bother IMHO a non-CM system may be a node as long as there is a way to make an incoming conection to it (ie one can make out it IP address and exact time range it's online). Think this way: a PSTN node has to be accesible only during ZMH, so why not have IP nodes comply with this requirement? Whether "IP-ZMH" shold be the same one as "PSTN-ZMH" is another question, time-related traffic costs changes not necesarily being an issue with Internet. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 3:04 #244 By: Lech Szychowski To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ > So are you saying that soon you will give up _your_ nodenumber, or that > you think it's the guy next-door who should give up _his_ nodenumber? If they are to continue have their systems being reachable/dialable (ZMH, FTS et al) there is no need for them to drop nodenumbers. What's raison d'etre of the nodelist? IMHO it's rather making it possible for one system/mailer to contact another than being a members list. I think one of the original reasons for introducing an institiution of a point was necessity to deal with such situation: FTS-compliant system not being able to support fixed time-and-number reachability. And that's exactly the situation we have now with dial-up dynamically IP/FQDN assigned nodes. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 3:16 #245 By: Lech Szychowski To: Pedro Lima Re: IP-access St: ------------------------------------------------------------------------------ > mg> the only way to reach immediatelly fidonet mailer is emsi. smtp or fts > mg> isn't a way. > Why not? (1) It looks that so far we've been - more or less - looking at TCP/IP just as a carrier layer for FTS protocols. (2) Dealing/talking with non-Fido protocols/software (such as SMTP MTAs) is not a Fido software responsibility/job. If I am to send mail to SMTP-talking system, I use SMTP-talking software myself. If I am to call FTS-aware/compliant system, I use FTS-aware/compliant software. I don't think anyone really wants to write an all-protocol mailer integrating all possible protocols - FTS-1, FTS-6, EMSI, SMTP, FTP, FSP and so on... :) Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 3:26 #246 By: Lech Szychowski To: rafal wiosna Re: IP-connectivity St: ------------------------------------------------------------------------------ > MW> systems do not allowe points to request files, etc. > Unlisted, that is. With Internet involved, regional/netional pointlists will stop being a solution to this problem - there will be no difference between "calling" your boss, someone in your region or some system in another zone. > If that's true no IP-only node should be allowed. They can't be > reached during zone's ZMH. If we extend the meaning of "reachability" to include IP as well as PSTN - and it looks like we'll almost certainly have to do this - then this problem gets reduced to the "IP-nodes vs IP-points" questuion. > So, I ask again: what's the point of becoming such a leaf node > then? Only to be able to vote? I'd say if someone wants to have some privileges he should also accept some obligations. Want to be a node? Sure, no problem. Just satisfy the "ZMH reachability" condition (possibly extended from PSTN-only to PSTN and/or IP). Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 3:31 #247 By: Lech Szychowski To: Steve Woodmore Re: IP-connectivity St: ------------------------------------------------------------------------------ > Incorrect, My system is online 24hrs/day via PSTN. So you can be (well, actually you already are :) a node. > You can also send internet email to my IP address 24 hrs/day This is a non-issue here, right? > What you cannot do is to connect to my system Via the internet, unless I > am online. And it means that (in both Maciek's and my opinions) this connectivity alone is not enough to make you qualify as an IP-node. Of course, it does not change your existing PSTN-based node status. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 3:37 #248 By: Lech Szychowski To: maciej grzeszczuk Re: IP-access St: ------------------------------------------------------------------------------ > carrying one aka? fe. at 2:480/70 modem answers on number > +48-22-8313212, and ifcico listens at 195.117.131.17:60179. > i don't want to use another useless aka just for my ip address > carried in phone field, as we could use specified flags for it. Ooops! Here we hit another problem: multiple-homed IP nodes. And it's not a theoretical one: 2:480/17 aka 2:48/48 (R48 Inet gate) can be reached at three different IP addreses, each of them in a different C class, each of them differently routed. Looks like we have to make some arrangements to be able to cope with such cases too. Another address, as with multiple PSTN lines? Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 3:40 #249 By: Lech Szychowski To: Rune Johansen Re: Proposal? St: ------------------------------------------------------------------------------ > - ICQ (paging/transfer system) Is ICQ worth thinking about? Is it at least available on more than one platform - or will it be soon? I doubt. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 3:43 #250 By: Lech Szychowski To: Pedro Lima Re: IP-connectivity St: ------------------------------------------------------------------------------ > denying the possibility of a connection, since AFAIK, it's not possible > to send e-mail to a disconnected IP address. :-) IP address? No. There are MX records, but I have yet to see MTA trying to query DNS for an MX record of an *.in-addr.arpa. entity :) Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 3:48 #251 By: Lech Szychowski To: Marco d'Itri Re: Why IP? St: ------------------------------------------------------------------------------ > - I can change the IP of my node and then reload the zone of my domain, > but if the nodelist has my IP I will be unreachable for some days So would you if you were to change your phonenumber in between the days nodelist is issued, right? Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 3:52 #252 By: Lech Szychowski To: Mats Wallin Re: IP-connectivity St: ------------------------------------------------------------------------------ > If you don't have the person in the nodelist, you will have to remember > and type the address manually. C'mon, all decent mail editors support aliases. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 3:58 #253 By: Lech Szychowski To: Marco d'Itri Re: IP-access St: ------------------------------------------------------------------------------ > If you use FTP [1] (as a server) you have the same capabilities of a > normal node, the only difference is that you don't use FTS-1 or EMSI. Nope. As you most probably know, FTP server cannot initiate file transfer to its client. It can't even notify the client it would like to :) Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 4:07 #254 By: Lech Szychowski To: Ward Dossche Re: Why IP? St: ------------------------------------------------------------------------------ > Or maybe the postmaster of z2.fidonet.org is concerned about the > possible deterioration of the records. Deterioration in what sense? There are three of us from R48 taking part in this echo and each of us is an un*x LAN/WAN/hosts administrator, working for quite large company or educational institiution like a university department. We do maintains multiple DNS domains at our sites as a part of our job and I dare say we ary quite capable of doing it. We are not perfect but we are no DNS newbies either. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 4:12 #255 By: Lech Szychowski To: Rune Johansen Re: IP-connectivity St: ------------------------------------------------------------------------------ > Would you then imply that Fido2Int-capable-only nodes should not be in > Fidonet? If they qualify as Fido nodes now, they are nodes and so they will be. If they stop accepting calls in ZMH (whether this means "PSTN-ZMH/calls" or "IP-ZMH/calls" is a different issue), they lose their node status. BTW: If we are to allow IP-only nodes, we have to define minimal required capabilities for such nodes. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 4:22 #256 By: Lech Szychowski To: Marco d'Itri Re: Observations and questions St: ------------------------------------------------------------------------------ >>That's why we HAVE to allow for IP addresses AND fqdn symbolic addresses in [...] > There is no need to have both, the FQDN is enough. Should be enough, that is :) IMHO it would be very nice if we could have both, just in case. You know, Murphy never sleeps; in a few days right after accepting an FQDN-only standard we might come upon a problem that we had not thought about - nobody's perfect. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 23 Sep 97 8:50 #257 By: Lothar Behet To: rafal wiosna Re: IP-connectivity St: ------------------------------------------------------------------------------ Hello rafal! On 19 Sep 97 wrote rafal wiosna to Mats Wallin: rw> If that's true no IP-only node should be allowed. They can't rw> be reached during zone's ZMH. Why not? May be not all, but some are even running CM. Gruss Lothar --- GoldED/2 3.00.Alpha5+ * Origin: Life: a bad game, but graphics is really good (2:2446/301) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2410/200 2432/0 200 321 700 2442/2000 2446/0 300 301 304 SEEN-BY: 2446/305 2448/6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 23 Sep 97 8:55 #258 By: Lothar Behet To: Ward Dossche Re: MAKENL ! ? St: ------------------------------------------------------------------------------ Hello Ward! On 21 Sep 97 wrote Ward Dossche to All: WD> I've already had some remarks re: the hyphens in the IP-number of the WD> test-node (only in the world-nodelist of zone-2 !!). People request me WD> to put dots there, this is what makenl says : The dot-problem may be solved by a special translation within the nodelist compiler, but supporting dns (names like vmp.nrh.de) would be important, too. WD> ************************* QUOTE ************************* WD> Begin processing NODELIST -- 21:26, Sunday, September 21, 1997 WD> ,3000,IP-test,somewhere,Lothar_Behet,000-194.231.142.17,300,VM WD> D:\MAKENL\CTL-FILE\TEST.CTL -- Invalid phone number -- WD> "000-194.231.142.17" WD> ************************ UNQUOTE ************************ WD> Which is fully compliant with FTS-0005. Regards Lothar --- GoldED/2 3.00.Alpha5+ * Origin: Life: a bad game, but graphics is really good (2:2446/301) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2410/200 2432/0 200 321 700 2442/2000 2446/0 300 301 304 SEEN-BY: 2446/305 2448/6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 23 Sep 97 11:14 #259 By: Lothar Behet To: Rune Johansen Re: Proposal? St: ------------------------------------------------------------------------------ Hello Rune! On 21 Sep 97 wrote Rune Johansen to All: The first part seems quite ok, except I would add Transx to the protocols. RJ> I am aware of that the placement of the receiving address is not by RJ> the time being finally placed, but I would assume that we will use the RJ> "phone" field of the nodelist entry to denote this. The phone field is RJ> being used for letting your application know where it should RJ> "connect", by looking at this field. This is supported by FTN-based software, except they need special entries, which are at this moment not compliant with FTS-0005. RJ> For DIRECT connections this would be a network address (IP number, RJ> DNS address, X25 id, ICQ number, whatever). For INDIRECT connections RJ> it would be the receivers address (email address, FTP server IP/DNS, RJ> whatever). I was not able to call by example a dns-entry with another (analogue or ISDN) line, could anyone else manage it? If this could be confirmed, dns-entries in the phone field would prevent "normal nodes" from dialing incompatible systems. RJ> Example for DIRECT connection: RJ> ...,SysOp,bbs.some.net,9600,CM,U,IP,ITEL,VMO RJ> ...,SysOp,12.34.56.78,9600,CM,U,IP,ITEL:1024,VMO,IFC,BNK:234 ISDN-lines use a speed of 300, if they are not capable for analogue connects. The same applies with IP-connections we are talking about. Another thing is a system, which supports direct dialin (i.e. ppp), which needs a real phonenumber and flags. Fido is at least a network of mailbox sysops. Should this information maintained by the the nodelist, too? IP-access for routing purposes is only one aspect, but IMHO at this moment the most important. But additional flags for dialup lines should be considered at this moment, because they will be the next step. They may be marked as user flags in the form of: ...,U,PPP,... ...,U,SLIP,... Regards Lothar --- GoldED/2 3.00.Alpha5+ * Origin: Life: a bad game, but graphics is really good (2:2446/301) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2410/200 2432/0 200 321 700 2442/2000 2446/0 300 301 304 SEEN-BY: 2446/305 2448/6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 23 Sep 97 11:18 #260 By: Lothar Behet To: Ger Vloothuis Re: Test of an IP-number in the nodelist / Proposal St: ------------------------------------------------------------------------------ Hello Ger! On 21 Sep 97 wrote Ger Vloothuis to Ward Dossche: GV> Triple 0 is the emergency phone number in Australia. Something like GV> 911 in the USA and 112 in GMS networks. Don't know if this would be a GV> problem but the emergency line operators would sure be hindered by a GV> lot of unintended modem calls. So that should be avoided. Do you use the raw nodelist data for dialing or is there any kind of dial translation in your nodelist compiler? I hope, that most (node-)setups will respect even the flags in the appropriate nodelist fields :) Regards Lothar --- GoldED/2 3.00.Alpha5+ * Origin: Life: a bad game, but graphics is really good (2:2446/301) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2410/200 2432/0 200 321 700 2442/2000 2446/0 300 301 304 SEEN-BY: 2446/305 2448/6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 7:06 #261 By: Lothar Behet To: All Re: Status 2:2/3000 St: ------------------------------------------------------------------------------ On 23 Sep 97 wrote Ward Dossche to Lothar Behet: WD> I understand some guys have already called you using VModem and the WD> test-nodelist-entry. WD> Can you report on this please? Yes, there were mailer connects with 5 different systems beside the regular transfers on the vmodem-line. 1 from sweden, 2 from USSR and 2 from germany. Furthermore there were 7 connects spawning to the bbs, whereas 4 of these did identify themselfs within the user application form. The other 3 did disconnect within about a minute. With my nodelist compiler (Fastlst V2.0) the entry in the nodelist 000-194-231-142-17 can not be translated to a callable ip-number for vmodem automatically, so i handle it with a private list for regular outgoing calls. Makenl 2.51 could not manage a usable ip# or dns entry for vmodem, whereas fastlst (in a private list) and Binkley XE4/Vmodem support different formats (using "*" and ".") for calling ip entries and fqdn (fully qualified domain name). There is still something to do about integrating ip nodes in the nodelist, but it works already. As soon as a standard format for ip nodes should be specified, at least the vital information may be extracted from the nodelist and converted to a private list according to the requirements of different programs for ip access. Regards Lothar --- GoldED/2 3.00.Alpha5+ * Origin: Life: a bad game, but graphics is really good (2:2446/301) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2410/200 2432/0 200 321 700 2442/2000 2446/0 300 301 304 SEEN-BY: 2446/305 2448/6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Thu 25 Sep 97 21:12 #262 By: Lothar Behet To: Marco d'Itri Re: It's in ... St: ------------------------------------------------------------------------------ Hello Marco! On 24 Sep 97 wrote Marco d'Itri to Ward Dossche: MI> ; U,xxxxx - marks non-standard port number, relates to _first_ MI> ; capability flag in the line (only one non-standard port MI> ; can be lested). MI> IMHO we should not specify non-standard ports this way because it MI> supports just one service and it's difficult to parse by programs. So MI> far the best proposal (except the extended nodelist one) is to add MI> something like TEL:60177. Seconded. There may be run more than one protocol on one Node and each may use it's own port. All are addressed to the same machine. MI> ; U,"foo.bar" - marks corresponding domain address (temporarily) MI> I don't think it's useful, the domain should go in the phone field, MI> if mailers can't use it, it's pointless to add it. With dns several tasks serving the same protocol may be spread over more than one machine. Additionaly the nodelist entry may stay, even if a service runs on another machine with another ip. I think, that every mailer over ip will work with fqdn, but beside makenl did somebody check nodelist compilers? With the working software on most systems at this moment, the fqdn in the flag field might not interfere with it, but who did check it until now? By example Frontdoor uses its own, but which limitations does it put an extended nodelist format, as many fd systems will stay online for probably several years. Regards Lothar PS: I will check Frontdoor this weekend, maybe somebody else could check other programs and post the results in this area. --- GoldED/2 3.00.Alpha5+ * Origin: Life: a bad game, but graphics is really good (2:2446/301) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2410/200 2432/0 200 321 700 2442/2000 2446/0 300 301 304 SEEN-BY: 2446/305 2448/6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 26 Sep 97 21:21 #263 By: Lothar Behet To: Ward Dossche Re: How to not get IP-implemented St: ------------------------------------------------------------------------------ Hello Ward! On 25 Sep 97 wrote Ward Dossche to All: WD> Four sysops are discussing the implementation of IP-connectivity in WD> the nodelist. These sysops are: WD> * Everybody; WD> * Somebody; WD> * Anybody; WD> * Nobody. [... some stuff deleted ...] WD> Sound familiar to anyone? Hey man, that's real fido :) Regards Lothar --- GoldED/2 3.00.Alpha5+ * Origin: Life: a bad game, but graphics is really good (2:2446/301) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2410/200 2432/0 200 321 700 2442/2000 2446/0 300 301 304 SEEN-BY: 2446/305 2448/6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Fri 26 Sep 97 21:50 #264 By: Lothar Behet To: Ward Dossche Re: MAKENL ! ? St: ------------------------------------------------------------------------------ Hello Ward! On 25 Sep 97 wrote Ward Dossche to Jan Ceuleers: >> Meanwhile, is it necessary to include leading zeroes, or is existing >> software able to use the dashes as separators? WD> The makenl-manual and FTS-0005 are quite strikt ... only dashes and WD> digits. Not only the manual, even the program knows this :( But it should be possible, to put the fqdn in the system name field. This would be vmp.nrh.de for 2:2/3000 and Vmodem (from Ray Gwinn's SIO package) services port 3141. I hope, this helps some people, who tried to connect to 2:2/3000. Regards Lothar --- GoldED/2 3.00.Alpha5+ * Origin: Maybe it read its own manual (2:2446/301) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2410/200 2432/0 200 321 700 2442/2000 2446/0 300 301 304 SEEN-BY: 2446/305 2448/6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 10:19 #265 By: Lothar Behet To: Mats Wallin Re: IP-access St: ------------------------------------------------------------------------------ Hello Mats! On 26 Sep 97 wrote Mats Wallin to maciej grzeszczuk: MW> It might be, that is unfortunatley the way this is done in FidoNet MW> today. If you want to have two phone numbers for your system, you need MW> two entries in the nodelist. There are two different things: The software we use today and a new nodelist standard, which supports extended capabilities in the future. At this moment, IP-lines are not supported by makenl and nodelist compilers directly, but some useful information can be spread, if a certain definition for the ip connectivity would be created. This information can be extracted by small utilities, which are used by nodes, who need it. The test ip nodes use an ip in place of the phone number, but this format is (mostly?) not usable directly for ip mailers. Because of makenl's limitations there might even happen unwanted calls to wrong numbers. IMHO we should do two things: 1. Define an intermediate standard which is supported by makenl and the existing nodelist compilers, that does not cause problems for unprepared systems. 2. Create an extended nodelist definition for the future. This will last some time. On short term, an intermediate may be defined by: The system name is used for the ip address or fqdn (in the last case not without sense:). As some people asked me, where "Somewhere" of 2:2/3000 is located, the location should still be available. The phone number is reserved for PSTN calls, which uses the standard flags. As flag space is limited and several lines might be needed, "-unpublished-" may be used for _additional_ lines. IP-only nodes are a seperate point to discuss and not part of this proposal. A certain amount of flags can be spread in the nodelist, but concerning to the limitations by makenl and some nodelist compilers (still to be checked!) might cause several lines for more than one protocol on the same node. MW> I think it's natural to do it the same way for a TCP/IP node, one MW> entry in the nodelist for each different address. Different (ip-)protocols may run on the same address, as they are serviced by different ports on the same system. Denis McMahon did a test on makenl, which failed: > ,10,test_10,nowhere,no-one,44-1234-567890,9600,CM,XA,H16,V34, > V32T,VFC,V110L,V110H,V120L,V120H,X75,X2S,IP123.221.65.9, > DNSalpha.beta.gamma.delta,VM65534,FTP65535,GUUCP,NEC,NC This sample entry still does not contain all possible combinations, but in practice a standard node would normally not use more than 3 protocols (IMHO). The intermediate definition might look like this: - the system name field contains an ip# or fqdn - user flags might be: - VMP[:3141] Vmodem, default port 3141 - IFC ifcico - BNP BinkD - ... to be continued ... One of the defined flags or "IP" might advice the (additional) nodelist utility, that this nodes system name contains an usable (ip) address. The utility program scans the nodelist for these entries and puts them in a private list, which is used by the "ip-mailer". Default ports should be noted in the nodelist comments and only deviating ports need to listed in the node entry. Some protocols, like TransX or FTP, may not be directly in the sense of Fido, but they are already in use. They should be listed too, as they decrease cost for transfer on long distance calls. In this case, at least the capability for a protocol is shown in the nodelist and the sysops may arrange the rest by netmail. If this is done in short term, then we have enough time to define a new nodelist format, which contains all required information for the next millenium. But in the meantime the interested sysops are able to retrieve the required information from the nodelist. Regards Lothar (aka 2:2/3000) PS: How do programs handle the T-flag? Do they only honor it, if it is exactly 3 characters long? Do they only use the first flag, starting with a "T"? --- GoldED/2 3.00.Alpha5+ * Origin: Life: a bad game, but graphics is really good (2:2446/301) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2410/200 2432/0 200 321 700 2442/2000 2446/0 300 301 304 SEEN-BY: 2446/305 2448/6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 15:01 #266 By: Lothar Behet To: Ward Dossche Re: IP-nodes ... St: ------------------------------------------------------------------------------ Hallo Ward! Am 25 Sep 97 schrieb Ward Dossche an All: WD> ; ********************************************************** WD> ; *** IP-entries *** WD> ; ********************************************************** WD> ; WD> ,3000,IP-test,somewhere,Lothar_Behet,000-194-231-142-17,300,CM,U,VM You may add BND or BNP and TELN, too :) All service the default ports. "IP-Test" may be changed to vmp.nrh.de for people, who prefer fqdn. Location is Uedem, Germany. WD> ; WD> ,3001,Eagle's_Nest,Montesilvano_Spiaggia_PE_I,Stefano_Brandimarte,000 WD> - WD> 195-103 121-1,300,CM,U,IP,TELN,IFC,BND ; WD> -+- DB 1.58/001877 WD> + Origin: Many Glacier (2:292/854) Gruss Lothar --- GoldED/2 3.00.Alpha5+ * Origin: Zum Maeusefangen brauchen sie eine Katze von (2:2446/301) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2410/200 2432/0 200 321 700 2442/2000 2446/0 300 301 304 SEEN-BY: 2446/305 2448/6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 17:48 #267 By: Lothar Behet To: Ward Dossche Re: A bit of steering ... St: ------------------------------------------------------------------------------ Hello Ward! On 26 Sep 97 wrote Ward Dossche to All: WD> I think we are now looking into 3 things: WD> 1) integrating the full-IP-address in the nodelist; WD> 2) integrating the domain-name in the nodelist; WD> 3) both. While 2 is the most practical, 1 (and 3) should be possible. WD> The next question is "where"? The phone-field is the obvious WD> sollution but then we need to do something about MAKENL and FTS-0005. First somebody has to check, which problems will arise with all the used software (nodelist compilers primarily), if fts-0005 (and makenl) would be changed. It's not a wise idea to implinent ip-access in nodelist, while others would run into problems. WD> Then flags are needed. Which-ones? All :) At this moment there are several direct protocols (Vmodem, BinkD, ifcico, ...) and some indirect (TransX, FTP, ...), which are already in use. WD> And finally, a draft-FTS needs to be written and presented to the WD> FTSC. Marco's extended nodelist format sounds god, but what about the existing software? After we have checked the compatibility with existing software, such a draft may be written. If a few people check it and deliver the information to this area, this task should be done in a short term. WD> Have I missed something? Do you know anybody, who never misses something? :) Regards Lothar --- GoldED/2 3.00.Alpha5+ * Origin: Life: a bad game, but graphics is really good (2:2446/300) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2410/200 2432/0 200 321 700 2442/2000 2446/0 300 301 304 SEEN-BY: 2446/305 2448/6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 9:26 #268 By: Steve Woodmore To: Pedro Lima Re: IP-connectivity St: ------------------------------------------------------------------------------ * Replying to a message in : SAVEAREA Hi Pedro Lima, hope you are having a nice day 21-Sep-97 05:35:49, Pedro Lima wrote to Steve Woodmore Subject: IP-connectivity PL> Then either the IP address isn't in fact yours, or you're PL> purposely denying the possibility of a connection, since AFAIK, PL> it's not possible to send e-mail to a disconnected IP address. :-) The IP address is mine, and you can send email to it at any time, what I said was you cannot connect to my system unless I am online -=> Steve Woodmore <=- --- Terminate 5.00/Pro * Origin: inal thought is a strange concept in fidonet (2:254/410.1) SEEN-BY: 2/200 201/329 212/8 235/100 251/25 254/60 61 62 261 410 620 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 442/403 465/50 SEEN-BY: 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 9:28 #269 By: Steve Woodmore To: Rune Johansen Re: IP-connectivity St: ------------------------------------------------------------------------------ * Replying to a message in : SAVEAREA Hi Rune Johansen, hope you are having a nice day 23-Sep-97 17:07:46, Rune Johansen wrote to Steve Woodmore Subject: Re: IP-connectivity RJ> Would you then imply that Fido2Int-capable-only nodes should not RJ> be in Fidonet? Correct, unless they had the ability to be online on PSTN during ZMH -=> Steve Woodmore <=- --- Terminate 5.00/Pro * Origin: You know when you have been Terminated... (2:254/410.1) SEEN-BY: 2/200 201/329 212/8 235/100 251/25 254/60 61 62 261 410 620 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 442/403 465/50 SEEN-BY: 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 14:46 #270 By: Sean Rima To: All Re: DSP IP Stats for Interest St: ------------------------------------------------------------------------------ Below are the stats for a 4 day period for the DSP IPMAILER. This is for info and to open the area on my IPMAILER. Sean BinkD Log analyzer. (C) /IP. 2:5020/269@fidonet. v0.13 (DOS) Analyzed period: from 23 Sep 13:02:26 to 27 Sep 14:15:16 ÚÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÂÄÄÄÄÄÄÄÄÂÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÂÄÄÄÄÄÄÄ¿ ³ Address ³Sessions³ Transfer, bytes ³Average³ ³ ³ ÃÄÄÄÄÄ rcvd ÄÄÄÄÂÄÄÄÄ sent ÄÄÄÄÄ´ CPS ³ ÃÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄ´ ³1:167/323.0 ³ 59 ³ 11,616,207³ 534,873³ 2583³ ³2:250/380.0 ³ 66 ³ 1,341³ 132,545³ 3938³ ³2:251/60.0 ³ 30 ³ 867³ 2,384,563³ 1665³ ³2:252/356.0 ³ 115 ³ 4,646,273³ 15,836,336³ 2603³ ³2:252/503.0 ³ 7 ³ 1,114³ 66,990³ 2961³ ³2:350/123.0 ³ 2 ³ ³ 411,383³ 95³ ³2:442/403.0 ³ 20 ³ 1,085,195³ 2,373,986³ 2230³ ³2:469/38.0 ³ 26 ³ 1,188³ 510,150³ 691³ ³2:490/63.0 ³ 7 ³ ³ 282,266³ 1069³ ³2:5020/180.0 ³ 230 ³ 34,401³ 1,801³ 260³ ³2:5020/204.0 ³ 2 ³ 545³ 3,576³ 589³ ÃÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄ´ ³> Totals:³ 564 ³ 17,387,131³ 22,538,469³ 1893³ ÀÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÁÄÄÄÄÄÄÄÄÁÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÁÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÁÄÄÄÄÄÄÄÙ *Mail* transfer ÚÄÄÄÄÄÄÄÄÂÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÂÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÂÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄ¿ ³ ³ Total ³ Sent ³ Recieved ³ ³ ÃÄqtyÄÂÄÄÄÄ bytes ÄÄÄÄÅÄqtyÄÂÄÄÄÄ bytes ÄÄÄÄÅÄqtyÄÂÄÄÄÄ bytes ÄÄÄÄ´ ÃÄÄÄÄÄÄÄÄÅÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄ´ ³ Mail:³ 39 ³ 276,480³ 33 ³ 273,174³ 6 ³ 3,306³ ³Arcmail:³ 740 ³ 34,344,424³ 481 ³ 16,964,256³ 259 ³ 17,380,168³ ³ Files:³ 103 ³ 5,304,696³ 4 ³ 5,301,039³ 99 ³ 3,657³ ÀÄÄÄÄÄÄÄÄÁÄÄÄÄÄÁÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÁÄÄÄÄÄÁÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÁÄÄÄÄÄÁÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÙ ----- MEMO.TXT ends ----- --- Msged/NT 4.10 * Origin: DSP BBS, Reading, Berks (2:252/356@fidonet) SEEN-BY: 25/10 50 201/329 212/8 235/100 250/102 183 191 210 607 251/25 SEEN-BY: 251/56 252/356 358 254/60 61 62 261 410 620 256/97 183 609 SEEN-BY: 257/71 609 258/6 259/66 280/100 292/854 857 876 335/533 362/21 SEEN-BY: 380/106 440/1 7 442/101 401 403 404 465/50 480/33 2432/200 SEEN-BY: 2501/204 2502/1000 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 23:49 #271 By: Ward Dossche To: Lech Szychowski Re: IP-connectivity St: ------------------------------------------------------------------------------ > What about using a very simple approach: if you don't feed any mail > to anyone (ie you are a "leaf" in a routing/distribution tree) then > there is (almost) no reason for you to be a node. Fine .... then the leaf-nodes dissappear ... which makes their former feeds leaf-nodes ... for which reason they need to dissappear etc... and 3 weeks later fidonet can close down. > Please keep in > mind that most of the dial-up connected systems would do fine as > points (nodelist problems saved herewith) of a boss that is a leased > line / permanent connection system. In a very strict and certain sense this is already the case now ... \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 0:00 #272 By: Ward Dossche To: Lech Szychowski Re: IP-connectivity St: ------------------------------------------------------------------------------ > Certainly. Policy4 never addressed issues we are discussing right now > because at the time it was created such issues were well beyond the > scope of interest and/or imagination. Time passes fast, technology > advances even faster and we just have to either write an amendment > to P4 or... (I wouldn't dare mentioning the alternative option). Changing FTS-0001 would be fine ... but someone has a copyright on it and changing the text is forbidden via that same copyright-statement. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 0:03 #273 By: Ward Dossche To: Lech Szychowski Re: Why IP? St: ------------------------------------------------------------------------------ >> Or maybe the postmaster of z2.fidonet.org is concerned about the >> possible deterioration of the records. > Deterioration in what sense? There are three of us from R48 taking > part in this echo and each of us is an un*x LAN/WAN/hosts administrator, > working for quite large company or educational institiution like > a university department. We do maintains multiple DNS domains at > our sites as a part of our job and I dare say we ary quite capable > of doing it. We are not perfect but we are no DNS newbies either. There is no need defending yourself or digging your heels in over what i call a non-issue. The postmaster here has my 100% support in how he manages the tables. You don't hear me saying what an important big-shot I am in business-life, so I'd appreciate if you wouldn't make such remarks neither. Please stick to the subject of this conference : "how to integrate IP in the nodelist?". Thank you. \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 6:22 #274 By: Pedro Lima To: Ward Dossche Re: A bit of steering ... St: ------------------------------------------------------------------------------ WD> a policy-document regulating the entry of IP- or likewise-addresses in WD> the internet-tables ... WD> This is complicating the issue uselessly and making it top-heavy. Agreed, we can do it later on. WD> Then flags are needed. Which-ones? How about: BND - BinkD (24554) IFC - ifcico (60179) ITN - telnet (23) RAW - raw socket (60179) VMP - VModem (3141) If I understood correctly, these are the current protocols which allow for a mailer-to-mailer connection. Non-standard ports may be listed after the flag. I'd suggest a "dual flag" like ITN,60177 instead of ITN:60177 or ITN60177 because I suspect current mailers determine flags by looking at what's separated with commas, and it would then be theoretically simple to prevent calls to IP-nodes on non-IP mailers. E-mail, TRANX, ICQ, ftp and similars may be considered as Internet-based connection possibilities, but as none of these allow for a direct mailer-to-mailer connection, I don't think they should be taken as "capability flags" justifying on their own a nodelist entry. However, they could work as "extension flags" to an otherwise justified entry (be it an IP entry or not). The way this could be implemented would require some studying, namely the size limit of the flag fields may complicate things. Anyway, this can also be considered later, I agree we should focus on the simplest for the time being. WD> Have I missed something? I don't think so. :-) Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 9:20 #275 By: Ward Dossche To: Lothar Behet Re: MAKENL ! ? St: ------------------------------------------------------------------------------ Lothar (and the other who made the remark/suggestion), > But it should be possible, to put the fqdn in the system name field. I definitely agree with that! \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 9:39 #276 By: Ward Dossche To: All Re: 000 = emergency-number in Australia St: ------------------------------------------------------------------------------ I checked with Alwyn Smith, the new ZC/3, and he confirmed ... > Yes, 000 is the Australian emergency number. I was thinking of 999 as a > prefix code to use ... Comments anyone? \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 22:53 #277 By: Ward Dossche To: All Re: 000-x-y-z-w -> 000-x.y.z.w St: ------------------------------------------------------------------------------ Yow All, See the message by Eckhard Mueller (2:2449/225) that has appeared (or soon will) with the same subj. I've tested the 'dash-to-dot'-conversion-aid which he describes. It is freqqable at this node with the following names : NLIP.ARJ 184 kB complete package for OS/2, NT, W95, DOS NLIP2.ARJ 12 kB if you need a version for OS/2 only NLIPNT.ARJ 15 kB if you need a version for NT and Win95 only NLIPDOS.ARJ 19 kB if you need a version for DOS only _and_ have already DOS4GW.EXE (V1.95) If you don't have DOS4GW, request NLIP.ARJ instead. After conversion and using this converted nodelist there are true-IP-addresses in the nodelist. Of course, keep the unconverted nodelist as otherwise updating due to a changed CRC will not work. Take care, \x/ard --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 1:56 #278 By: maciej grzeszczuk To: Lech Szychowski Re: IP-access St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Sat, 27 Sep 97 01:43:00 +0200 Lech Szychowski napisal byl w fido.ip_connect: > For the time being that's AFAIK true. Which does not mean we cannot > witness some FTS-1 capable software appear soon (although I have to > admit I doubt it). what interaction do you imagine while smtp of ftp session, huh? 289 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: mocz na nasza nacje (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 2:01 #279 By: maciej grzeszczuk To: Lech Szychowski Re: IP-access St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Sat, 27 Sep 97 02:37:00 +0200 Lech Szychowski napisal byl w fido.ip_connect: > Ooops! Here we hit another problem: multiple-homed IP nodes. [...] > with such cases too. Another address, as with multiple PSTN > lines? if we don't want to solve the problem with delivering mail to the system via secondary link, when primary is down (and the mail is packed for the primary one) we should use multiple nodelist entries. because of the fact that some of the mailers are aware of this problem (fe. t-mail, when you can specify what addresses should be polled when the primary node is down). 290 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: tak, tak, wanna (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 12:15 #280 By: Lech Szychowski To: Ward Dossche Re: A bit of steering ... St: ------------------------------------------------------------------------------ > Because including a node would become rather complicated. After having > dealt with the fido-structures a future IP-node must equally have its > IP-address or similar in the tables maintained by the postmaster at > z2.fidonet.org. Suppose nodelist lists just the FQDN of the IP-system (if it list the IP address, then all problems are already solved). Then it's true changing its address/location requires both changing ist nodelist entry and its DNS entry - but, as DNS is quite independent we cannot avoid this problem no matter what we do. Well, I guess we could, if we were to say that nodelist is to be maintained by the z2.fidonet.org hostmaster and z2.fidonet.org DNS entries are to be generated automatically (some kind of script could do it without much hassle) from the nodelist entries. > This means 2 interventions where one would do when using the phone- > field. Correct. But it's IMHO inevitable. Please realize then we'd have the same problem with PSTN addresses if phone numbers were subject to changes as frequent as the dynamically assigned IP addresses. Well, let me clarify one more thing: you are not trying to imply that Z2 Fido IP-nodes should all be listed under z2.fidonet.org domain? I hope not :) > It also means that from a fido-environment we will be deciding the > behaviour of people in an internet-environment, we will have to write a > policy-document regulating the entry of IP- or likewise-addresses in the > internet-tables ... I think this is not necssary. Let me give you an example. My system has FQDN/CNAME of lech.pse.pl, IP address of 194.92.3.7, Fido name/address of 2:48/48 and PSTN address of +48-22-6286745. Suppose I have my FQDN listed in the nodelist (for the purpose of this example technical/syntactical details of this listing are irrelevant). (1) If I am to change my PSTN address (phone number), I notify my RC, he changes R48 nodelist segment and passes the changed segment up the line. One person involved, one manual action. (2) If I am to change my Fido address, see (1) (3) If I am to change my IP address, I notify the hostmaster of pse.pl domain, he changes the relevant DNS record and reloads DNS base at the primary server for this domain. One person involved, one manual action. (4) If I am to change my FQDN, I have to notify both my nodelist segment maintainer (so that he could change my nodelist entry to reflect my new FQDN) and my domain hostmaster (so that he could change the DNS records defining the FQDN-to-IP relationship). That's two persons involved, two actions. Situation (4) seems to be the only one that involves much manual work. Fortunately, it can be made a bit easier if my domain hostmaster lets me keep my old name as an alias (better yet, if he lets me get the alias before I get my FQDN listed in the nodelist I could just be using this alias as my FQDN for the purpose of the nodelist entry from the very beginning). Then no change in the nodelist would be necessary. [note to all you folks who are concerned by the "CNAME vs MX" problem: Yes, I am aware of that. But as long as there is no standard MX lookup mechanism involved, it's perfectly safe to have an alias used as well as a CNAME. And IMHO MX lookups will not be not an issue here, because "IP-capable" Fido mailers will rather look for A/CNAME records, resolving addresses, not SMTP capabilities. Please correct me if I fail to see something important here.] > 1) integrating the full-IP-address in the nodelist; > 2) integrating the domain-name in the nodelist; > 3) both. My preferences are (from best to worst): 2, 3, 1. Why? The answer is: in short, for all the reasons people had in mind when they created DNS. > Have I missed something? I think it might be a good idea to try to adopt some standard way of letting mailers know that some Fido/Internet addresses are actually addresses of the very same system, so that when an attempt to connect to one of them fails a mailer could automatically try the other ones. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 12:22 #281 By: Lech Szychowski To: Steve Woodmore Re: IP-connectivity St: ------------------------------------------------------------------------------ > Surely the bare minimum requirement for a node number is to be reachable > during ZMH?. Please define the term "reachable during ZMH". I believe that's the main reason for this little misunderstanding we seem to have here now. Specifically, whether "reachable" means "via PSTN network" or "via IP network" - and what protocole(s) you have in mind. > All I am thinking of id what if I want to reach an SMTP node with an > urgent netmail, and I don't have an internet account? IMHO SMTP is irrelevant here. SMTP-capable MTAs have been around for quite a while, there is quite a number of them and they seem to deal with all the SMTP/DNS related issues as effectively as state-of-the-art permits. SMTP/DNS approach differs from FTS/nodelist one in so many ways that an attempt to integrate them both IMHO will either fail miserably or take a long long time (and then fail as well, being too complicated and too late). > Personally I think we should not even consider SMTP/IP nodes only as a > viable option, they really do raise more problems than they solve. I concurr :) Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 12:27 #282 By: Lech Szychowski To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ >> Personally I think we should not even consider SMTP/IP nodes only as a >> viable option, they really do raise more problems than they solve. > We are not discussing the solving of problems which have not been > raised. Whenever we include this four letters (SMTP) in our discussion, we raise this problem. Although it might not be self-evident for many of us (we tend to have most of our experience rather in the field of Fido, right?), we actually are. > The true subject is 'how to embrace new transmission-technologies'. The > hidden agenda is 'how to nuke FTS-0001'. A usefull dangerous statement > for idiots who have no intention to be constructive nor to foster > advancement. Nuke? Well, I'd rather call it "extend so that it would not ignore the existence of IP network (namely one, the Internet) as a carrier media", "carrier" being a keyword. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 12:30 #283 By: Lech Szychowski To: maciej grzeszczuk Re: IP-access St: ------------------------------------------------------------------------------ >> For the time being that's AFAIK true. Which does not mean we cannot >> witness some FTS-1 capable software appear soon (although I have to >> admit I doubt it). > what interaction do you imagine while smtp of ftp session, huh? SMTP/FTP are IMHO out of question. But to say "FTS-1 == EMSI" is a bit of exaggeration, don't you agree? :) Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 12:33 #284 By: Lech Szychowski To: maciej grzeszczuk Re: IP-access St: ------------------------------------------------------------------------------ > if we don't want to solve the problem with delivering mail to the system > via secondary link, when primary is down (and the mail is packed for the I hope we do want to solve this problem. > that some of the mailers are aware of this problem (fe. t-mail, when you > can specify what addresses should be polled when the primary node is down). Yes, but you have to do it manually - which, excuse le mot, sucks when we have to think about more than 5-7 nodes; remember, to do it right one has to do it each time the nodelist is changed. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 11:38 #285 By: maciej grzeszczuk To: Ward Dossche Re: A bit of steering ... St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Fri, 26 Sep 97 22:59:07 +0200 Ward Dossche napisal byl w fido.ip_connect: > The next question is "where"? The phone-field is the obvious sollution but > then > we need to do something about MAKENL and FTS-0005. > Then flags are needed. Which-ones? > And finally, a draft-FTS needs to be written and presented to the FTSC. > Have I missed something? yes. you've missed my question. i've asked about four times, what to do with the nodes that have both, standard modem and tcp connetions? possibly solutions: a/ ,U,IP(ADR) flag, that would carry the address. (imho the best idea) b/ create another entry in the nodelist (with separate nodenumber), that would carry all the information that primary does, but have address in the phone field (we're duplicating a lot of data then) c/ put address in the system name field (as it's incompatible with existing mailers as well as a/ point, we should better use the a/ one - mailers should be modified do use the ,U,IP(ADR) flags. old ones could use some kind of subst.lst list - created locally by sysop, contained the nodenumber and proper internet address) 303 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: ciri kocha krapa... w pupe... (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 11:56 #286 By: maciej grzeszczuk To: Steve Woodmore Re: IMPORTANT! standard of protocol for ip-nodes proposal. St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Sat, 27 Sep 97 07:58:56 +0200 Steve Woodmore napisal byl w fido.ip_connect: > Surely the bare minimum requirement for a node number is to be reachable > during ZMH?. [...] > Personally I think we should not even consider SMTP/IP nodes only as a > viable option, they really do raise more problems than they solve. smtp-only nodes should forget about gaining node status. but ip node (of course, if it's on line during zmh, and have his own, separate ip number (or valid fqdn name, that points to his actual ip)) accepting emsi connetions should be treated as a fully capable node (pvt status should be added, because we should not confuse standard mailers with modem connected). but, we have to think about the standard of connetion that every ip node allows. my proposal is to set VMODEM as a standard protocol, which must be handled by every ip node (some kind of fts-0001 for ip nodes). vmodem exist on every hardware or software platform (seen working with amiga, pc running os/2, win95, win3.1, linux, freebsd, openbsd, and all unices running on the 'big' computers). also ifcico from ifmail packet also can easily handle both incoming and outgoing connections. all tcp fossils are vmodem aware. we shouldn't allow binkd as a standard of ip node (ip system running only binkd, not vmodem, should not be allowed to gain node status), as it's compatible only with itself. 304 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: radykal = wesola kupa (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 12:00 #287 By: maciej grzeszczuk To: Pedro Lima Re: IP-access St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Fri, 26 Sep 97 10:08:19 +0200 Pedro Lima napisal byl w fido.ip_connect: > mg> separate aka just for duplicate all information, with one differs - the > mg> phone number field - is a bad idea... > Naturally, all the capability flags are different as well. what is the more importand, the node number differs as well. what for? and what we're duplicating then, is the sysop's name (usually the longer field), system's name, city, some of the flags... what for? 305 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: graba, jestes zajebisty. dziekuje. dam ci wina. (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 27 Sep 97 18:49 #288 By: Eric Vaneberck To: All Re: FAQ about Fido Via Internet St: ------------------------------------------------------------------------------ The autor of this FAQ is Alex Konshin 2:5030/217 FIDO via Internet. FAQ (lang:english) FIDO via Internet. FAQ (lang:english) + | This is just a first attempt to write a Fido over IP FAQ that summarizes | experience of Fidonet sysops from Regions 45, 46, 47, 50 and 51, Zone 2. | Warning: some links from this page are pointing overseas and may be | really slow. - Contents: * BinkD * General info * Binkd-related WWW links * Miscellaneous software for your operating system: * OS/2 * *NIX * DOS * Windows 3.x * Windows'95 * Windows NT * Vmodem under OS/2 troubleshooting * Space delimiter in nodelist number field * Where can I get nodelist containing ip addresses of Fidonet nodes? * IP-list flags used in Region 50 * fidonet.net domain * Credits, etc... - BinkD (OS2/Win32/*NIX): General info Binkd 0.9.1 was released July 10. What's new? * What is binkd? BinkD is a new BSO (Binkley Style Outbound) mailer that works with TCP/IP o only. It's protocol (called BINKP) is already supported by Argus. BinkD is available with sources and is designed keeping in mind portability between Windows'95/WinNT/OS2/Unix platforms. It is very simple in setup, extremely small in size and provides probably the best throughput over TCP/IP links. * Ok, sounds good! Where can I get it? Try this: ftp.prospect.com.ru, login:binkd, ftp.falcon.spb.su. These two sites will probably contain the latest versions. You can also try my mirror site in Canada , but it might be not up to date. Other binkd-related sites: * Binkd home page by Dima Maloff: http://www.magadan.su/~maloff/binkd/ * Binkd User Reference: http://www.doe.carleton.ca/~nsoveiko/fido/binkd/man/binkd.man.html * Alex Konshin's binkd page: http://www.falcon.spb.su/fido.BinkD.html * List of binkp-capable nodes by Ruslan Hakurate: http://www.aha.ru/~hadja/binkd.txt * Argus home page (binkp-compatible mailer for Win95/WinNT): http://www.ritlabs.com/argus/ - Miscellaneous software for your operating system: * OS/2 : Basically, you have two options: Vmodem and BinkD. * Vmodem.exe is a part of the SIO communication drivers distribution. Read vmodem.txt carefully, it is a comprehensive manual. * Brief outline: put something like this to your config.sys: DEVICE=C:\OS2\SIO16\SIO.SYS (COM2) (COM3,INTERNET,NONE) (COM4,INTERNET,NONE) it tells SIO to create 2 virtual COM ports that will be available via vmodem. Run vmodem.exe. Now you can access these ports from any of your communication software. See Vmodem troubleshooting section for further problems. * *NIX : You can install ifmail together with the patch ftp://ftp.card.ru/pub/sts/iftel3.tgz that allows it's interoperability with Vmodem. This patch is for ifmail v.2.8 and is said to cause problems when applied to other versions. Alternatively, you can install BinkD which is available with more or less portable sources and binaries exist for many popular platforms. Both ifcico (ifmail's mailer part) and binkd can share the same Binkley style outbound and run concurrently on the same machine/node. * DOS : There's rlfossil - a quick'n'dirty utility that emulates FOSSIL interface for telnet/rlogin connections. * Windows 3.x : comt.zip - I haven't tried this myself, but for sure it's for native Win applications only. In other words, it's useless for DOS mailers running in a DOS box. * Windows 95 : * Your first option is to install BinkD. * Second option (DOS boxes only): you can run rlfossil using a second network card and packet driver or ndis3pkt.zip. If you're running ndis3pkt, make sure that you assign a unique ip address for every DOS session using it's interface. * [from Sergey Astashonok, 2:450/58]: Combination of COM/IP v1.03b with T-Mail/NT seems to be working. Use AT &D2 S1009=3142 s1008=3142 s1001=1 s1002=1 s1004=0 s1005=0| init string to connect to Vmodem. DTR drop can't force it to hangup, so use ~~~+++~~~ATH0| instead of DTR drop. * Windows NT : * Your first choice seems to be BinkD. * There's wntfossil 1.0 beta3. (the latest version should be available at http://www.woodruff-software.com), that is said to allow using telnet connections, but so far they succeeded only in running a BBS through it. This thing runs on NT 4.0 Server only. How to make it work with T-Mail (Nikolay Buhalov, 2:50/352.5, coadmin@finet.rosmail.com): * Install Woodruff fossil for NT v1.0 for the same port as your IP connection. * Configure it for TELNET (go Control Panel -> WinFossil -> change COMx to TELNET) * Configure your T-Mail to use TELNET instead of port number x * There's Argus - a native Windows'95/NT mailer with GUI, that also supports TCP/IP transport. - Vmodem troubleshooting: * regular SIO supports up to 4 ports only, but there's a 16-port (commercial?) version also available; * vmodem is not capable of storing it's configuration (i.e. no ATZ command); * mailer says that it can't initialise vmodem. Run mode com3,dtr=off,idsr=off,odsr=off before starting your mailer. * refuses to work without any visible reasons. That happens fairly often - up to several times a day. The reasons for such behaviour are unknown, but a suspicious number of dead sockets (netstat -s) accumulates. Only one workaround is known so far: correctly close all the running instances of your mailer, issue the aforementioned mode command (just in case...), close vmodem.exe by pressing Ctrl-C ... and pray for the following not to happen: * vmodem doesn't start saying that the port is busy. That usually happens while trying to restart vmodem. Looks like the port keeps some remains of the previous ip connect (netstat doesn't show any open sockets, though). No workaround is known. Try this: do wait for a few minutes; try to start vmodem; while vmodem not running. * vmodem supports two protocols: telnet and proprietary VMP. Configure vmodem to use another port for inbound connections in order to be able to run standard telnetd (\TCPIP\ETC\SERVICES for IBM TCP/IP or \MPTN\ETC\SERVICES for IBM MPTN): VMOTelnet 60177 - Space delimiter in nodelist number field * You'll need to specify port number after ip address (separated by space) in order to connect to nonstandard ports. That can be a problem, because the ways to include a space in the nodelist phone number field are not trivial. * T-Mail has an undocumented feature that substitutes underbars in phone number for spaces. The following is from a sample subst.lst file: 2:5020/400 Phone 194*87*1*79_60177 * for BinkleyTerm and Xenia you should escape space with a backslash to get the same result: ,118,Comp,Moscow_Russia,Alexey_Kshnyakin,158*250*32*40\ 60177 60177,38400,XA,MO,CM,TCPIP,IFC - Where can I get nodelist with IP addresses of nodes? * There's ip-list for Region 50 of Zone 2, maintained by Basil Dolmatov (2:5020/140). It is distributed via r50-list fileecho and is available from many nodes with ip connectivity, e.g.: http://www.falcon.spb.su/fido.vmodem.html * Ruslan Hakurate, 2:5020/1001, maintains a list of BinkP compatible nodes which can be conveniently included into BinkD config. * Alex Konshin, 2:5030/217, the original author of this FAQ, maintains am online FIDO-via-IP query/registration form. * see also section on fidonet.net domain - ip-list flags: * VMP - proprietary vmodem protocol, supported by vmodem only ('coz the author didn't provide any protocol description for porting to another platforms and implementing in compatible software). By default it answers at port 3141. If you type atdt#194*58*242*1 to vmodem, it'll connect to my system using VMP (# means connect to port 3141 using VMP). Note, that dots in IP address are substituted for asterisks. * TEL - telnet protocol. It's supported by almost all of the software, but ifmail requires patches to do it correctly. Vmodem uses it if no # prefix is specified and by default tries to establish connection with port 23. You'll need to specify port number after ip address (separated by space, e.g. atdt#194*87*1*79 60177) in order to connect to nonstandard ports. That can be a problem, because the ways to include a space in the nodelist phone number field are not trivial (see above). This example shows a default vmodem-compatible telnet port for ifmail, i.e. if a node has IFC flag, it typically accepts compatible telnet connections to port 60177. * IFC - for ifmail nodes. Proprietary telnet daemon answers connections to port 60179 (by default), and vmodem-compatible takes care of port 60177. * BND - for BinkP compatible nodes. Deafult port is 24554. So far is supported by BinkD only. * TCP or TCPIP - for any of the above. Indicates general TCP/IP capability. - fidonet.net domain * What is fidonet.net? fidonet.net is an Internet domain registered to facilitate development and deployment of fido-over-internet technology of all kinds and flavours. Here is a recent (as of 22/09/97) whois info on it: Domain Name: FIDONET.NET Administrative Contact: Ustinov Fyodor (UF7-ORG) ufm@PROSPECT.COM.RU 7 095 564 8272 Fax- 7 095 564 8945 Technical Contact, Zone Contact: Valdov, Dmitry (DV15-RIPE) dv@KIS.RU +7 831 2342022 +7 831 2343420 Billing Contact: Ustinov Fyodor (UF7-ORG) ufm@PROSPECT.COM.RU 7 095 564 8272 Fax- 7 095 564 8945 Domain servers in listed order: NS.FIDONET.NET 195.98.32.193 NS1.FIDONET.NET 194.85.143.33 * How fidonet.net subdomain and host names are assigned? Subdomain and host names are formed in the same way as for the fidonet.org domain. For example, if your 4D Fidonet address is A:BBB/CCC[.DDD], then your domain address would be [pDDD.]fCCC.nBBB.zA.fidonet.net. Such unique address mapping allows one to obtain a node's IP address knowing only it's FTN address. Binkd is expected to support such address mapping directly starting from the next version (0.9.2). * What is the difference between fidonet.net and fidonet.org than? Fidonet.org domain was registered with the purpose of exchanging mail between Internet RFC822-style and Fidonet mail systems. In other words, hosts in the fidonet.org domains are Fidonet-Internet gateways that are supposed to accept mail by both SMTP and FTS-1 (or, better EMSI) protocols. There is no such restriction on hosts in fidonet.net domains (see below). * What do I do to get a fidonet.net hostname? * Your system should have binkp capability. This protocol is currently supported by binkd and Argus mailers. * You should contact a person who is responsible for DNS in your Fidonet network. You can use nslookup to figure out if your network already has a DNS server. If there is none, please contact Fyodor Ustinov, 2:5020/79, ufm@prospect.com.ru to arrange a DNS record for you. * I am using Vmodem. Why should I install something else in order to be listed in fidonet.net? Common sense is that Fidonet systems should be able to exchange mail directly. In order to make this possible, they should support some protocol which will be the 'least common denominator' for a given transport. FTS-1 is such a protocol for conventional dial-up transport. Vmodem is not a candidate for the 'least common denominator' because * it exists for OS/2 only * it's protocol is proprietary * it merely adapts dial-up technology for TCP/IP transport retaining many drawbacks of the former and without taking full advantage of the latter. On the contrary, binkp is * an open specification protocol * designed to operate on TCP/IP networks * has a portable implementation that is distributed in sources under terms of GNU public license. So, it seems to be a natural choice for the 'least common denominator' protocol. - Credits, etc... Fido, FidoNet and the dog-with-diskette are U.S. registered trademarks of Tom Jennings. Dog with diskette picture that appears on the top of this page was found on Lee Kindness's WWW Fidonet Resource pages and (finally) is used with his kind permission ;). - Send your questions and/or answers to: + | Alex Konshin (the original compiler of | Nick Soveiko (maintainer of | this FAQ) | the English FAQ + | alex@falcon.spb.su | nsoveiko@doe.carleton.ca | Fido: | Fido: | 2:5030/217@fidonet | 2:5030/217.57@fidonet | (fido.falcon.spb.su,VMP,TEL,BND) | 2:5030/69.101@fidonet | | 2:5020/118.52@fidonet Last : September 22, 1997. --- Terminate 5.0+ * Origin: Terminate registration site 32-10-813088 (2:293/2009) SEEN-BY: 201/329 212/8 235/100 254/60 270/101 280/100 291/709 1928 SEEN-BY: 292/812 818 854 857 862 865 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 9:43 #289 By: Steve Woodmore To: Lech Szychowski Re: IP-connectivity St: ------------------------------------------------------------------------------ * Replying to a message in : SAVEAREA Hi Lech Szychowski, hope you are having a nice day 27-Sep-97 03:31:00, Lech Szychowski wrote to Steve Woodmore Subject: IP-connectivity LS> And it means that (in both Maciek's and my opinions) this LS> connectivity alone is not enough to make you qualify as an LS> IP-node. Of course, it does not change your existing PSTN-based LS> node status. So you are just concentrating on IP only nodes then?, and forgetting all the other ways that fidonet mail can be shifted over the internet. -=> Steve Woodmore <=- --- Terminate 5.00/Pro * Origin: Moved to London (2:254/410.1) SEEN-BY: 2/200 201/329 212/8 235/100 251/25 253/417 254/60 61 62 261 SEEN-BY: 254/410 620 280/100 292/854 857 876 335/533 362/21 380/106 SEEN-BY: 440/410 442/403 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 9:44 #290 By: Steve Woodmore To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ * Replying to a message in : SAVEAREA Hi Ward Dossche, hope you are having a nice day 27-Sep-97 22:04:40, Ward Dossche wrote to Steve Woodmore Subject: Re: IP-connectivity WD> We are not discussing the solving of problems which have not been WD> raised. The true subject is 'how to embrace new WD> transmission-technologies'. The hidden agenda is 'how to nuke I don't have any hidden agenda, all I want is a way for people using the nodelist to be able to see if there is someone who will supply their Echomail over the internet. Taking the long view, I can why their is resistance to this, once we have established this, then why bother with geonets?, someone half way round the world is only a local call away then. In fact why even Bother with regions and Zones, why not just have one large net. -=> Steve Woodmore <=- --- Terminate 5.00/Pro * Origin: Leaving NET440 :( (2:254/410.1) SEEN-BY: 2/200 201/329 212/8 235/100 251/25 253/417 254/60 61 62 261 SEEN-BY: 254/410 620 280/100 292/854 857 876 335/533 362/21 380/106 SEEN-BY: 440/410 442/403 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 18:56 #291 By: Pedro Lima To: Lech Szychowski Re: IP-connectivity St: ------------------------------------------------------------------------------ Hi, LS> we just have to either write an amendment LS> to P4 or... (I wouldn't dare mentioning the alternative option). If we go by the book, then P4 doesn't allow for ammendments, not even when considering the possibility of local policies (the closest possible thing will be such an "ammendment" being approved as zonal policies in all zones). This leaves us with no option but to replace it if we want to change something in it. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 18:57 #292 By: Pedro Lima To: Lech Szychowski Re: IP-access St: ------------------------------------------------------------------------------ Hi, LS> You are right. But no DNS record should be cached for a period of time LS> longer than its TTL says. So I understand. It's possible, then. LS> Right again - and again a "but" :) But the problem is so far almost no LS> DNS servers support triggered update scheme; classical DNS servers LS> use polling approach. Actually, I don't see it as a serious problem. Such nodes would be "up" at a distinct time of day (declared on their Txx flags?), probably during an hour or two, and given a short TTL, the records would then probably be already updated by the time someone tries to connect, but as my knowledge of DNS and bind isn't that strong, maybe I'm missing something? Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 6:01 #293 By: Pedro Lima To: Ward Dossche Re: A bit of steering ... St: ------------------------------------------------------------------------------ Hi, Sorry to quote myself... PL> I'd suggest a "dual flag" like ITN,60177 (...) PL> to prevent calls to IP-nodes on non-IP mailers. Unless a specific flag is used to denote IP nodes, such as "IP" (preferably to "TCP", IMHO). Also, new IP protocols may surface, and like this, one can maintain his mailer configuration. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 6:06 #294 By: Pedro Lima To: Lech Szychowski Re: IP-access St: ------------------------------------------------------------------------------ Hello, LS> Another address, as with multiple PSTN lines? Until a new nodelist format accepting several entries for the same node gets to be the FidoNet standard, I think that's the logical choice. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 8:30 #295 By: Pedro Lima To: Lech Szychowski Re: IP-connectivity St: ------------------------------------------------------------------------------ Hi, LS> if system operator really wants to be a node he can very well LS> become a traditional technology based one. The problem with that approach is that whoever wants to connect to such a node would then be subject to the telco charges, and it's probably cheaper to do it via Internet. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 18:58 #296 By: Pedro Lima To: Lech Szychowski Re: IP-connectivity St: ------------------------------------------------------------------------------ LS> system not being able to support fixed time-and-number LS> reachability. And that's exactly the situation we have now with dial-up LS> dynamically IP/FQDN assigned nodes. No, I could be able to be in the Internet every day at the same hour using my dynamic IP account. The problem is how would someone know where to connect, which brings us to the subject of dynamic DNS. :-) Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 7:19 #297 By: Pedro Lima To: Steve Woodmore Re: IP-connectivity St: ------------------------------------------------------------------------------ SW> The IP address is mine, and you can send email to it at any time I thought an e-mail in the form user@[aaa.bbb.ccc.ddd] would be directed to port 25 of the machine at aaa.bbb.ccc.ddd, but maybe it isn't as simple as that. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 7:16 #298 By: Pedro Lima To: Lech Szychowski Re: IP-connectivity St: ------------------------------------------------------------------------------ LS> IP address? No. There are MX records, but I have yet to see MTA LS> trying to query DNS for an MX record of an *.in-addr.arpa. entity :) What happens then to an e-mail in the form user@[aaa.bbb.ccc.ddd]? Reverse lookup the name of aaa.bbb.ccc.ddd, lookup its MX record and the mail sent there? Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 7:22 #299 By: Pedro Lima To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ WD> Changing FTS-0001 would be fine ... but someone has a copyright on it WD> and changing the text is forbidden via that same copyright-statement. Is the reference 'FTS-0001' copyrighted? Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 19:00 #300 By: Pedro Lima To: Lothar Behet Re: Proposal? St: ------------------------------------------------------------------------------ Hello, LB> dns-entries in the phone field would prevent "normal nodes" from LB> dialing incompatible systems. The problem of implementing it is only MakeNL. LB> Another thing is a system, which supports direct dialin (i.e. ppp), LB> which needs a real phonenumber and flags. That's an idea I've playing around with, but I didn't want to further confuse the issues at hand. If that dialin connection allows for Internet connectivity (the system would act like a free ISP), then a real IP address must be used for the remote (unless using something like IP-Masquerading?) and a protocol allowing fully automatized configuration of the remote would be required, as I don't see how to give the configuration information in the nodelist. Most likely that dialin connection, although using IP-based protocols, doesn't grant access to the Internet, and the "private network" IP addresses may be used to minimize problems. If the negotiation protocol doesn't allow for automatic configuration of the remote, then we'd have to previously adopt a configuration standard for it. LB> But additional flags for dialup lines should be considered at this LB> moment, because they will be the next step. They may be marked as user LB> flags in the form of: ...,U,PPP,... ...,U,SLIP,... Does SLIP allow for automatic configuration of the remote? Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 12:12 #301 By: Ger Vloothuis To: Pedro Lima Re: Test of an IP-number in the nodelist / Proposal St: ------------------------------------------------------------------------------ Hi Pedro, Pedro Lima wrote in a message to Ger Vloothuis: GV> Triple 0 is the emergency phone number in Australia. Something like GV> 911 in the USA and 112 in GMS networks. Don't know if this would be a GV> problem PL> What's your international prefix? If '0' or '00', then indeed a PL> problem may occur if no specific translation (e.g. translate to PL> one's own phone number) or flag filtering is configured. The prefix is +61-. There would not be any country with an '0' or an '00' prefix because this is the ITU recommended international access code. If there is no translation specified in the nodelist compiler, then the mailer would dial the number as listed, wouldn't it? Regards Ger --- WtrGate v0.93.p1 Unreg * Origin: Teletechnique Melbourne (3:633/284) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 270/101 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2410/200 2432/0 200 321 700 2433/0 225 1200 1201 1204 SEEN-BY: 2433/1800 2448/10 6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 12:16 #302 By: Ger Vloothuis To: Ward Dossche Re: How to not get IP-implemented St: ------------------------------------------------------------------------------ Hi Ward, Ward Dossche wrote in a message to All: [....] WD> It may end up with Everybody blaming Somebody when Nobody does what WD> Anybody could have done. Maybe, some day, wheather and other unpredictables permitting, 'Somebody' will come up with the right idea. I don't think it will be just 'Anybody'. But bouncing of 'Everybody's' thoughts does 'Nobody' any harm. WD> Sound familiar to anyone? Yes, that is the way things work in most places. It actually does work most of the time :-) Regards Ger --- WtrGate v0.93.p1 Unreg * Origin: Teletechnique Melbourne (3:633/284) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 270/101 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2410/200 2432/0 200 321 700 2433/0 225 1200 1201 1204 SEEN-BY: 2433/1800 2448/10 6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 12:29 #303 By: Ger Vloothuis To: Denis McMahon Re: IP-access St: ------------------------------------------------------------------------------ Hi Denis, Denis McMahon wrote in a message to Pablo Saratxaga: [....] DM> time, the only thing that's missing is it's integration into the DM> nodelist. If you look at nodelist integration, there should be a mechanism which can recognise either IP address or domain address. So nobody would be dependant on DNS if there was an IP address only, but if a domain address was registered it would make life easier. Coming to think of it, why not keep the existing V7 Nodelist unchanged, and make the mailer try to resolve the IP number from the default Fido domain address: "p#.f#.n#.z#.fidonet.org" and if the result comes up with a valid IP number then an FTP or email transfer mechanism can start. If the IP number comes up as a CNAME only, then the mailer falls back from FTP transfer to Mailtunnel or Transx. Both can have 'well known' email user addresses at the remote domain and probe messages can be sent out to establish the best connectivity methods. If there is no IP number at all, then the mailer falls back to the normal phone line method. The user flags in the nodelist could indicate what type of FTP or Email transfers are accepted. No need to change anything in the nodelist really. Maybe just add a user flag UIP[xxx.xxx.xxx.xxx] to indicate FTP capability or UMX[xxx.xxx.xxx.xxx] to indicate mailtunnel/transx capability for those who need it. Regards Ger --- WtrGate v0.93.p1 Unreg * Origin: Teletechnique Melbourne (3:633/284) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 270/101 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2410/200 2432/0 200 321 700 2433/0 225 1200 1201 1204 SEEN-BY: 2433/1800 2448/10 6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 16:12 #304 By: David Moufarrege To: Ward Dossche Re: How to not get IP-implemented St: ------------------------------------------------------------------------------ Hello Ward! In a message Ward Dossche wrote to All: WD> Sound familiar to anyone? Sure, I have used the same text for other topics -=David=- _____________________________ | e-mail : david@kraut.xg.com | | FidoNet: 1:2613/404 | | 1:13/0 | ----------------------------- ... You can't step twice in the same river. --- * Origin: Kraut Haus * Rochester, NY * 716-359-0871 (1:2613/404) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 270/101 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2410/200 2432/0 200 321 700 2433/0 225 1200 1201 1204 SEEN-BY: 2433/1800 2448/10 6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 16:14 #305 By: David Moufarrege To: Ward Dossche Re: IP-nodes ... St: ------------------------------------------------------------------------------ Hello Ward! In a message Ward Dossche wrote to All: WD> At midnight the following will be added to the nodelist ... Which nodelist? -=David=- _____________________________ | e-mail : david@kraut.xg.com | | FidoNet: 1:2613/404 | | 1:13/0 | ----------------------------- ... Youngsters remember anything if it happened or not. --- * Origin: Kraut Haus * Rochester, NY * 716-359-0871 (1:2613/404) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 270/101 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2410/200 2432/0 200 321 700 2433/0 225 1200 1201 1204 SEEN-BY: 2433/1800 2448/10 6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 29 Sep 97 8:31 #306 By: Mats Wallin To: Lech Szychowski Re: IP-connectivity St: ------------------------------------------------------------------------------ LS> C'mon, all decent mail editors support aliases. Sure, but why should I have to define a lot of aliases when I instead can use the nodelist for this information? Mats mw@defsol.se --- * Origin: Definite Solutions, Stockholm, Sweden (2:201/329.11) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/27 212/8 235/100 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 9:28 #307 By: Rune Johansen To: Lech Szychowski Re: Proposal? St: ------------------------------------------------------------------------------ >> - ICQ (paging/transfer system) > Is ICQ worth thinking about? Is it at least available on more than > one platform - or will it be soon? I doubt. Just trying to point out that we should make this undependant on the transfer methods that we know now, to allow for other methods later. --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 222/0 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 9:36 #308 By: Rune Johansen To: Ward Dossche Re: 000 = emergency-number in Australia St: ------------------------------------------------------------------------------ >> Yes, 000 is the Australian emergency number. I was thinking of 999 as a >> prefix code to use ... > Comments anyone? Yes: In the nodelist, if you run across a number that you do not want to add a dialling prefix to (local number) you strip it. As the 000- is NOT a local number that you explicit want to not add a dialling prefix to, there should be no problem. If this number is to make problems for the people down under, it would have to be noted as 61-000 to make Australian nodes trying to dial them. Or, maybe even 61-7-000 to make them dial it within area code 7. So, within the nodelist format of a phone number, 000- is NOT a national number --000 is a national number. --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 222/0 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 9:39 #309 By: Rune Johansen To: Lech Szychowski Re: IP-connectivity St: ------------------------------------------------------------------------------ > What about using a very simple approach: if you don't feed any mail > to anyone (ie you are a "leaf" in a routing/distribution tree) then > there is (almost) no reason for you to be a node. Please keep in > mind that most of the dial-up connected systems would do fine as > points (nodelist problems saved herewith) of a boss that is a leased > line / permanent connection system. True. If you are a node, with a PSTN/ISDN node number, and you have the possibility to call an ISP to fetch mail from your boss via IP, you don't need to be listed as IP capable. Your boss has to. If someone want to crash you a mail from an IP-only node, you cannot. You'll just hav to route it via the nearest IP capabable node, or your own IP connections. --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 222/0 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 29 Sep 97 1:32 #310 By: Lech Szychowski To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ >> What about using a very simple approach: if you don't feed any mail >> to anyone (ie you are a "leaf" in a routing/distribution tree) then >> there is (almost) no reason for you to be a node. > Fine .... then the leaf-nodes dissappear ... Nope. They do not disappear. They became points,... > which makes their former feeds leaf-nodes ... ...and they still get their feed from their bosses,... > for which reason they need to dissappear etc... ...which means these bosses have no reason to disappear. Maybe I was a bit unclear when I wrote about "leaf in a distribution tree"; what I had in mind was fromn the _technical_ point of view those leaf nodes (the ones that feed mail to no one, including points) might as well be points. Of course, they might choose to apply for the IP-node status - but then I suggest we should adopt some mimimal set of conditions they have to satisfy to be granted this status, reachability and connectivity being main issue here. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 29 Sep 97 1:36 #311 By: Lech Szychowski To: Ward Dossche Re: IP-connectivity St: ------------------------------------------------------------------------------ >> advances even faster and we just have to either write an amendment >> to P4 or... (I wouldn't dare mentioning the alternative option). > Changing FTS-0001 would be fine ... but someone has a copyright on it > and changing the text is forbidden via that same copyright-statement. It's not necessary to change FTS-1. We may just create a new document, titled say "Non-PSTN Fidonet nodes". There we can list requirements, suggestions etc. FTS-1 can stay (as it is now) the main document for PSTN nodes, IP nodes having their own counterpart. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 29 Sep 97 9:15 #312 By: rafal wiosna To: Lothar Behet Re: IP-connectivity St: ------------------------------------------------------------------------------ þ (RAF, Tue Sep 23 1997) Lothar Behet -> rafal wiosna: rw>> If that's true no IP-only node should be allowed. They can't be rw>> reached during zone's ZMH. LB> Why not? May be not all, but some are even running CM. Because they cannot be reached from the PSTN Fido and that's against P4. But don't get me wrong -- I'm not against it, I'm just poiting out that today's P4 and other regulations are not "compatible" with the whole IP-node idea. - Rafa’ "WXR" WiOS/2na. --- GoldED/2 2.51.A1026+ 30PL2 * Origin: Pho c0? (2:480/33) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 29 Sep 97 1:33 #313 By: maciej grzeszczuk To: Lech Szychowski Re: IP-access St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Sun, 28 Sep 97 11:30:00 +0200 Lech Szychowski napisal byl w fido.ip_connect: > > what interaction do you imagine while smtp of ftp session, huh? > SMTP/FTP are IMHO out of question. But to say "FTS-1 == EMSI" is > a bit of exaggeration, don't you agree? :) emsi (vmodem) would be standard for ip-nodes. as the base protocol. that would simplify many things... 309 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: ps. polewiak pewnie jest juz w lodzi... (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 29 Sep 97 1:36 #314 By: maciej grzeszczuk To: Lech Szychowski Re: IP-access St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Sun, 28 Sep 97 11:33:00 +0200 Lech Szychowski napisal byl w fido.ip_connect: > I hope we do want to solve this problem. but it shouldn't be our primary case now.. > Yes, but you have to do it manually - which, excuse le mot, sucks when > we have to think about more than 5-7 nodes; remember, to do it right > one has to do it each time the nodelist is changed. sure. but for now, we haven't got better idea. we don't have better idea to solve the same problem with standard modem connected nodes with multiple lines (like 2:480/19 and 2:480/89). 310 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: linia lotnicza: las vegas - las palmas - las kaba (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 29 Sep 97 12:36 #315 By: maciej grzeszczuk To: Ward Dossche Re: 000 = emergency-number in Australia St: ------------------------------------------------------------------------------ From: maciej grzeszczuk Sun, 28 Sep 97 08:39:39 +0200 Ward Dossche napisal byl w fido.ip_connect: > > Yes, 000 is the Australian emergency number. I was thinking of 999 as a > > prefix code to use ... > Comments anyone? 999 is an emergency number in poland... 318 -- = wasza KrAp = krap@psych.uw.edu.pl = http://www.psych.uw.edu.pl/~krap = = phone 602-339173 = PGP 50D98803B12327E7 216A787AB7EFD5FA * in arp we trust * --- ifmail v.2.9 * Origin: znam ludzkie cialo od A do D! (2:480/70) SEEN-BY: 48/70 201/329 212/8 235/100 254/60 280/100 292/854 857 876 SEEN-BY: 335/533 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 SEEN-BY: 481/21 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sun 28 Sep 97 12:05 #316 By: Lothar Behet To: Lech Szychowski Re: IP-access St: ------------------------------------------------------------------------------ Hello Lech! On 27 Sep 97 wrote Lech Szychowski to maciej grzeszczuk: LS> Ooops! Here we hit another problem: multiple-homed IP nodes. LS> And it's not a theoretical one: 2:480/17 aka 2:48/48 (R48 Inet gate) LS> can be reached at three different IP addreses, each of them in LS> a different C class, each of them differently routed. LS> Looks like we have to make some arrangements to be able to cope LS> with such cases too. Another address, as with multiple PSTN LS> lines? How do you handle this on FTN level? I think, if you are not regularly visiting your several sites, you will install a forward schedule to the (temporary) main site. Wether you handle this forwarding on ip or ftn base, doesn't matter, but it is possible. This is not a problem of ip, but a problem of your system setup independant of the used connection technology. Regards Lothar --- GoldED/2 3.00.Alpha5+ * Origin: Spelling chekker four sail. Wurks grate on (2:2446/301) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2410/200 2432/0 200 321 700 2442/2000 2446/0 300 301 304 SEEN-BY: 2446/305 2448/6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 29 Sep 97 8:32 #317 By: Lothar Behet To: Ward Dossche Re: 000 = emergency-number in Australia St: ------------------------------------------------------------------------------ Hello Ward! On 28 Sep 97 wrote Ward Dossche to All: >> Yes, 000 is the Australian emergency number. I was thinking of 999 >> as a prefix code to use ... WD> Comments anyone? IMHO any prefix we select at this moment may be used later as an official code or already be in use in at least one country. Regards Lothar --- GoldED/2 3.00.Alpha5+ * Origin: Life: a bad game, but graphics is really good (2:2446/301) SEEN-BY: 24/905 906 201/329 212/8 235/100 240/5490 5701 244/1 254/60 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 465/50 480/33 SEEN-BY: 2410/200 2432/0 200 321 700 2442/2000 2446/0 300 301 304 SEEN-BY: 2446/305 2448/6000 2471/1020 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Mon 29 Sep 97 6:48 #318 By: Pedro Lima To: maciej grzeszczuk Re: IP-access St: ------------------------------------------------------------------------------ mg> what is the more importand, the node number differs as well. what for? We're aiming at accomodating IP (and other network technologies) nodes in the nodelist with the maximum of versatility and simplicity and minimally upsetting the existant standards, namely by making it compatible with the existant mailers and their nodelist compilers. By your suggestion, it would be almost certain that the length limit of the flag fields is exceeded, which is bound to create problems. Regards, Pedro --- * Origin: Kaos BBS * +351-1-8862878 * 24h (2:362/21) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 361/7 362/21 24 47 380/106 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 23:38 #319 By: Rune Johansen To: Lothar Behet Re: Proposal? St: ------------------------------------------------------------------------------ > The first part seems quite ok, except I would add Transx to the protocols. Hmm, yes, you are right. >> For DIRECT connections this would be a network address (IP number, >> DNS address, X25 id, ICQ number, whatever). For INDIRECT connections >> it would be the receivers address (email address, FTP server IP/DNS, >> whatever). > I was not able to call by example a dns-entry with another (analogue or > ISDN) > line, could anyone else manage it? I tried a few. On one, my modem just answered "3366" without dialling at all. On a couple of other, it started to auto-answer itself, without dialling :-)) >> Example for DIRECT connection: >> ...,SysOp,bbs.some.net,9600,CM,U,IP,ITEL,VMO >> ...,SysOp,12.34.56.78,9600,CM,U,IP,ITEL:1024,VMO,IFC,BNK:234 > ISDN-lines use a speed of 300, if they are not capable for analogue > connects. > The same applies with IP-connections we are talking about. I just put in a speed there, to make it "conformant" to the nodelist format, so people would see where it might be placed. >Another thing is a system, which supports direct dialin (i.e. ppp), which >need > a real phonenumber and flags. Fido is at least a network of mailbox sysops. > Should this information maintained by the the nodelist, too? No. direct dialin has nothing to do with fidonet, the fidonet techologies or the messaging scheme in fidonet. PPP is a protocol that encapsulates other network protocols like IP, IPX, Appletalk, Banyan Vines etc., so they can be used by the respective protocol stacks. If you look at it from a IP protocol stack's view, the PPP is a network card interface that layer 3 of the ISO stack can use. This service is usually supplied by the operating system, so your legacy IP stack can use it. The fidonet techologies would then use the IP stack via the (usually) 4. level (TCP) to make a reliable connection to the other end. Some applications create their own, private IP stack, but thats due to the non-existance of a stack in the rest of the OS. >IP-access for routing purposes is only one aspect, but IMHO at this moment >the > most important. > But additional flags for dialup lines should be considered at this moment, > because they will be the next step. > They may be marked as user flags in the form of: > ...,U,PPP,... > ...,U,SLIP,... You do have a point about being flexible in the way the flags can be used. Almost the same way that I want to have options for making ICQ a usable transport protocol. But, PPP and SLIP is at a much lower level in the protocol stacks, lower than we have ever encountered in the network transportation debate that we are into. --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 222/0 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 23:50 #320 By: Rune Johansen To: Pedro Lima Re: A bit of steering ... St: ------------------------------------------------------------------------------ > If I understood correctly, these are the current protocols which allow > for a mailer-to-mailer connection. Non-standard ports may be listed > after the flag. I'd suggest a "dual flag" like ITN,60177 instead of > ITN:60177 or ITN60177 because I suspect current mailers determine flags > by looking at what's separated with commas, and it would then be > theoretically simple to prevent calls to IP-nodes on non-IP mailers. That's why a IP flag should be there too. If you separate the port numbers from the protocol, you cannot be sure who is connected to who. It is easier to parse this: "U,IP,ITN:60177,VMP,IFC:2013" than this: "U,ITN,60177,IFC,2013,VMP". When mailers see "IP", they can steer clear of dialling the node directly, instead of having to recognise all the new protocols that would be added when they emerge. > E-mail, TRANX, ICQ, ftp and similars may be considered as Internet-based > connection possibilities, but as none of these allow for a direct > mailer-to-mailer connection, True. The DIRECT mailer-to-mailer capabilities are the ones that first needs to be included in the nodelist. Then the indirect ones can be involved. --- BBBS/2 v3.42 ToMmIk-6v * Origin: BarCode BBS - now with ISDN at 47-67061044 (2:210/20) SEEN-BY: 20/11 201/300 307 329 330 203/614 210/20 27 212/8 222/0 235/100 SEEN-BY: 254/60 280/100 292/854 857 876 335/533 362/21 380/106 465/50 SEEN-BY: 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 1:21 #322 By: Lech Szychowski To: Pedro Lima Re: IP-access St: ------------------------------------------------------------------------------ > LS> Another address, as with multiple PSTN lines? > Until a new nodelist format accepting several entries for the same node > gets to be the FidoNet standard, I think that's the logical choice. Given this conditions (old format) that's the only choice :) Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 1:24 #323 By: Lech Szychowski To: Pedro Lima Re: IP-connectivity St: ------------------------------------------------------------------------------ > What happens then to an e-mail in the form user@[aaa.bbb.ccc.ddd]? > Reverse lookup the name of aaa.bbb.ccc.ddd, lookup its MX record and the > mail sent there? Nope. If MTA decides this mail will go via SMTP protocol, it tries to connect to port 25 at the IP address specified in square brackets. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 1:26 #324 By: Lech Szychowski To: Pedro Lima Re: IP-connectivity St: ------------------------------------------------------------------------------ > I thought an e-mail in the form user@[aaa.bbb.ccc.ddd] would be directed > to port 25 of the machine at aaa.bbb.ccc.ddd, but maybe it isn't as > simple as that. It is, unless you work very hard to change this (provided the MTA you use is flexible enough to allow to be tweaked that much). Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 1:31 #325 By: Lech Szychowski To: Pedro Lima Re: IP-connectivity St: ------------------------------------------------------------------------------ > LS> if system operator really wants to be a node he can very well > LS> become a traditional technology based one. > The problem with that approach is that whoever wants to connect to such > a node would then be subject to the telco charges, and it's probably > cheaper to do it via Internet. True enough. Therefore I suggested that we need an IP-specific equivalent of minimal requirements for a node. If we choose to follow standard (PSTN) rules as close as possible (which seems a logical and natural choice), then IP-node would have to be accesible (ie accept incoming sessions) during the equivalent of ZMH and support some minimal set of protocols (and we have to define this set). Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 1:37 #326 By: Lech Szychowski To: Steve Woodmore Re: IP-connectivity St: ------------------------------------------------------------------------------ > So you are just concentrating on IP only nodes then?, and forgetting all > the other ways that fidonet mail can be shifted over the internet. Hmm... The Internet seems to be an IP based network - and using the IP protocol (most likely plus the TCP layer) is the only way of sending data thru this network. I can't see your point here. Well, to be honest, I think I see a possible point here: are you trying to say that we should also provide enough flexibility to be able to accomodate also addresses like X.25, X.400 etc? Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 1:44 #327 By: Lech Szychowski To: Pedro Lima Re: IP-connectivity St: ------------------------------------------------------------------------------ > If we go by the book, then P4 doesn't allow for ammendments, [...] > This leaves us with no option but to replace it if we want > to change something in it. That's a hell of a task. But I'm afraid you are right. If we are to introduce fundamental changes into Fidonet (and using worldwide packet networks as carrier media is a fundamental change) then we have to accept that we might need some more fundamental changes concernig other issues - P4 being one of them. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 1:56 #328 By: Lech Szychowski To: Pedro Lima Re: IP-access St: ------------------------------------------------------------------------------ > LS> You are right. But no DNS record should be cached for a period of time > LS> longer than its TTL says. > So I understand. It's possible, then. Yes. But at the expense of generating much additional netork traffic and increasing server load - and therefore should be avoided whenever possible. > Actually, I don't see it as a serious problem. Such nodes would be "up" > at a distinct time of day (declared on their Txx flags?), probably > during an hour or two, and given a short TTL, the records would then > probably be already updated by the time someone tries to connect, Well, it might (even more: it will) work this way. What we are looking at here is a classical "register yourself" scheme: node goes on-line, notifies DNS server, server updates its table, short expiry time makes everyone look the name up at the nameservers listed as authoritative for the domain and these queries get answered with the correct address. But we have to be aware that implementing this approach we just will have to think about decentralizing DNS services. I'm not trying to say it's good or bad - I just want us to be aware of as many implications as possible. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 2:00 #329 By: Lech Szychowski To: Pedro Lima Re: IP-connectivity St: ------------------------------------------------------------------------------ > No, I could be able to be in the Internet every day at the same hour > using my dynamic IP account. The problem is how would someone know where If you are, then - as with answering PSTN calls - this part of node requirements is IMHO satisfied. I believe most of us will agree this (plus following the technical standards) is enough to constitute a valid base for requesting node status. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 2:05 #330 By: Lech Szychowski To: maciej grzeszczuk Re: IP-access St: ------------------------------------------------------------------------------ > emsi (vmodem) would be standard for ip-nodes. as the base protocol. > that would simplify many things... Sure, we might decide to make it so. But for sake of not introducing any possible misunderstanding I'd rather not use the name "EMSI" here - at least until we decide/agree EMSI should be the standard. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 2:11 #331 By: Lech Szychowski To: maciej grzeszczuk Re: IP-access St: ------------------------------------------------------------------------------ >> I hope we do want to solve this problem. > but it shouldn't be our primary case now.. Agreed. Let's just not forget about it. >> Yes, but you have to do it manually - which, excuse le mot, sucks when [...] > sure. but for now, we haven't got better idea. we don't have better idea Well, we have the idea. We just do not have (yet) the technical capabilities. > to solve the same problem with standard modem connected nodes with > multiple lines (like 2:480/19 and 2:480/89). I believe the solution we'll come to might work for both problems. To be the right one, it should IMHO allow for specifying an arbitrary set of IP addresses, phone numbers and whatever else we can think about. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 2:12 #332 By: Lech Szychowski To: Mats Wallin Re: IP-connectivity St: ------------------------------------------------------------------------------ > LS> C'mon, all decent mail editors support aliases. > Sure, but why should I have to define a lot of aliases when I instead > can use the nodelist for this information? You're right. Eventually one should be able to use nodelist for this. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 2:15 #333 By: Lech Szychowski To: Rune Johansen Re: Proposal? St: ------------------------------------------------------------------------------ > Just trying to point out that we should make this undependant on the > transfer methods that we know now, to allow for other methods later. Point taken. Of course I agree :) I think if we adopt the "flags" approach we may just have new flags introduced any time a new protocol becomes reliable/usable/popular enough. Leszek. --- FastEcho+ 1.45 * Origin: Abandon hope all ye enter here (2:480/33.7) SEEN-BY: 201/329 212/8 235/100 254/60 280/100 292/854 857 876 335/533 SEEN-BY: 362/21 380/106 465/50 480/19 33 43 49 70 80 88 91 481/21 SEEN-BY: 484/2 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 10:20 #334 By: Steve Woodmore To: Pedro Lima Re: IP-connectivity St: ------------------------------------------------------------------------------ * Replying to a message in : SAVEAREA Hi Pedro Lima, hope you are having a nice day 28-Sep-97 07:19:56, Pedro Lima wrote to Steve Woodmore Subject: IP-connectivity SW>> The IP address is mine, and you can send email to it at any time PL> I thought an e-mail in the form user@[aaa.bbb.ccc.ddd] would be PL> directed to port 25 of the machine at aaa.bbb.ccc.ddd, but maybe PL> it isn't as simple as that. yes it is as simple as that or it isn't, any emails for me get held at mail.demon.co.uk until I log on and collect them. Like I say you can email me 24hrs, but only connect to me if I am online. -=> Steve Woodmore <=- --- Terminate 5.00/Pro * Origin: Descended from Apes (2:254/410.1) SEEN-BY: 2/200 201/329 212/8 235/100 251/25 254/60 61 62 261 410 620 SEEN-BY: 280/100 292/854 857 876 335/533 362/21 380/106 440/410 442/403 SEEN-BY: 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Tue 30 Sep 97 11:05 #335 By: Sean Rima To: maciej grzeszczuk Re: IMPORTANT! standard of protocol for ip-nodes proposal. St: ------------------------------------------------------------------------------ 28 Sep 97 11:56, maciej grzeszczuk wrote to Steve Woodmore: > we shouldn't allow binkd as a standard of ip node (ip system running > only > binkd, not vmodem, should not be allowed to gain node status), as it's > compatible only with itself. Sorry you are wrong there. IFcico also uses BinkD, as does Argus and 2 other mailers in the pipleline. Also Binkd is available for OS2, Windows95 , NT and all versions of Unix. I also believe it is ported Amiga and the ST. Why should I and around 180 established Ipmailers have a forced change because BinkD doesn't meet VModem. I think you would find that this would end as 2 standards and not one. Sean IPMailer : alice.pcug.co.uk BinkD Port 24554 --- Msged/NT 4.10 * Origin: DSP BBS, Reading, Berks (2:252/356@fidonet) SEEN-BY: 201/329 212/8 235/100 251/20 25 252/356 358 254/60 61 62 SEEN-BY: 254/261 410 620 257/71 280/100 292/854 857 876 335/533 362/21 SEEN-BY: 380/106 442/403 465/50 480/33 2432/200 5100/8 ----[ IP_CONNECT ]------------------------------------------------------------ On: Sat 23 Aug 97 8:50 #21 By: Alberto Pasquale To: Ward Dossche Re: IP addressing St: ------------------------------------------------------------------------------ Hello Ward! May I bring your attention to what has already been "standardized" in OS2Net ? ;S Internet flags: ;S ;S Flag Meaning ;S ===== ================================================================ ;S IP Internet Protocol. Node accepts Telnet sessions. ;S VM Vmodem capability, Host accepts Vmodem connection. ;S ========================================================================== ;S Internet-address is specified in phone number field. ;S ;S Because Internet nodes can't handle modem calls, nodelist ;S speed field should be 300 when IP or VM flag is used. ;S ;S Default port should not be specified. ;S 3141 is default Vmodem port. ;S 23 is default Telnet port. ;S ;S ========================================================================== ;S If both IP and VM flags are used, other than default port is not possible. ;S ========================================================================== ;S ;S Samples in format: ;S ;S ,phone_field,baud_field,flags ;S ;S ,192.168.1.25:669,300,IP - port 669 accepts telnet sessions ;S ,192.168.1.25:669,300,VM - port 669 accepts Vmodem sessions ;S ,bbs.os2net.net,300,IP,VM - accepts telnet and Vmodem sessions ;S ,bbs.os2net.net,300,VM - accepts Vmodem sessions only ;S ,bbs.os2net.net,300,IP - accepts telnet sessions only ;S =========================================================================== ;S ;S Bandwith, maximum speed of ISDN or Internet connection: ;S ;S Flag Meaning ;S ===== ================================================================ ;S Bxxx xxx kbit/s max speed ;S =========================================================================== ;S ;S Samples: ;S ;S B28 28800 kbit/s max speed ;S B256 256 kbit/s max speed ;S B2000 2 Mbit/s max speed ;S ========================================================================== Of course my nodelist compiler FastLst already provides full support for this kind of "phone numbers". Anyway I suppose that non-aware compilers will just prefix the IP with the international dialing code... People who do not need to call IP addresses can just ignore these entries. Alberto --- timEd/2 1.10+ --- DB 1.58/001877 * Origin: Many Glacier (2:292/854) SEEN-BY: 201/329 212/8 254/60 280/100 292/854 857 876 335/533 341/70 SEEN-BY: 362/21 380/106 465/50 480/33 2432/200