Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!iuvax!ndcheg!uceng!kamat From: kamat@uceng.UC.EDU (Govind N. Kamat) Newsgroups: comp.mail.uucp Subject: Re: Domain Registration Message-ID: <1137@uceng.UC.EDU> Date: 5 Jun 89 01:00:40 GMT References: <1105@mailrus.cc.umich.edu> <7518@bsu-cs.bsu.edu> <1120@uceng.UC.EDU> <107917@sun.Eng.Sun.COM> Distribution: na Organization: College of Engg., Univ. of Cincinnati Lines: 58 In article <107917@sun.Eng.Sun.COM> plocher@sun.UUCP (John Plocher) writes: >In article <1120@uceng.UC.EDU> kamat@uceng.UC.EDU (Govind N. Kamat) writes: >>Domains are administrative, not topological. But that does not in any >>way preclude setting up a domain for XNET, using your own set of names >>as I mentioned above. > > Which of these are legal? > > 121.45.33/123.XNET.NET > @Washington_D.C.XNET.NET > >Following your reasoning, Both. ... In reality, Neither. > > The first (121.45.33/123) is the "name" of a fidonet node. > This is the name by which other Fidonet nodes identify each > other. But, the syntax [0-9]*.[0-9]* is explicitly dissallowed > by the domain naming RFCs. Not true. For instance, IP addresses are used in the IN-ADDR.ARPA domain. The domain system (DNS) places no restrictions on the content of domain names; in fact, even 8-bit characters may be used. However, the various protocols in use on the Internet -- telnet, SMTP, etc. do have restrictions. Therefore, it is *recommended* that only alphabets, digits and hyphens be used in host names, so that such software won't break. > The second case is made up, simply to show that you can not use > arbitrary "names" for your machines, even if you "hide" them > behind a domain name. This "name" uses "@" and "<",">" and "." > in ways which are not consistant with the RFCs. Our domain server seems to be perfectly happy with "@ < >" in domain names. These characters are specified as special in the mail RFC, #822, which predates the domain RFCs. If they occur in the so-called "local-part", they have to be placed inside a quoted string. Even while processing XNET mail as a special case, as would be done today, this cannot be avoided. In other words, XNET participating in the domain system does not introduce any additional complexities. >I have to agree, in a general way, with the person you quoted. It is >not just Bitnet which must change the naming scheme to match. Others >have had to also. Don't get me wrong: I am not trying to belittle the problems in converting to a different naming convention. That is necessarily a painful process. However, it is not reasonable to expect the world at large to continue to provide special treatment for various "internally well-connected" networks like the hypothetical XNET. Which is what Rahul Dhesi was suggesting. > -John Plocher -- Govind N. Kamat College of Engineering kamat@uceng.UC.EDU University of Cincinnati Cincinnati, OH 45221, USA