home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky alt.bbs:8220 comp.bbs.misc:1893
- Path: sparky!uunet!olivea!hal.com!decwrl!concert!gatech!rpi!usc!howland.reston.ans.net!spool.mu.edu!yale.edu!yale!mintaka.lcs.mit.edu!ai-lab!hal.gnu.ai.mit.edu!petrilli
- From: petrilli@hal.gnu.ai.mit.edu (Chris Petrilli)
- Newsgroups: alt.bbs,comp.bbs.misc
- Subject: Re: DTS-0001 proposed standard
- Message-ID: <1k6e3cINN9t0@life.ai.mit.edu>
- Date: 27 Jan 93 16:43:24 GMT
- References: <25JAN93.01484297.0042@UNBVM1.CSD.UNB.CA> <1k1706INN4f5@life.ai.mit.edu> <1993Jan26.142954.1306@globv1.hacktic.nl>
- Distribution: usa
- Organization: Department of Redundency Department
- Lines: 70
- NNTP-Posting-Host: hal.ai.mit.edu
-
- "Peter" = peter@globv1.hacktic.nl (Peter Busser)
- "Me" = petrilli@hal.gnu.ai.mit.edu (Chris Petrilli)
-
- Me> The problem I have is that it creates yet another set of "islands" in
- Me> the network that have to be interfaced. Given, those islands are
- Me> larger, but they are still islands.
-
- Peter> That is the difference between the bottom up and top down
- Peter> approach. You propose a top down approach. The proposed bottom
- Peter> up approach (i.e. the TDS-0001 document) will solve the short
- Peter> term problems of a small group of people. And a top down
- Peter> approach would solve long term problems. So I don't think that
- Peter> either way is "better" than the other as they have different
- Peter> goals. In short, it is good to have a short term solution, but
- Peter> it would be even better to have a long term solution, a real
- Peter> generic standard, as well.
-
- The problems I see stem from the incompatibilities between the long
- term and short term goals. It seems to me that FidoNet (and the other
- networks) long for connectivity to the Internet, but right now, the
- only way to do so is thorugh all kinds of gateways that are flakey at
- best. If they really want to work with the Internet, then they should
- be moving toward the Internet standards, otherwise, it'll never be
- anything more than a kludge. The introduction of a new small set of
- standards is fine, but it should be a transisition point to the final
- stage, and the standards offered are nothing like what would be
- required, and would present the need for yet another revision of the
- software, rather than just a small modification.
-
- Me> It would be better to ignore the "operating system specific"
- Me> concepts and go for REAL standards that are usable by everyone
- Me> rather than a subset. There's so much that is left out of this
- Me> document that it's difficult to use.
-
- Peter> Well, then you're going to talk about functionality and not
- Peter> about implementation. That would still allow for different
- Peter> implementations of the same standard to be incompatible. Have a
- Peter> peek in comp.mail.misc, there is a discussion about "the UNIX
- Peter> mailbox format". All mailboxes use a RFC-822 type message
- Peter> layout, but there are nevertheless many incompatibilities.
-
- If the standard is complete they will interoperate. The RFC822
- standard does not specific mailbox format, it simply specifies mail
- header and body formats and needs. The reason there are
- incompatibilities (and there are not many) is that there are two or
- three different mods to the standard "Unix mailbox," such as the RAND
- MH pack format and the Bablyl format. If you offer a standard to this
- then you fix the problem if people comply with it. I though the goal
- was to make them interoperate rahter than making them all look alike?
-
- Me> The best thing to do is just simply adopt the Internet standard
- Me> which will allow for almost anything on the face of the planet.
- Me> The date standard I mentioned is the european one, not the
- Me> american one.
-
- Peter> That won't help the people that are stuck with FidoNet stuff at
- Peter> all. The only thing for MS-DOS that looks like a BBS (flames
- Peter> >/dev/null please) and does USENET and e-mail is Waffle.
-
- It has always been my opinion that FidoNet is broken since it is
- incompatible with all the standards that are dominant in other
- computer communities, and it was short sighted not to adopt the
- Internet standards. You're gonna have to change some day, might as
- well change now.
-
- Chris
- --
- | Chris Petrilli ____
- | petrilli@gnu.ai.mit.edu ``Quayle is a bozoe.'' \ /
- | I don't even speak for myself. \/
-