Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!cs.utexas.edu!bcm!tmc.edu!sob
From: sob@tmc.edu (Stan Barber)
Newsgroups: news.software.nntp
Subject: Re: NNTPD hates Message-IDs with TWO '@'s in them.  (BIG log file attached to this posting)
Message-ID: <6039@gazette.bcm.tmc.edu>
Date: 18 Jun 91 04:05:58 GMT
References: <1991Jun13.043253.20660@zoo.toronto.edu> <6014@gazette.bcm.tmc.edu> <1991Jun17.192620.27934@zoo.toronto.edu>
Sender: usenet@bcm.tmc.edu
Organization: Baylor College of Medicine, Houston, TX
Lines: 38
Nntp-Posting-Host: tmc.edu

In article <1991Jun17.192620.27934@zoo.toronto.edu> henry@zoo.toronto.edu (Henry Spencer) writes:
>The problem here is that the relationship between 1036 and 822 is confused
>and ambiguous.  1036 does *not* "follow 822/1123 for all headers defined
>in those documents"; it often tightens up the rules in one way or another.
>A superficial reading of 1036's comments on message IDs suggests a loosening
>of rules, although this is arguably contradicted by the phrasing about news
>articles conforming to 822.  I'm not saying that "this is ok"; I'm saying
>that it is not entirely clear that it is *not* ok, because it's less than
>crystal clear just whose rules apply.
I very much disagree. RFC1036 CLEARLY says that where RFC1036 and RFC822
differ, RFC822 is taken to be correct. I don't believe that this is 
can be debated to any person who reads the RFC and does not try to read
something into it. When you and Geoff wrote Cnews, you claimed to following
RFC1036. Obviously, you followed YOUR interpretation of RFC1036. I am saying
that interpretation is flawed. You can believe that mine is. I am willing
to cope with software that does not conform to my interepretation of RFC1036.
You stand firm on what you believe it sez. Sounds like a good approach to
interoperability to me. (sarcasm in previous sentence....)
>Stan Barber wrote....
>>Does this mean we need a new RFC that sez which RFC822/1123 fields we will
>>follow as defines and which we won't?
>
>No, we need some clarifications in 1036 about its relationship to 822.  Plus
>preferably a tightening-up of its spec for message IDs so that they are
>clearly a subset of the 822 ones.  (We don't want full 822 message IDs,
>including obscenities like quoted white space, to be legal in news... if
>only because most news systems have never accepted them.)

I think you mean the answer is really YES. You say "NO" and then go on to
state that we need some clarification. How does this community clarify things?
Urban Myths? Certainly, the various software implementing news transports
are in disagreement about this, so that is no valid source of clarity unless
it is something like an RFC. Please, what do you propose to do to insure
clarity?
-- 
Stan           internet: sob@bcm.tmc.edu         Director, Networking 
Olan           uucp: rutgers!bcm!sob             and Systems Support
Barber         Opinions expressed are only mine. Baylor College of Medicine