coolidge@brutus.cs.uiuc.edu (John Coolidge) (09/16/89)
Got these this morning: Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU> (UMass-Mailer 4.04) Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>?(UMass-Mailer?4.04) Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>(UMass-Mailer4.04) Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>-(UMass-Mailer-4.04) Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>.(UMass-Mailer.4.04) These are all the same article, posted to rec.music.gaffa. An analysis of the paths shows the following sites are in all five paths: [...]!tut.cis.ohio-state.edu!att!dptg!rutgers!mit-eddie!GAFFA.MIT.EDU\ !Love-Hounds-request Looks like some mailer somewhere or another is broken... Side note: as far as the "real" message id (the part between the angle brackets) is concerned these are the same article. Should news be patched to use just the text between the brackets as the message id? As far as I know, neither B 2.11 or C does this, but it looks ok from my reading of RFC1036... --John -------------------------------------------------------------------------- John L. Coolidge Internet:coolidge@cs.uiuc.edu UUCP:uiucdcs!coolidge Of course I don't speak for the U of I (or anyone else except myself) Copyright 1989 John L. Coolidge. Copying allowed if (and only if) attributed. You may redistribute this article if and only if your recipients may as well.
karl@socrates.cis.ohio-state.edu (Karl Kleinpaste) (09/16/89)
coolidge@brutus.cs.uiuc.edu writes:
Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU> (UMass-Mailer 4.04)
Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>?(UMass-Mailer?4.04)
Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>(UMass-Mailer4.04)
Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>-(UMass-Mailer-4.04)
Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>.(UMass-Mailer.4.04)
...
[...]!tut.cis.ohio-state.edu!att!dptg!rutgers!mit-eddie!GAFFA.MIT.EDU\
!Love-Hounds-request
...
The breakage starts before it got here (for the 2nd time, anyway) - we
got all 5. It's been re-injected and sent around to us multiple
times; we show up in the Path: headers more than once. This means
that someone is doing a serious no-no, sending back to sites already
in Path:. Bad juju.
Script started on Sat Sep 16 12:24:25 1989
[1] [12:24pm] tut:/usr/spool/news/rec/music/gaffa# egrep Path:\|Message-ID: 301[4-8]
3014:Path: tut.cis.ohio-state.edu!att!dptg!rutgers!mit-eddie!GAFFA.MIT.EDU!Love-Hounds-request
3014:Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU> (UMass-Mailer 4.04)
3015:Path: tut.cis.ohio-state.edu!husc6!mailrus!cs.utexas.edu!tut.cis.ohio-state.edu!att!dptg!rutgers!mit-eddie!GAFFA.MIT.EDU!Love-Hounds-request
3015:Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>?(UMass-Mailer?4.04)
3016:Path: tut.cis.ohio-state.edu!husc6!mailrus!cs.utexas.edu!uwm.edu!gem.mps.ohio-state.edu!tut.cis.ohio-state.edu!att!dptg!rutgers!mit-eddie!GAFFA.MIT.EDU!Love-Hounds-request
3016:Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>(UMass-Mailer4.04)
3017:Path: tut.cis.ohio-state.edu!husc6!mailrus!uwm.edu!uakari.primate.wisc.edu!uflorida!gatech!purdue!tut.cis.ohio-state.edu!att!dptg!rutgers!mit-eddie!GAFFA.MIT.EDU!Love-Hounds-request
3017:Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>-(UMass-Mailer-4.04)
3018:Path: tut.cis.ohio-state.edu!husc6!ginosko!aplcen!uakari.primate.wisc.edu!uwm.edu!csd4.csd.uwm.edu!bionet!apple!hercules!joyce!ames!purdue!tut.cis.ohio-state.edu!att!dptg!rutgers!mit-eddie!GAFFA.MIT.EDU!Love-Hounds-request
3018:Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>.(UMass-Mailer.4.04)
[2] [12:24pm] tut:/usr/spool/news/rec/music/gaffa# exit
script done on Sat Sep 16 12:24:44 1989
--Karlhenry@utzoo.uucp (Henry Spencer) (09/17/89)
In article <1989Sep16.160104.14146@brutus.cs.uiuc.edu> coolidge@cs.uiuc.edu writes: >Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU> (UMass-Mailer 4.04) >Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>?(UMass-Mailer?4.04) >Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>(UMass-Mailer4.04) >Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>-(UMass-Mailer-4.04) >Message-ID: <8909130935268E6.AJTQ@Mars.UCC.UMass.EDU>.(UMass-Mailer.4.04) > >Side note: as far as the "real" message id (the part between the angle >brackets) is concerned these are the same article. Should news be patched >to use just the text between the brackets as the message id? As far as >I know, neither B 2.11 or C does this, but it looks ok from my reading >of RFC1036... If you read RFC1036 very carefully, the simplified RFC822 syntax it defines does **NOT** permit parenthesized comments. (The human-readable name in the From: line is an explicit exception.) A message-ID line cannot include anything but the message-ID, which must be delimited by `<' and '>'. The above Message-ID lines are all illegal. The proper response is an error message and a refusal to forward the article. An up-to-date C News will discard the first one for having white space in the message-ID, but the current relaynews code doesn't enforce the <> rule. Probably it should -- mkhistory does. -- "Where is D.D. Harriman now, | Henry Spencer at U of Toronto Zoology when we really *need* him?" | uunet!attcan!utzoo!henry henry@zoo.toronto.edu
coolidge@brutus.cs.uiuc.edu (John Coolidge) (09/17/89)
henry@utzoo.uucp (Henry Spencer) writes: >In article <1989Sep16.160104.14146@brutus.cs.uiuc.edu> I write: >>[Message-ID's in the form Message-ID: <id>(comment)] >> >>Side note: as far as the "real" message id (the part between the angle >>brackets) is concerned these are the same article. Should news be patched >>to use just the text between the brackets as the message id? As far as >>I know, neither B 2.11 or C does this, but it looks ok from my reading >>of RFC1036... >If you read RFC1036 very carefully, the simplified RFC822 syntax it defines >does **NOT** permit parenthesized comments. (The human-readable name in >the From: line is an explicit exception.) A message-ID line cannot include >anything but the message-ID, which must be delimited by `<' and '>'. The >above Message-ID lines are all illegal. The proper response is an error >message and a refusal to forward the article. That's what I thought. I disagree, though, with the proper response: I'd rather see news simply ignore anything past the trailing '>' character, possibly also posting an error notice to the admin or some log file. That way some human intervention can try and discover the cause of the offending header and berate them for non-compliance, but the article is still accepted. In other words: 'be generous with what you accept and conservative with what you generate'. --John -------------------------------------------------------------------------- John L. Coolidge Internet:coolidge@cs.uiuc.edu UUCP:uiucdcs!coolidge Of course I don't speak for the U of I (or anyone else except myself) Copyright 1989 John L. Coolidge. Copying allowed if (and only if) attributed. You may redistribute this article if and only if your recipients may as well.
bill@twwells.com (T. William Wells) (09/18/89)
In article <1989Sep17.014958.26095@utzoo.uucp> henry@utzoo.uucp (Henry Spencer) writes:
: An up-to-date C News will discard the first one for having white space in
: the message-ID, but the current relaynews code doesn't enforce the <> rule.
: Probably it should -- mkhistory does.
I'd be happier if it did.
---
Bill { uunet | novavax | ankh | sunvice } !twwells!bill
bill@twwells.com