weening@polya.Stanford.EDU (Joe Weening) (10/21/89)
When an article is cross-posted to several groups, some of which are
unknown, should the news software preserve all of the names in the
"Newsgroups:" line? Here is an example of a problem that just
occurred. An article was received on one of our systems (Neon) with
the following header:
Path: neon!shelby!lll-winken!brutus.cs.uiuc.edu!gem.mps.ohio-state.edu!tut.cis.ohio-state.edu!ucbvax!ucdavis!caldwr!ack
From: ack@caldwr.UUCP (David Ackerman)
Newsgroups: misc.headlines,ca.general,ca.earthquakes
Subject: seismograms of the 10-17-89 San Francisco Bay quake
Keywords: seismogram earthquake
Message-ID: <594@caldwr.UUCP>
Date: 20 Oct 89 02:37:47 GMT
Organization: California Department of Water Resources
Lines: 2482
It arrived on another system (Polya) by a different route, as:
Path: polya!Teknowledge.COM!uw-beaver!cornell!mailrus!tut.cis.ohio-state.edu!ucbvax!ucdavis!caldwr!ack
From: ack@caldwr.UUCP (David Ackerman)
Newsgroups: misc.headlines
Subject: seismograms of the 10-17-89 San Francisco Bay quake
Keywords: seismogram earthquake
Message-ID: <594@caldwr.UUCP>
Date: 20 Oct 89 02:37:47 GMT
Organization: California Department of Water Resources
Lines: 2482
Somewhere along the line, "ca.general" and "ca.earthquakes" got
removed, with the result that people on Polya who subscribed to these
groups but not to misc.headlines never got to see the article.
Perhaps we can identify the culprit, and fix their software. I have a
feeling this is happening fairly often but goes unnoticed.karl@cheops.cis.ohio-state.edu (Karl Kleinpaste) (10/21/89)
The article is intact at Tut. --Karl
henry@utzoo.uucp (Henry Spencer) (10/22/89)
In article <12563@polya.Stanford.EDU> weening@Polya.Stanford.EDU (Joe Weening) writes: >When an article is cross-posted to several groups, some of which are >unknown, should the news software preserve all of the names in the >"Newsgroups:" line? ... Yes, for the reasons you go on to describe: it is quite possible for an article to reach a site that knows about the groups via one that doesn't. Rewriting of headers is evil and should be avoided if humanly possible. (There are one or two cases, like "Path", where it can't be avoided.) C News does not gratuitously rewrite headers. -- A bit of tolerance is worth a | Henry Spencer at U of Toronto Zoology megabyte of flaming. | uunet!attcan!utzoo!henry henry@zoo.toronto.edu
jbuck@epimass.EPI.COM (Joe Buck) (10/24/89)
In article <12563@polya.Stanford.EDU> weening@Polya.Stanford.EDU (Joe Weening) writes: >When an article is cross-posted to several groups, some of which are >unknown, should the news software preserve all of the names in the >"Newsgroups:" line? Up until version 2.10.2, unknown newsgroups were stripped. 2.11 and C news preserve all newsgroups. I'm not sure what 2.10.3 did. The RFCs specify that all newsgroups should be preserved. In your example, you have > Newsgroups: misc.headlines,ca.general,ca.earthquakes It went through a machine running old news that stripped the "ca" groups. The fastest way to move news from SF to LA often goes out of state, so this type of thing used to happen a lot before 2.11 became common. >Perhaps we can identify the culprit, and fix their software. I have a >feeling this is happening fairly often but goes unnoticed. Based on the paths you show, shelby, lll-winken, brutus.cs.uiuc.edu, or gem.mps.ohio-state.edu must be guilty, and I very much doubt that it's the Ohio State machine. -- -- Joe Buck, just visiting/consulting at Entropic -- write me at: jbuck@janus.berkeley.edu ...!{uunet,ucbvax}!janus.berkeley.edu!jbuck
coolidge@brutus.cs.uiuc.edu (John Coolidge) (10/24/89)
jbuck@epimass.EPI.COM (Joe Buck) writes: >>Perhaps we can identify the culprit, and fix their software. I have a >>feeling this is happening fairly often but goes unnoticed. >Based on the paths you show, shelby, lll-winken, brutus.cs.uiuc.edu, >or gem.mps.ohio-state.edu must be guilty, and I very much doubt that >it's the Ohio State machine. 'Tis not us. We're running C News, which doesn't do such evils. In any case, we carry the ca.* groups (and anything else our feeds ship us). --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.
alden@gem.mps.ohio-state.edu (Dave Alden) (10/24/89)
In article <1989Oct23.201529.23634@brutus.cs.uiuc.edu> coolidge@cs.uiuc.edu writes: >jbuck@epimass.EPI.COM (Joe Buck) writes: > >>Based on the paths you show, shelby, lll-winken, brutus.cs.uiuc.edu, >>or gem.mps.ohio-state.edu must be guilty, and I very much doubt that >>it's the Ohio State machine. > >'Tis not us. We're running C News, which doesn't do such evils. In any >case, we carry the ca.* groups (and anything else our feeds ship us). > >--John 'Tis not us either - I've checked and we passed the article along complete with ca.general and ca.earthquakes in the Newsgroups line. We're currently running B News 2.11.18 (soon to upgrade, very soon :-) ...dave ps Thanks for the compliment Joe - I owe you a beer. :-) -- Dave Alden, System Programmer MaBellNet: (614) 292-4962 Department of Mathematics Internet: alden@shape.mps.ohio-state.edu The Ohio State University UUCP: ...!{att,pyramid,killer}!osu-cis! 231 W 18th Avenue shape.mps.ohio-state.edu!alden Columbus, OH USA 43210 BITNET: alden@ohstpy
weening@Gang-of-Four.Stanford.EDU (Joe Weening) (10/27/89)
In article <3672@epimass.EPI.COM> jbuck@epimass.EPI.COM (Joe Buck) writes:
Up until version 2.10.2, unknown newsgroups were stripped. 2.11 and C
news preserve all newsgroups. I'm not sure what 2.10.3 did. The RFCs
specify that all newsgroups should be preserved.
The site that I think is doing this is running 2.10.3. If anyone
knows for sure whether 2.10.3 has this problem, please let me know.
Based on the paths you show, shelby, lll-winken, brutus.cs.uiuc.edu,
or gem.mps.ohio-state.edu must be guilty, and I very much doubt that
it's the Ohio State machine.
As one or two people have pointed out, the system modifying the header
was not one of these (it was on the other path of the two I showed).
Since I'm now fairly sure who it is, we can move this discussion to
private mail. Thanks for all of the comments.
--
Joe Weening Computer Science Dept.
weening@Gang-of-Four.Stanford.EDU Stanford University