jeh@dcs.simpact.com (04/30/91)
I was recently checking the 'junk' newsgroup here, looking for articles that were junked but shouldn't have been, and found lots of them :-( . The culprit in each case was the Distribution: header. Here is an example: > X-NEWS: simpact junk: 1122 > Relay-Version: VMS News - V6.0-3 14/03/90 VAX/VMS V5.2; site dcs.simpact.com > Path: simpact!crash!ncr-sd!sdd.hp.com!cs.utexas.edu!cse!evax!lindahl > Newsgroups: comp.windows.x > Subject: X man pages on the SUN > Message-ID: <LINDAHL.91Apr27012548@arrisun3.utarl.edu> > From: lindahl@arrisun3.utarl.edu (Charlie S. Lindahl) > Date: 27 Apr 91 06:25:48 GMT > Sender: usenet@cse.uta.edu (USENET Dummy account for GNUS and TheNews) > Distribution: comp ^^^^ > Organization: Automation and Robotics Research Institute, U Texas @ Arlington > Lines: 59 Our news config files are set up to state that we accept comp.windows.x (all of comp, in fact) and that we accept distributions world,usa,ca,la, and sdnet . (los angeles being close enough to here so as not to matter) I wasn't aware that it was common practice to put newsgroup top-level-hierarchy names into the "Distribution:" header. To make matters worse, many of the missing articles have complete newsgroup names in the Distribution: header, e.g. "comp.windows.x" . For news.admin, news.software.b readers: What is going on here? I thought "Distribution" headers were strictly for geographical distribution names. I know that in the past there was confusion between hierarchy names and distribution names, ie they shared the same 'namespace' because they were declared in the same place in the news config files. A recent update to the VMS news package, "ANU NEWS", fixed this... and we started rejecting articles with headers like those described above. For news.software.anu-news readers: ANU NEWS should probably be fixed so that if an item arrives with a "distribution" field that is acceptable according to the systems file newsgroup filter (but not necessarily the distribution filter), the item will be accepted. In the meantime, I suppose I can go edit the distribution filter to say that we (and our news neighbors!) are in distributions comp,news,rec, etc., in addition to world,usa, etc. However, this seems counterproductive to the recent separation of newsgroup and distribution filters. --- Jamie Hanrahan (x1116), Simpact Associates, San Diego CA Internet: jeh@dcs.simpact.com, or if that fails, jeh@crash.cts.com Uucp: ...{crash,scubed,decwrl}!simpact!jeh
wisner@ims.alaska.edu (Bill Wisner) (04/30/91)
Older versions of B news would default an article's distribution to the first component of the newsgroup name. Many sites are still living in the past. It's my opinion that newsgroup names in distribution headers are put there by users who don't understand the distribution mechanism. Bill Wisner <wisner@ims.alaska.edu> Gryphon Gang Fairbanks AK 99775 bnug, dude yeah .
dsroberts@biivax.dp.beckman.com (05/03/91)
I have run into a bit of confusion on my part with the Distribution: for BIONET. One of the "owners" of bionet claims that since bionet is not part of the USENET hierarchy, a valid distribution for bionet is bionet, it works as both a newsgroup hierarchy and geographic distribution. Is distribution: limited to strictly geographic names like "world" or "ca", or are hierarchies with special distribution like "bionet" and "vmsnet" actually valid distributions? -- --------------------------------------------------------------------------- Don Roberts Internet: don@beckman.com Beckman Instruments, Inc. Yellnet: 714/961-3029 2500 Harbor Bl. Mailstop X-12 FAX: 714/961-3351 Fullerton, CA 92634 Disclaimer: Always ---------------------------------------------------------------------------
sloane@kuhub.cc.ukans.edu (05/04/91)
In article <1991May3.082931.284@biivax.dp.beckman.com>, dsroberts@biivax.dp.beckman.com writes: > I have run into a bit of confusion on my part with the Distribution: for > BIONET. One of the "owners" of bionet claims that since bionet is not part of > the USENET hierarchy, a valid distribution for bionet is bionet, it works as > both a newsgroup hierarchy and geographic distribution. Is distribution: > limited to strictly geographic names like "world" or "ca", or are hierarchies > with special distribution like "bionet" and "vmsnet" actually valid > distributions? Sorry, but there aren't any hard rules about ANYTHING in the news game. If the "owners" of bionet think it is both a distribution and a hierarchy, then it must be. If that is what they put in their articles, and you want to see the articles, then you have to play the game their way. I have found that the old B news syntax in the NEWS.SYS file of mixing the groups and distributions together is the easist way to deal with problems like this. -- USmail: Bob Sloane, University of Kansas Computer Center, Lawrence, KS, 66045 E-mail: sloane@kuhub.cc.ukans.edu, sloane@ukanvax.bitnet, AT&T: (913)864-0444