allbery@NCoast.ORG (Brandon S. Allbery) (12/18/89)
As quoted from <CRUM.89Dec16164428@lipari.usc.edu> by crum@lipari.usc.edu (Gary L. Crum): +--------------- | I attempted to post an article to comp.sys.sun around 15-Nov, and about | one month later (11-Dec) the article appeared in USC's comp.sys.sun news | spool. Worse than the delay is the fact that the "Date:" line on the | message lists 5-Dec, not the day I composed the message. In this case, | it's no big deal, but this is yet another reason that topics like Suns | should not be left to moderated newsgroups only. +--------------- I hate to tell you, but this is SOP for moderated groups. In particular, updating the Date: header is *required*. I used to leave the Date: alone in comp.sources.misc; even with the minimal delay I hold to, articles had a tendency to be junked because they were too old according to receiving machines' expiration periods. For a longer-latency moderated group like comp.sys.sun, *every* article would be rejected as too old if the date weren't changed. And moderation delays are a fact of life, especially in sources newsgroups. The moderator in most newsgroups (comp.sources.misc is a special case, I'm just a noise filter ;-) is expected to test the software and make sure that it works properly, and try to trap any trojan horses or etc. hiding in the source. Not to mention little things like detecting problems caused by antisocial news or mail systems that decide to munch on the submission.... You'd probably hate comp.sources.misc even more, because articles go out with *my* name on them. There's a good reason for that, too: if something has to be canceled for some reason (it has happened a few times, once because someone accidentally let AT&T trade-secret source creep into his submission) I need to be the article's "owner" or I can't cancel it. And while I was able to fudge around that when c.s.m was run from ncoast, Rick Adams would take a rather dim view of my forging cancel messages on uunet. So I *have* to be the poster-of- record. (The actual submitter is always listed in the "Submitted-by:" auxiliary header.) You'd be amazed how much of a moderator's job is constrained by the capabilities, or lack thereof, of the Usenet news system. You're overreacting. There are good reasons for both of the things you complained about; don't assume the moderator is censoring postings for "ideological correctness". ++Brandon -- Brandon S. Allbery allbery@NCoast.ORG, BALLBERY (MCI Mail), ALLBERY (Delphi) uunet!hal.cwru.edu!ncoast!allbery ncoast!allbery@hal.cwru.edu bsa@telotech.uucp *(comp.sources.misc mail to comp-sources-misc[-request]@backbone.site, please)* *Third party vote-collection service: send mail to allbery@uunet.uu.net (ONLY)* expnet.all: Experiments in *net management and organization. Mail me for info.
tale@cs.rpi.edu (David C Lawrence) (12/18/89)
In <1989Dec17.172253.21978@NCoast.ORG> allbery@NCoast.ORG (Brandon S. Allbery): > You'd probably hate comp.sources.misc even more, because articles go > out with *my* name on them. There's a good reason for that, too: if > something has to be canceled for some reason [...] I need to be the > article's "owner" or I can't cancel it. More precisely, the Sender, or From if no Sender, headers should agree. Some news systems don't even check that since cancel messages are so ridiculously simple to forge. A lot of newsreaders have a hard time dealing with the theoretical rules, though. In rn, local users who post using NNTP's mini-inews often can't cancel their articles from rn because it sees the Sender: line as usenet@wherever. In GNUS, even when the Sender: line indicates the user attempting the cancel, it won't let you if the From: line doesn't match. > And while I was able to fudge around that when c.s.m was run from > ncoast, Rick Adams would take a rather dim view of my forging cancel > messages on uunet. It _isn't_ a forgery. Dan Heller (comp.sources.x) and I just had a discussion about this a couple of months ago. The Sender: line was provided for expressly the purpose of making postings for other people. The fact that it is so often filled with {news,daemon,usenet}@site is a side-effect of NNTP posting. If you set the Sender: line to yourself before issuing the posting, and make the From: line the name of the submitter, it is not a forgery either then or when you send a cancel. > So I *have* to be the poster-of- record. (The actual submitter is > always listed in the "Submitted-by:" auxiliary header.) Dan and I got into the discussion when he was bemoaning how many people think that he wrote the packages posted. Making From: the user who submitted the package provides information to news readers in a manner they are accustomed to receiving it. Replies to the author of the package are easier, KILL files work more efficiently (ie, no scanning the body of the article for Submitted-by:) if you are so inclined to nuke sources by programmer, and summaries (like = in rn) provide a little bit more information. > You'd be amazed how much of a moderator's job is constrained by the > capabilities, or lack thereof, of the Usenet news system. True enough, but this is not one of those cases. Dave -- (setq mail '("tale@cs.rpi.edu" "tale@ai.mit.edu" "tale@rpitsmts.bitnet"))
allbery@NCoast.ORG (Brandon S. Allbery) (12/19/89)
As quoted from <~#}VS@rpi.edu> by tale@cs.rpi.edu (David C Lawrence): +--------------- | In <1989Dec17.172253.21978@NCoast.ORG> allbery@NCoast.ORG (Brandon S. Allbery): | > You'd probably hate comp.sources.misc even more, because articles go | > out with *my* name on them. There's a good reason for that, too: if | > something has to be canceled for some reason [...] I need to be the | > article's "owner" or I can't cancel it. | | More precisely, the Sender, or From if no Sender, headers should | agree. Some news systems don't even check that since cancel messages +--------------- I had things set up this way originally; I couldn't cancel things. Rn wouldn't let me, readnews wouldn't let me, vnews wouldn't let me. So I gave up. I find myself wondering whether the RFCs and the newsreaders have different ideas of the rules.... ++Brandon -- Brandon S. Allbery allbery@NCoast.ORG, BALLBERY (MCI Mail), ALLBERY (Delphi) uunet!hal.cwru.edu!ncoast!allbery ncoast!allbery@hal.cwru.edu bsa@telotech.uucp *(comp.sources.misc mail to comp-sources-misc[-request]@backbone.site, please)* *Third party vote-collection service: send mail to allbery@uunet.uu.net (ONLY)* expnet.all: Experiments in *net management and organization. Mail me for info.
chip@chinacat.Lonestar.ORG (Chip Rosenthal) (12/19/89)
allbery@ncoast.ORG (Brandon S. Allbery) writes: >[regarding cancelling] I couldn't cancel things. Rn wouldn't let me, >readnews wouldn't let me, vnews wouldn't let me. So I gave up. In the case of "rn" it's a bug. About line 250 of intrp.c is the line: newsuid = atoi(index(d,':')+1) The result is that "newsuid" is always zero. The fix is to change it to: newsuid = atoi(index(d+1,':')+1); The reason is because at this point, "d" is pointing to the ":" in front of the password field of the news admin's passwd entry, and thus it atoi's the user's password.