Article View: news.admin.policy
Article #173Re: A R G I C allows itself far too much
From: Josh Smith
Date: Mon, 28 Dec 1992 18:50
Date: Mon, 28 Dec 1992 18:50
55 lines
3565 bytes
3565 bytes
Ken Bibb (kbibb@maui.qualfomm.com) writes: > But can [Uunet] refuse service because a *human* isn't behind the posts? Tim Pierce (twpierce@unix.amherst.edu) replies: > Who can prove that? I don't believe it's true. I'd lay money that > Argic's choice of messages to follow up is motivated by a similar kind > of pattern-scanning mechanism to Kibo's (i.e. find /usr/spool/news > -exec grep -i turk) but I think that someone actually attempts to > write at least some of the prose. While the text was certainly written by a human at some point, it appears that posts are being made without any human interaction. That is, the poster is not only pattern-scanning, but is also auto-posting replies based on the results of the scan. I've sent a couple of messages to this group and soc.culture.turkish in an attempt to gather evidence for this hypothesis, and it's pretty convincing. This morning's effort should be fairly conclusive if Argic is in fact just a machine. > But besides that, why should it? UUNET's job is to funnel news to and > from paying customers. Why should it decline that service from > anatolia just because someone's put an automaton behind it? I don't know enough about UUNET or the legality/morality of flooding the net with posts, and Argic is "flooding" a small enough segment that this probably isn't a problem. What if someone set up a daemon to follow up to every post made to every group by quoting the entire post, and adding the single line "Good points, John" or "I disagree, John" (chosen randomly) at the bottom? Given the equipment to do this quickly, this could effectively double the size of Usenet overnight. What if it posted five or ten followups to every message? What if it started following up to its own followups? Imminent death of the net predicted! (grin) Seriously, though, it seems like auto-posters could easily end up costing a lot of people a lot of money, completely aside from the hassle factor of having half of the messages in any group generated by a daemon with nothing new to say. (Several newsgroups already fit this description thanks to Argic.) I don't, unfortunately, have any solutions. I do think that it's a problem, though, and that if it happened on a wider scale and/or in more heavily traveled groups, that there would be a significant hue and cry to do _something_. I'm sure the readers of Argic's target groups would appreciate any suggestions anyone has, other than the standard "use a killfile", which does a reasonable job of eliminating he hassle factor, but doesn't do anything to reduce the costs (hundreds, if not thousands of dollars :^) of Argic's automated tirade. Evidence for the "Argic is an automaton" hypothesis coming tomorrow to a newsgroup near you. -- Josh Smith, User Support Coordinator :: irilyth@fenris.claremont.edu Harvey Mudd College, Claremont CA :: consult(std-disclaimer.pl). "It takes a big man to cry, but it takes a bigger man to laugh at that man."
Technical Details
Message-ID:
<1992Dec28.185055.18248@muddcs.claremont.edu>Path:
arc1-ca.pugleaf.net!sparky!uunet!elroy.jpl.nasa.gov!news.claremont.edu!fenris!irilythReferences:
<BzxMJI.LG4@mtholyoke.edu> <root.725487038@maui> <Bzxrs9.Bq1@unix.amherst.edu>Show Full Headers
Complete Article Headers
Newsgroups: news.admin.policy Path: arc1-ca.pugleaf.net!sparky!uunet!elroy.jpl.nasa.gov!news.claremont.edu!fenris!irilyth From: irilyth@fenris.claremont.edu (Josh Smith) Subject: Re: A R G I C allows itself far too much Message-ID: <1992Dec28.185055.18248@muddcs.claremont.edu> Originator: irilyth@fenris Sender: news@muddcs.claremont.edu (The News System) Organization: Evil Geniuses For A Better Tomorrow References: <BzxMJI.LG4@mtholyoke.edu> <root.725487038@maui> <Bzxrs9.Bq1@unix.amherst.edu> Date: Mon, 28 Dec 1992 18:50:55 GMT Lines: 55