Xref: utzoo news.misc:1416 news.software.b:1355 Path: utzoo!attcan!uunet!husc6!uwvax!oddjob!ncar!ames!pasteur!agate!brahms!desj From: desj@brahms.berkeley.edu (David desJardins) Newsgroups: news.misc,news.software.b Subject: Re: Wish List re: Crossposting Summary: A new proposal designed to reduce unintentional crossposting. Message-ID: <9879@agate.BERKELEY.EDU> Date: 12 May 88 01:50:53 GMT References: <3938@gryphon.CTS.COM> <439@bacchus.DEC.COM> <52859@sun.uucp> Sender: usenet@agate.BERKELEY.EDU Reply-To: desj@brahms.Berkeley.EDU (David desJardins) Followup-To: news.misc Organization: UC Berkeley Math Dept Lines: 31 In article <52859@sun.uucp> chuq@sun.UUCP (Chuq Von Rospach) suggests that when a message is crossposted, the posting program should automatically generate a Followup-To: line directing followups only to the first group in the Newsgroups: line. Unfortunately, this proposal would require changing the installed news software at each and every network site. And everyone knows that this simply isn't going to happen. So I think that, if we really want to have an effect, that we consider a modification of Chuq's proposal. What I suggest is that sites which receive crossposted articles with blank or missing Followup-To: lines could automatically modify those articles to include Followup-To: lines to the first group in the Newsgroups: line. This would have the same effect as Chuq's proposal, not only at the site with the modification but at every downstream site. Obviously, users at either end could override this; the poster by including an explicit Followup-To: line, and the recipient by modifying the Newsgroups: line of his followup. If this filter were installed at only the backbone sites, and gradually propagated to others as they updated their news software, then it seems that it would take effect over a large fraction of the net almost immediately. It could be distributed as a patch to the news software. The more that I think about this, the better it sounds. I'd like to hear some serious discussion of any disadvantages people see to this approach, and whether it should be pursued. (I'm crossposting to news.software.b so that the people who might actually put something like this into the software will see the proposal, but the discussion should remain in news.misc, I think.) -- David desJardins