home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!charon.amdahl.com!pacbell.com!sgiblab!sdd.hp.com!wupost!emory!tridom!wabwrld!willard!morpheus!dawson
- From: dawson@morpheus.bssi.bls.com (Willard Dawson)
- Newsgroups: comp.bbs.waffle
- Subject: Re: News/mail reader
- Message-ID: <1992Nov17.150618.21522@morpheus.bssi.bls.com>
- Date: Tue, 17 Nov 1992 15:06:18 GMT
- References: <TeRauB3w165w@dogear.spk.wa.us> <Z4scuB14w165w@xocolatl.com>
- Organization: BellSouth Systems Integration
- Lines: 60
-
- eoin@xocolatl.com (John F. Snyder) writes:
-
- >bobk@dogear.spk.wa.us (Bob Kirkpatrick) writes:
- >
- >-> kjhoule@iowegia.uucp (Kevin Houle) writes:
- >->
- >-> > For the time being, I've written a program that does just that. Given a
- >-> > root directory on the command line, the program recurses through the
- >-> > root and sub-directories and deletes duplicate articles based on the
- >-> > Message-ID header. As duplicates are removed, the filenames are
- >-> > resequenced in ascending order starting with the lowest original
- >-> > article. If anyone is interested, let me know.
- >
- >-> I'm VERY interested. Depending on size, perhaps an encoded posting?
- >
- >Me too.
- >
- >Is the resequencing a configurable option?
- >
- >I can see folks missing articles after the numbers have been
- >compressed and/or having their join pointers reset because it is now
- >greater then the actual number of the messages.
-
- Methinks Kevin has overlooked this problem? I'd be interested in reading
- how one would work through the problem of updating each user's join file
- to reflect the highest message read, and keeping it properly sync-ed with
- the new "highest numbered" message per newsgroup. It all sounds quite
- nasty, and like it would be probably be compute-intensive... you'd have
- to update each line in each join file.
-
- Let's suppose you have 1000 users, each of which is subscribed to all of
- Usenet, plus a few alt groups, and maybe some local groups. How about
- 1000 lines per join file... gee, that's 1,000,000 lines to reset, each
- time you resequence the articles.
-
- That sounds like a lot of work, and I believe is all really quite unnecessary.
- Waffle will continue to work just fine if you simply delete old articles.
- Resequencing is not necessary. If you want to avoid the join file reset
- problem, have your "expire" utility save the highest numbered file, rather
- than delete it. This is not an universally good workaround, as there is
- at least one situation I can think of where I'd prefer not to keep the
- highest numbered article, even if its deletion means that users' join files
- get reset back to 0.
-
- The proper solution to the join file reset problem is to fix Waffle! What
- ought to occur, on the occasion that a user's join file indicates an article
- higher than is available, is that Waffle ought to reset the join file line
- to the highest numbered article in the group, such that that article is
- lower-numbered than the previous entry in the join file.
-
- >
- > eoin@xocolatl.com | Practice Random Kindness /~~~~~~~~~~~~~\ /
- > GEnie: J.Snyder18 | * and < D A R W I N ><
- > CI$: 70313,3111 | \|/ Senseless Acts of Beauty \_____________/ \
- > Prodigy: NOT!!!! | PGP 2.0 User _l _l _l _l
- --
- Willard Dawson, BellSouth Advanced Networks, 1100 Johnson Ferry Road, Ste. 890
- Atlanta, GA 30342, Voice: +1 404 303 2343, UUCP: gatech!vdbsan!dawson
- Internet -> X.400: /G=W/S=DAWSON/O=BSAN/ADMD=BELLSOUTH/C=US/@sprint.com
- Percent hack: dawson%morpheus@bselab.bls.com :: Standard disclaimer applies.
-