Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.2 9/3/84 (WLS Mods); site astrovax.UUCP Path: utzoo!watmath!clyde!bonnie!akgua!astrovax!wls From: wls@astrovax.UUCP (William L. Sebok) Newsgroups: net.news Subject: Re: Readnews skipping postings Message-ID: <483@astrovax.UUCP> Date: Sun, 4-Nov-84 14:45:24 EST Article-I.D.: astrovax.483 Posted: Sun Nov 4 14:45:24 1984 Date-Received: Tue, 6-Nov-84 04:28:16 EST References: <5516@brl-tgr.ARPA> Distribution: net Organization: Princeton Univ. Astrophysics Lines: 66 > Has the new (2.10.2 or whatever) readnews incorporated some new version of > code or added some processing that attempts to be "smart" about skipping > over postings that it thinks you've already seen? I ask this because I > just noticed a strange behavior which I cannot explain, and which I do not > recall seeing under the older readnews. > .... > I normally run readnews with simple shell scripts that call readnews on a > group of newsgroups, with the format "readnews -n net.xxx,net.zzz,net.qqq" > and so on, grouping the newsgroups into categories as I wish. I just was > reading using the following script: > > readnews -n net.flame,net.followup,net.misc,net.politics,net.women > > I was reading messages in net.misc, and suddenly noticed that I had > skipped over some posting numbers. Normally, this wouldn't be unusual -- > I would assume that the item had been in both net.flame and net.misc, > and I had already read it as the system went through net.flame. However, > I had just started this process, having "q'd" out of it before lunch, > and had begun in net.misc; there was nothing new in net.flame when I > began reading again. I entered the missing article numbers, and found > that they were NOT addressed to net.flame at all -- they were posted > to net.motss, net.politics, and net.misc (in that order). I then "q'd" > out, edited my .newsrc net.misc entry to point to before those items, > and ran readnews with the script again. Again it skipped those items!!! > > I then "q'd" again, re-edited .newsrc, and did a "readnews -n net.misc" > directly. THEN it would display those skipped items. I surmised that > having the "net.motss" entry as the first posting had some effect. > I did a "readnews -n net.misc -x -t Homophobic" to look for any > similar postings, and found some which I had never seen before, and some > which I recalled reading. (Why would some be skipped and some > displayed? I saw some followups but had not seen the base items!) OK, > so I thought that maybe adding "net.motss" to my shell script might have an > effect. I again re-edited .newsrc to point before those skipped items, > and also added to the shell script, so it read: > > readnews -n net.flame,net.followup,net.misc,net.politics,net.women,net.motss > > I then executed it again. IT AGAIN SKIPPED THE ITEMS IN QUESTION!!!! 2.10.1 and 2.10.2 vnews both have this behaviour (see my previous posting about bugs). Another way to invoke this behaviour (in vnews, I haven't tried it in readnews) is by unsubcribing to groups with the options line in the .newsrc For instance, the beginning of my options line reads: options -n all,!fa,junk,control,net.followup,!net.ai,!net.micro,!net.puzzle, net.flame,!net.religion,!net.politics,!net.motss,!net.kids,!net.music, ... I like to do it this way because I can overide it and see the articles in these groups if I have extra time on my hands. Also "N" will refuse to go to a group if one has unsubscribed to it the normal way. Anyway when one unsubscribes from a group as above an article will not be shown to you if it has any such group name in its header preceding the groups you normally read. The fix? I don't know, I haven't been ambitious enough yet to dig into that part of the code. I have changed here the behavior of "N" to allow you to go to normally unsubscribed groups. This is a part of a fair amount of other work that I will post when I think it has stabilized. With this change I can go back to using the normal way to unsubscribe to groups. -- Bill Sebok Princeton University, Astrophysics {allegra,akgua,burl,cbosgd,decvax,ihnp4,noao,princeton,vax135}!astrovax!wls