Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!uwm.edu!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!ira.uka.de!smurf!urlichs
From: urlichs@smurf.sub.org (Matthias Urlichs)
Newsgroups: news.software.b
Subject: Re: "x" field for C news
Message-ID: <90.150.19:47:11@smurf.sub.org>
Date: 30 May 90 17:47:10 GMT
References: <EMV.90May29180815@duby.math.lsa.umich.edu>
Organization: University of Karlsruhe, FRG
Lines: 57

In news.software.b, article <EMV.90May29180815@duby.math.lsa.umich.edu>,
  emv@math.lsa.umich.edu (Edward Vielmetti) writes:
< 
< Are x-ed out groups transmitted downstream?
< 
No. (Verified by inspecting the source code.)

< If they are, is there any way of causing them not to be, short of
< banging them out of all of the entries in the sys files?
< 
However, junked articles are transmitted downstream. (x-ed out groups aren't
even placed in junk.)
If you don't want that to happen either, I think this fix will do it.
(WARNING: Looks good, currently in use here, but not thoroughly tested.)

NB: You can also bang them out in _your_ sys file entry. But I thing the "x"
entries in the active file are a much better idea.

The change has to be made to relay/procart.c:insart().
Mine currently looks like this (latest patch level - sorry, patch date):

STATIC void
insart(art)
register struct article *art;
{
        if (!(art->a_status&(ST_DROPPED|ST_REFUSED|ST_NEEDATTN))) {
                if (!art->a_filed)                      /* paranoia */
                        (void) fprintf(stderr, "%s: %s not filed by copyart!\n",
                                progname, art->h.h_msgid);
                history(art, STARTLOG);         /* history may be unwritable */
                if (art->a_status&(ST_DROPPED|ST_REFUSED|ST_NEEDATTN))
                        uninsart(art);          /* it was; can't keep the article */
                else if (art->a_status & ST_JUNKED)
                        (void) puts(' - x-ed out');      /* also ends the log line */
                else {
                        transmit(art, exclude); /* writes systems on stdout */
                        (void) putchar('\n');   /* ends the log line */
                        ctlmsg(art);            /* NCMP */
                }
#ifdef FLUSHLOG
                (void) fflush(stdout);          /* crash-proofness */
#endif
        }
        art->a_status &= ~ST_REFUSED;   /* refusal is quite casual & common */
}

BTW, there still are a few comparisons of a_status with things other than
ST_GOOD. Shouldn't these all be changed, just for paranoia's sake?

BTW2, is the problem with refiling fixed by now? (the active file entry
"sci.aquaria =rec.aquaria" causes articles with a Newsgroups line of
"sci.aquaria,rec.aquaria" to be filed under rec.aquaria twice.)

BTW3, if you have any "x" type active entries, the newest rn patches regard
these groups as nonexistent. Good idea if you ask me.
-- 
Matthias Urlichs