Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!mcvax!unido!siedap!tim
From: tim@siedap.UUCP
Newsgroups: comp.bugs.misc
Subject: UKUUCP / YORKBOX Bug - (nf)
Message-ID: <4300001@siedap.UUCP>
Date: Mon, 17-Nov-86 13:41:00 EST
Article-I.D.: siedap.4300001
Posted: Mon Nov 17 13:41:00 1986
Date-Received: Wed, 19-Nov-86 02:31:23 EST
Lines: 26
Nf-ID: #N:siedap:4300001:000:1278
Nf-From: siedap!tim    Nov 17 19:41:00 1986

I have a problem when calling this version of uucp :

static char SccsId[] = "@(#)cico.c	1.7 (UKUUCP)	3/18/86";

It has a "sleep(2)" after the "alarm(0)" at the end of the over+out loop.
BUT why oh why has someone added a "sleep(2)" AFTER the over+out loop
in ukuucp ?  What happens is that ukuucp obediently sleeps at that
point - before calling cleanup() and without having set to ignore
SIGHUP and SIGINTR. So if I call over X25 and the remote host has a
YorkBox and I do NOT sleep - I zip thro' into "cleanup()" and close
the line. This hangs up my PAD, which terminates the connection,
which I ASSUME (I don't know the YorkBox too well) causes the YorkBox
to send a SIGHUP to the calling process (uucico) to inform it of the
termination. Since I'm pretty quick this signal arrives DURING the
sleep - aborting / crashing uucico. So what you see in the Audit file
with -x9 debug is the send and receive of the 'O' message (that works
fine) and then nothing more - no exit code message. Since we didn't
get as far as cleanup() the LCK file for my machine has not been
deleted - so I can't call again until the lockfile has aged past
the expiry date.

Any ideas for a  c l e a n  solution, anyone ?

Tim Reeves, Siemens A.G., Munich

	    { ...mcvax!unido!ztivax! } siedap!tim