[10013] in Athena Bugs
Re: serious zephyr screwup on e40-008-4
daemon@ATHENA.MIT.EDU (cfields@Athena.MIT.EDU)
Wed Oct 21 18:05:45 1992
From: cfields@Athena.MIT.EDU
Date: Wed, 21 Oct 92 18:05:25 -0400
To: jtidwell@MIT.EDU
Cc: bugs@MIT.EDU, bug-dialup@MIT.EDU, ckclark@MIT.EDU
> For what it's worth, the person whose zephyr messages I got was logged
> in on a machine in E51, which I could not possibly have ever logged in
> on. :-) However, he was logged in on e40-008-4 around that time, or
> shortly before:
>
> cfirth ttypd TORTELLINI.MIT.E Wed Oct 21 16:07 - 16:11 (00:04)
Actually, I would guess this bug is a case of cfirth's zwgc not dying
when he logged out of the dialup. If you look, I bet you'll find that
you were also on ttypd when you logged in; his zwgc was still running
attached to your tty. (So was yours, of course.)
> I restarted my zwgc and the problem went away
I bet the problem went away only coincidentally, as the dialup process
reaper ran and killed all processes not owned by people logged in.
> ...so it's certainly possible that we're talking about the same bug
> here.
My explanations notwithstanding, these two bugs are tellingly
different:
The setup:
A logs in and out of machine X.
A logs into machine Y.
B logs into machine X.
The bug Calvin reports:
A receives zephyrgrams that B should have gotten.
The bug you report:
B receives zephyrgrams that A should have gotten.
Another difference:
In your bug, both you and the intended recipient of the
zephyrgram get a copy; in Calvin's (from my experience
with the same bug), only the wrong person gets a copy
of the zephyrgram.
Craig
