[2020] in Release_7.7_team
8.3.24
daemon@ATHENA.MIT.EDU (Jonathon Weiss)
Thu Dec 23 19:45:01 1999
Message-Id: <199912240044.TAA09066@speaker-for-the-dead.mit.edu>
From: Jonathon Weiss <jweiss@MIT.EDU>
To: release-team@MIT.EDU
Date: Thu, 23 Dec 1999 19:44:51 -0500
So, I was supposed to send this out early in the week, but between it
slipping my mind, and my getting tied up doing the moira purge, I
didn't get to it till now which may make it moot.
It was suggested to me that we should think about pushing out 8.3.24
with the patch for the 'recursive mutex enter' panic problem. Teh
thought was that we try to get it out before New Year's, in case
someone found an attack that killed off named or inetd or somethign
that caused machines to panic, and decided that the long weekend was a
good time to try it out on all of MIT. Since this could cause disk
corruption, there some reason for concern.
The flip side is of course, that pushing out a patch release that
fixes this problem, at the same time as we change the order of
operation in the update process to try to insure that this doesn't
happen during the update, makes me nervous. I'd like to see this
release go out carefully, with more observation than most.
Of course while we could probably put the release into the dev cell
next week, I don't think there is enough time to get it into the
athena cell with a sufficient amount of care (I'm not really sure
there was if I'd sent this mail on Monday when I should have.)
However, even if we can't get this out this year, we should push
forward with it now that the term is over.
On an unrelated note, if we have a sufficient version of the sun
patches to cover all of the y2k issues, I don't see a particular
reason to update them to newer versions. Unless, of course, it fixes
some other problem we're having.
Jonathon
