[2167] in Release_7.7_team
Minutes for 2000-03-15 meeting
daemon@ATHENA.MIT.EDU (Greg Hudson)
Wed Mar 15 14:29:21 2000
Date: Wed, 15 Mar 2000 14:28:54 -0500 (EST)
Message-Id: <200003151928.OAA25901@small-gods.mit.edu>
From: Greg Hudson <ghudson@MIT.EDU>
To: release-team@MIT.EDU
Attending: aurora rbasch othomas zacheiss ajfox wdc ghudson miki tb jweiss
1. Release progress
The IRIX update is ready to test.
Greg and miki are going to meet at 3 to discuss getting a Sun update
ready.
Linux should be more or less set, although we might need a patch
release to make it work. (Right now we think it should be fine
as-is.)
We should decide on the cutoff for Red Hat 6.2 to be released such
that we will use it for Athena 8.4. We decided the cutoff was around
beta, maybe a little earlier. (This assumes AFS will work with the
kernel it uses.)
3. Solaris 8.3.27 with patch for proE
We will try to get this out into the dev cell tonight, and the Athena
cell next Monday evening.
The Solaris patch is a libc patch, so we should be a little careful.
Greg will make sure it won't mess up private machines. (It won't; it
modifies libC.so and libdemangle.so, both of which are on local disk
and only affect C++ programs.)
3. Figurehead position
wdc may stop coming to release team meetings. jweiss tentatively
agrees to replace him as figurehead, subject to approval by Roger and
Vijay.
4. Private workstation group
The athena-pws group is winding down. One of its recommendations is
that resources be found to package stock Kerberos and AFS (and maybe
ssh) for non-Athena Unix workstations. Support may be an issue.
Mostly not relevant to the Athena release team, but interesting.
5. Bug-tracking
Bug tracking could use more resources; there are a lot of unanswered
bug reports.
The IS strategic plan suggests that we should minimize handoffs, but
for now we will continue to hand off bug reports for non-IS locker
software to other groups and close the bug report.
We will close bug reports once we have checked in a fix, even if it
hasn't been deployed yet. We should let the user know our plans for
deploying the fix.
If a non-IS person responds to a bug report, we will decide on a case
by case basis whether we need to follow up or not before closing the
bug report.
6. More software on local disk?
Bill suggested arranging things like the netscape locker so that we
could keep them on local disk but have the locker override them. The
way this worked with Andrew is that the Andrew software would look at
/mit/andrew before /usr/andrew. The team did not particularly like
the idea that if any user attaches the locker, all users on that
machine get the locker version of the software.
We discussed some cluster workstation performance issues. A lot of
students prefer the Linux machines because they are perceived as
faster. We considered various schemes for getting software like
netscape and Star Office on local disk but did not conclude that any
of them are workable. We do think that the situation on Linux where
we have all the system software local is probably a benefit.
Greg will do some relevant performance measurements (AFS versus local
disk performance on Solaris and Linux) and send mail to the team.
