[3762] in Release_7.7_team
Minutes of 2002-03-05 meeting
daemon@ATHENA.MIT.EDU (Greg Hudson)
Wed Mar 5 15:22:16 2003
Date: Wed, 5 Mar 2003 15:21:59 -0500
Message-Id: <200303052021.PAA10632@equal-rites.mit.edu>
From: Greg Hudson <ghudson@MIT.EDU>
To: release-team@MIT.EDU
Attending: ghudson aurora zacheiss rbasch amb wdc miki jweiss
1. GNOME 2.2 configuration issue
GNOME 2 doesn't understand GNOME 1 preferences. (This includes panel
preferences and many global defaults such as keyboard and mouse
settings, the desktop background, etc..) Debian has a script which
attempts to convert preferences, which we can try to adapt. Our
options are:
* Do nothing, and let everyone's preferences be reset.
* Run the script silently; if it breaks, let the user lose.
* Adopt the Debian approach, which (Greg thinks) is:
- On first GNOME 2 login, ask if the user wants to convert prefs
- If so, try to convert
- Start the desktop stuff
- Ask if the desktop looks okay as is
- If not, blow away GNOME 2 prefs and log out; next login will
have no user preferences
The consensus tended toward the third option, but we felt we should
ask Oliver.
2. GNOME 2.2 impact on documentation
We don't document much of GNOME, so the impact should mostly be on
screenshots which Heather regenerates every year anyway.
3. xscreensaver
In an effort to minimize local modifications of GNOME software, we
probably want to grab the name xscreensaver. This means:
* We will install xscreensaver in /usr/athena/bin.
* We will ask bug-sipb to change the name of the sipb xscreensaver
by July, so that people who "add -f sipb" won't lose when GNOME
software tries to run sipb xscreensaver.
4. Nautilus
After alpha, we should investigate how well Nautilus works
(particularly on slow Sun hardware) and decide whether to incorporate
it into the default session.
5. Disconnect ops
Progress on making disonnected operation work better has been stalled
due to lack of resources.
6. Alpha release
... is due April 1. Hopefully it will be ready by then.
7. Build hardware
We have plans based on the amount of money we can spend and on what
Sun does (details may not be quite accurate, but should be close):
* $30K: Replace both build machines with dual-processor 280Rs.
Rewrite build scripts to do package builds in parallel.
Expect substantial improvement.
* $15K: Replace one build machine with dual-processor 280R, rewrite
build script. Not as nice as having both machines be fast, but
significantly better than nothing.
* $6.5K: Replace one build machine with single-processor 280R, which
is based on a 900MHz Ultrasparc III instead of a 500MHz or 650MHz
Ultrasparc IIi. Expect moderate improvement, but this may be a
waste of money if Sun shortly comes out with a Netra-class machine
based on the Ultrasparc III.
* Less: wait for Sun to come out with new inexpensive server
hardware, and buy one or two of those, depending on the price and
on budget constraints. Expect moderate improvement.
In the meantime, we will temporarily put the wash on a Sunblade 150 in
Greg's office instead of the Netra in the machine room, which will
yield some small improvement. (500MHz to 650MHz processor, larger L1
cache.)