[6629] in Release_7.7_team
Minutes from release-team meeting 2/26, 3pm, N42-203
daemon@ATHENA.MIT.EDU (Jonathan Reed)
Sat Feb 27 13:12:23 2010
Content-Type: text/plain; charset=us-ascii
Mime-Version: 1.0 (Apple Message framework v1077)
From: Jonathan Reed <jdreed@MIT.EDU>
In-Reply-To: <F9BD6457-3491-429C-BE11-CDB771E90A43@mit.edu>
Date: Sat, 27 Feb 2010 13:12:11 -0500
Message-Id: <1365258C-7F37-4FD0-B933-7805C5265530@MIT.EDU>
To: release-team@MIT.EDU
Content-Transfer-Encoding: 8bit
> Agenda:
>
> 1) Commit bits for -trainees (broder)
Evan would like the trainees to get svn bits. There is general approval for this, and discussion of various methods, as we want the bar to be slightly higher than "ask". We also lack a real code review process.
We should start doing explicit ACKs or NAKs for svn commits. The ACK or NAK needs to come from someone with -root access. Currently mail to goes to source-reviewers. We should possibly set the Reply-To to debathena@ in the post-commit hook so that discussion happens on that list.
People who have attended all 3 -trainees meetings and request bits will get them. Evan will deal with compiling that list.
Going forwarding, the process will be: submit a patch that gets ACK'd, and we will give you bits to commit it, if you want them.
Also, debathena-root should get added to the .k5login for the "athenasnap" account on svn.mit.edu
-trainees will get bits after third meeting and if they ask for it
> 2) Hotline needs to be able to rename/renumber machines and force updates. What's the best way? (jdreed)
There is general approval for the method in Trac #468. jdreed will implement
> 3) Feedback on Quick Reference Cards to replace pocket ref (jdreed)
For those playing along at home, the idea is to replace the Pocket Ref with a seris of "Quick Reference Cards" in PDF format, with one "card" per section of the pocketref. WE will also provide a 4-up PDF to print pocket versions. This has the advantage that users can only print what they need (SIPB frequently reports giving out an entire pocket ref just for, e.g. AFS.).
There is general approval of this. A PDF of a draft AFS version is in ~~jdreed/Public/quickref_afs.pdf. Feedback on content is welcome, but the fonts, layout and logo have been finalized after working with the Communications team.
> 4) Other business
Mark will be out of town for the first ~2.5 weeks of March. We're still planning a cutover to defaulting to double-sided, no-header-pages for the cluster printers, but if we can't make a cutover date of March break, we will hold off until June. And perhaps holding off until June is better in general.
CUPS was upgraded on 2/23 from 1.4.2+patches to 1.4.3. This promptly broke lpq (and users noticed) so it was downgraded again on 2/25. Once we get Karmic, we will focus on documenting the SysV commands (lp, cancel, lpstat, etc) and moving users away from lpq, and lprm. The wrappers will remain, but starting in August, they will begin to whine at users who use lprng-specific syntax.
Java and printer dropdown problems have gotten better now that we decreased the browsepoll interval, but they occasionally still suck since we restart cups inside the chroot.
> 5) Karmic update and Trac review
#493. Hal keeps track of mounts in /media/.hal-mtab. We think we can ask hal to unmount anything under /media at logout time, and then forcibly umount all instances of /media and re-mount it for the next login session. The proposed tmpfs fix should still be tested if possible.
We need to actually test a karmic -cluster install.
-Jon
