For minutes of previous meetings, please see FoundationsTeam.
Present
ColinWatson - chair
Apologies
RobbieWilliamson - schedule conflict
ScottJamesRemnant - ill
Agenda
- Feature approval deadline
- Welcome Gary Lasker to the Foundations team for this cycle, rotating from Canonical OEM Services
- AOB
Actions from this meeting
- Colin to see if -Wa,-mimplicit-it=thumb fixes qt4-x11 problems on armel, and if not punt to mobile team / #ubuntu-arm
Feature approval deadline
- Outstanding drafts / work items (all present promised to get these done):
- foundations-lucid-daily-builds (James)
- foundations-lucid-distributed-development (James)
- foundations-lucid-fix-iscsi-root (Colin)
- foundations-lucid-multiarch-support (Steve)
- foundations-lucid-puppet-installer (Mathiaz, server team)
- foundations-lucid-ratings-and-reviews-in-software-center (Michael)
- foundations-lucid-robust-python-packaging (Michael)
- foundations-lucid-upstart-policy (Scott)
- foundations-lucid-upstart-server-review (Scott)
- foundations-lucid-user-contributed-metadata-for-software-center (Michael)
- Michael says that MPT wants to have a software-center-ui-changes specification; this needs to be ready by tomorrow (and we will need help from other teams to get it done)
AOB
- qt4-x11 fails to build on armel and blocks the rest of KDE (ScottK); -Wa,-mimplicit-it=thumb may help, Colin will test
Activity reports
Colin Watson
Short week (post-UDS holiday).
- Wrote up performance review.
- Drafted specs: foundations-lucid-uec-installer-enhancement,
- foundations-lucid-ubiquity-partitioner-optimisation, foundations-lucid-btrfs-support, foundations-lucid-oem-dvd-iso, foundations-lucid-gfxboot-update, foundations-lucid-reliable-device-id-in-grub
- Initial work on foundations-lucid-gfxboot-update.
- Initial work on foundations-lucid-uec-installer-enhancement.
- Merged lp:~cjwatson/eucalyptus/question-ordering.
- Fixed eucalyptus #458904 (when installing a node, euca_find_cluster
- fails to locate the cluster controller if instances are running) the new and preferred way.
Dealt with most of the emacs22 -> emacs23 transition.
Evan Dandrea
- On Holiday
James Westby
- On Holiday
Loïc Minier
Misc
- Not much done this week; was a bit sick and jetlagged after UDS and
- was on leave for 2 days (Thursday and Friday)
Matthias Klose
- On Holiday
Michael Vogt
update-manager
- trying to reproduce irritating battery dialog in auto-launch mode
- (no success)
- add inline restart required dialog
software-center
- spec update
- sub-categories discussion with mpt
- phonecall with mpt
- Software-center: review and471 branch
- Software-center: spec importance ordering
specs
- Foundations-lucid-restart-required-dialog:
- initial work
- plus mail design team for input
- Foundations-lucid-supported-timeframe-information:
- - work on script suitable for cron.germinate - push lp:~mvo/ubuntu-maintenance-check/python-port
- Adding work items to specs
- Spec writing (lots)
Foundations-lucid-better-archive-crawler
- write spec
- added action items
- start on implementation at lp:~mvo/+junk/lp-changelogs-crawler
- add workaround for the getPublishedSources() limitations (#487522)
- work with gary/leonadrdr on debugging #487522
apt
- discussion with julianK on moving to cmake
- work with donkult on some apt pinning semantics
merges
- Merge pbuilder, libpciaccess (sync), tuxtype (sync),
- notification-daemon
sponsoring
- Merge/review/upload compiz-nowrapper branch
- Review/sponsor app-install-data-parner for karmic (brian)
misc
- Update-notifier: upload new version, remove auto-open of reboot
- required dialog
- Review update-notifier patch from pitti
- Compiz: review no_wrapper branch and send comment to amaranth
- admin stuff (expenses, holiday etc)
- App-install-data-partner: merge from brian, fix some issues, branch
- karmic, update lucid branch
- Work on computer-janitor, fix some bugs, upload new version
- Mail IS ticket about changelogs.ubuntu.com
- Write gconf python tool that can diff user settings vs schema
- defaults (for compiz)
- Compiz: look at session registration problem (with amaranth)
Scott James Remnant
- Caught up on expenses
- Boot Performance:
- Set up the two Dell Mini 10v netbooks at home to do daily
- bootcharts and installer testing
- Updated BIOS of both and got Wake-On-LAN working to make the
- process completely automated
- Wrote server-side bits to download CD images, update TFTP and NFS
- roots, and wake up the computers in turn to do test installs
- Wrote server-side bits to collect bootcharts and installer logs,
- make some HTML out of them and boot time progress graphs, and publish on rookery
- Set up the two Dell Mini 10v netbooks at home to do daily
- libnih:
- Since this is now in use by multiple pieces of software, finally
- took the step of declaring the current API as a stable one and turned it into a separate package of its own that things dynamically link to rather than including in their own source trees
- Allocator alignment fix for SPARC
- Updated the suppressions files to work with the latest valgrind,
- and fixed a few test case leaks in the process
Uploaded to lucid and cowboyed through NEW and into main
(the code already entirely exists in main three times <g>)
- Since this is now in use by multiple pieces of software, finally
- ureadahead:
- Ported to external libnih
- Updated packaging to dh 7 now I've learned that
- Fixed apport hook installation
Uploaded to lucid and cowboyed through NEW and into main
(it's already in karmic-updates after all <g>)
- Committed change to the platform seed
- (requires meta update and upload, didn't do one as there hasn't
- been one for lucid yet)
- (requires meta update and upload, didn't do one as there hasn't
- Removed old sreadahead source from the archive Once the new meta package has been uploaded, I can drop the transitional binary package for sreadahead
- mountall:
- Ported to external libnih
- Updated packaging to dh 7 now I've learned that
- Uploaded to lucid (this includes rcS fix that I should prepare a
- karmic-proposed upload for)
- gdm:
- Been tracking a failure to start gdm, especially on HDD machines,
- which I think is caused by races with udev device creation - a more precise "start on" line seems to fix it for me. Will analyse gdm and Xorg to make sure I've got it right and upload (may be worthy of karmic-proposed upload too)
- Been tracking a failure to start gdm, especially on HDD machines,
- bootchart:
- Removed bootchart-java from the archive. It doesn't work in lucid
- anyway, and the whole world is moving to prefer the Python one
- Updated the python bootchart (pybootchartgui) to SVN trunk, fixes a
- lot of bugs
- Updated bootchart packaging to only recommend pybootchartgui, and
- to Breaks bootchart-java
- Implemented cropping and annotation in pybootchartgui
- (this gets used by the daily bootcharts stuff)
- Updated bootchart package to automatically crop Ubuntu bootcharts
- to avoid the "extra 45s" complaint
- Uploaded updated packages to lucid There's an effort upstream to merge an improved version of our C-based collector with the Python chart generation into a single
"bootchart" package: http://github.com/mmeeks/bootchart We'll likely adopt this at some point, once I've had a chance to play with the taskstats-based collector and see if it works.
- Removed bootchart-java from the archive. It doesn't work in lucid
Steve Langasek
one week UDS; one week vacation
Release management
- continue with lucid archive opening: merges/updates of bsdmainutils,
lintian, emacs-goodies-el, devscripts, cdrom-detect, choose-mirror, iso-scan, debian-installer-utils, netcfg
- SRU processing
Packages
- merges/syncs for lucid: jquery, bsd-mailx, cln, db4.6, egenix-mx-base
- fix hdparm to not call hdparm -f for every drive at boot time, slowing
down boot for no reason