For minutes of previous meetings, please see FoundationsTeam.
EvanDandrea - chair
it was a short week because of the holidays. i fixed bug 494704 and worked on some pycon 2010 stuff. today i plan on trying my hand at packaging a python module that isn't in uni and catching up with various folks about next steps
- done: good chunk of
foundations-lucid-ubiquity-partitioner-optimisation, some tweaks to various packages needed for server eucalyptus work, quite a bit of time off doing chromeos work
- todo: want to try to get foundations-lucid-reliable-device-id-in-grub out of the way now that upstream has fixed it to not really need device.map in most cases
- blocked: foundations-lucid-gfxboot-update relies on DX work that isn't done yet; have discussed with dbarth
started to land my branches on ubiquity. Working with the X team to split i18n data out of xkb-data in support of console-setup changes. Just finished moving from gtkcomboboxes to gtkcomboentrys, to allow for type to select. Looking into decoupling bulletproofx from gdm.
- I'm continuing to do some work on software center
- I've uploaded some code and one bug fix that Michael has merged
- I'm getting going on bug triage and fixing
- Oh, and I've got Lean training in Lexington tomorrow and Fri
- I've been working on the importer, I tracked down the bug that was killing codehosting, and I'm now getting all the branches back up to date
- todo: the latter of those parts, it's worked through the backlog, but there are a bunch of failures to triage
- blocked: nothing. also, I'm out of the country for the next few weeks.
- so you'll find me at odd hours, so mail may be best to track me down if desired
- done: most gcc-4.4 ARM updates done, minor gdb update, eglibc-2.11 testing and upgrade, gcc-4.5 prepared for testing, some packaging of new upstream versions over the holidys
- next week: gcc-4.5 testing, openjdk-6 update, prepare JCK testing with sbeattie, outstanding merges
- blocked: can't think of anything
software-center: subcategory work, ratings&reviews client side work (needs LP server side, but some client side work can already be done)
- maintaince-timeframe: submited LP soyuz branch for review (spec blocked until that is done)
- "fun" with nvidia (#494627)
glanced over the upgrade tester results at http://people.ubuntu.com/~mvo/automatic-upgrade-testing/ (quite some failures)
- working this week on getting our 4th and final hardy point release rolling
be prepared to receive calls for help testing
Outstanding actions from last meeting
Actions from this meeting
Chair for next meeting
- Short week because of holidays
- Fixed bug 494704 (python-soaplib) using UDD. Submitted, reviewed, landed.
- Did some Pycon 2010 prep work
- Very short week (holidays).
- Imported openssh package history into bzr based on new Debian import.
- Got an OpenSSH Upstart job working.
- Cut off another 20 seconds from the benchmark time (down from 46.8 to 26.2 seconds) by suppressing unnecessary recalculation of menu choices. Now at 35% of original time, with a goal of 20%.
- Short week; off on Monday and vacation from the 24th to the end of the year.
- Replied to Dylan's email on changes to the slideshow and how to get up and running with testing the installer.
- Asked Dylan, Roman, and Ago to become Ubuntu Members and MOTUs, respectively. Will continue to pester until they submit :).
- Talked with QA about usb-creator test cases and added a persistence test case to Ara's new usb-creator set.
- Looking into bug 471498, so I can safely re-enable writing hostname configuration when running as oem-config.
- Short week (holidays)
- software-center - working on code to reselect an app in all cases where the list view is refreshed and keep corresponding menu items always in sync
- OEM and Foundations team meetings
- Calls with mvo and mpt
Distributed Development and Daily Builds
- Spent most of the time investigating the bug that was hammering codehosting. It got so bad that we had to halt imports. (Fixed today)
- Worked on some bugs that were affecting some peoples use of bzr-builddeb.
- Filing bugs on LP about problems I was seeing with the API, and tracking down the information they needed.
- A little sponsor review.
- Archive admin duties, including writing a script to hugely speed up
- sync processing, so there is less time spent on AA duties in future (hopefully).
- Preparation for travel.
- Short week; on leave on the 21st, 24th, and 25th of December.
- Update-manager: add percent JS method for the webkit view (for the release notes viewer)
- fix potential failure when multile meta-pkgs are installed
honor ForceOverwrite=no properly in auto-test setupd
- Auto-upgrade-tester: review logs, report bacula failure (#499389),
- work on apt-history feature
- look at and471 history branch
- work on search/raking strategies for xapian
- review spec changes
- Support-timeframe: look into soyuz integration, mail julian
- travel preparing portland
- Synaptic: i18n merges
- Review/test/upload new apt (with history branch merged) 0.7.25ubuntu1
- hunt/fix hidden ABI break
- Python-apt: migrate my branches from people.u.c to code.launchpad.net
- Synaptic: fixes in search history
- Synaptic: work on origin display logic
Scott James Remnant
- Short week due to holidays.
- Hard Drive performance:
- Did some investigation into kernel page cache flags, there's something in here that marks "readahead" but I think it's just kernel's own readahead at the moment, not readahead() pages
- Re-permitted nobootwait for virtual devices
- Reworked mountpoint tagging quite a bit
- Fixed bug where mountpoints not listed in /etc/fstab would always need to be remounted
- (util-linux) added documentation for fstab options, and made mount ignore them
- Fixes for crashes on boot caused by strcmp() on a NULL device name
- Fixed delay for the resume invocation
- Switch to mounting devtmpfs on /dev
- Fair amount of work on figuring out why things weren't working out, narrowed down to bad TTY and VT handling code. Worked on a temporary patch that only uses /dev/tty7 as the "console" and forces a VT_WAITACTIVE after the VT switch.
- Fixed "plymouth deactivate" crashes when no renderer was active, or when --show-splash had never been called.
- Fixed "plymouth quit" after "plymouth deactivate"
- Removed code that disabled plymouth when alternate init= used.
- Always start plymouth, let it use the splash kernel argument to decide whether or not to use the graphical or text mode renderer (since we always want to use plymouth for prompting)
- Learned a bit about console logging.
- another round of cryptsetup SRU fixes
- merges: acpid, hsqldb, irda-utils, libpam-ldap, xkeyboard-config, db4.6, gpredict, kallery, ruby-gnome2, kazehakase, zsync, libloader, labplot, couchdb
- syncs: im-switch, mutt, imview, powermgmt-base, netatalk
- archive admin duties