FoundationsTeam/Meetings/2008/0723

For minutes of previous meetings, please see https://wiki.ubuntu.com/PlatformTeam.

Present

None

Everything except desktop should be viable; the desktop images are rsyncable but not yet worth testing.

Arne: language-selector - finish the language-support package split we have started in Hardy an\

d make it selectable in the UI. Good progress. Expect it to be ready for FF

Asac: flash experience - partly done
Asac: safe upgrading - probably on track
Asac: NM 0.7 and 3g-networking-intrepid - should enter intrepid after alpha3, probably won't be very well tested until beta
Calc: ooo-intrepid-upload-schedule - slightly delayed, probably Aug 8 for FF
Doko: java in main - getting in shape, will be done for FF. Java applet situation probably won't improve this cycle.
liw: cleanup-cruft: progressing nicely (cleaning up prototype code from the sprint)
liw: python-memory-profiling-tool: stalled
liw: get-rid-of-python-central-and-support: slow progress this week, need to finish the doc/spec
liw: gobby-server-persistent-state: code works, waiting for reactions from IS and upstream
liw: fast-lsb-release: not started yet
liw: super-piuparts, lintian-harness-improvements: post feature freeze
liw: no-fsck-at-boot: to be reassigned, but have concluded that it's not doable without new magic
slangasek: pam configuration magic - not on the schedule officially, but probably can be done in time for FF
TheMuso: Desktop Accessibility Review spec - Some of the spec will be completed by FF.
TheMuso: Dmraid - Need to complete more reading, but confident that this will be completed by FF.

To do:

Ubuntu Distributed Development:

Ran first full test import of packages. It has almost completed. At this point there are 13176 attempted imports and 214 failures. Many of the failures are understood and should be fixed. Those packages will be tested again this week.
At the sprint had many discussions with Robert and others about various issues. I have a better understanding of some of the ideas now, and a good idea of what needs to be done in the next few weeks.
Gave a lightning talk at LugRadio Live on the topic, but I wasn't very happy with it.
Update to python-debian with my changelog.py improvements. Hopefully in time for lenny.

Other:

Discussions with Arne on font-selector. We have an idea of something that should be achievable without too much work, though it is far short of what is wanted from the tool. Any more will require modifications to fontconfig to expose more information, or re-implementing much of the configuration logic in the tool itself.

Finished (I hope) personal development review stuff for this cycle.
Intrepid sprint in London.
Discussions with Matthias Klose, Michael Vogt, Steve Langasek, Robert Collins about changes to Python packaging. It's still somewhat unclear to me why things are the way they are now, but collecting the information I get to the spec page (GetRidOfPythonCentralAndSupport).
Finished a prototype of system-cleaner, with Michael Vogt. The command line version now cleans up a couple of different kinds of cruft. Michael also made a .glade file for prototyping the GUI. Gave a demo at the sprint.
Toyed with running fsck in non-writing mode from cron. Unfortunately, unless the filesystem is quiet, this will result in fsck reporting problems even when there are none. Back to the drawing board, or possibly this spec can be dropped. It would be possible to use LVM snapshots to do this, but that's not a general solution.
Fell in love with Ken Wimer's dark theme for GNOME.
All the usual random, possibly useful development discussions that happen at these sprints happened. Gave unasked, but perhaps non-unwanted opinions on various topics, including Bazaar stuff. Annoyed Robert with hints that he should make bzr be able to handle medium sized BLOBs (in the gigabyte to tens of gigabytes range).

Accessibility:

Updated accessibility stack to latest GNOME development release versions, a couple of packages were sponsored uploads.
Started looking into the python-xlib API to be used with ubiquity to properly check that at-spi gets loaded. At the moment a timeout is used, which doesn't work in all cases.
Announced and updated packages in the accessibility-dev PPA for users to use/test.
Started preparing an SRU to fix a memory leak in the hardy version of at-spi.
Started reading further into how to use policykit/dbus from python, for the accessibility desktop review spec.

Dmraid:

Managed to graft some dmraid disk metadata onto a pair of Kvm/Qemu disks. These disks are 80G when fully used, however for testing they will likely never get that big. This is a good intermediate step to allow for VM testing of dmraid.
Spent time with Colin learning about the internals and functionality of partman, needed to revamp the partman dmraid UI for d-i.
Started reading libparted code with the eventual goal of properly educating it about dmraid devices.
Finished breaking dmraid out into library/executable packages.

Misc:

Release work:

fix metalink files for Ubuntu 8.04.1 and fix scripts to generate them correctly in the future (bug #246397)
fix arts ftbfs to help get kubuntu CD images down to size
release team meeting for intrepid alpha 3
preparations for alpha-3:
- smash the live CDs down to size
- work with BenC to fix livefs build failures caused by grub
- iterate candidate images

Packages:

work with kirkland on common functionality in lsb-base for an init script 'status' argument (upload sponsored)
cleaned out my sponsorship queue:
- bug #189375 deferred (no debdiff provided)
- poked at bug #138189 (pulling in python-dev as a runtime dep), now fixed via sync from Debian
- work with Colin K on UUID grub (bug #244141)
- sponsored fix for time(1) manpage (bug #71778)
work on getting samba 3.2.0 in a state that it can be considered for intrepid
work on merging new upstream version of pam, so Kees can have SHA256 support for intrepid

Specs:

conference call with kirkland, dendrobates, and jdstrand about where we want to be for the pam package in intrepid; preparing to turn PAMConfigFrameworkSpec into a proper blueprint, since this now ties in pretty closely to the pam_ecryptfs stuff as well as the pam_smbpass cleanup and should really get here for intrepid

Misc:

Distro sprint:

Work with Colin et al. on trimming recommends from CDs
discussion with Henrik & Steve B. about tracking and documenting regressions in a release
sponsor Colin K's upload of grub (bug #244141)
hallway track with Lars & Robert about replacing python-support/python-central
and a bunch of other stuff

FoundationsTeam/Meetings/2008/0723 (last edited 2009-01-06 22:03:40 by ppp-70-246-95-66)