Differences between revisions 12 and 13
Revision 12 as of 2010-03-23 22:20:52
Size: 43123
Editor: pool-74-107-129-37
Revision 13 as of 2010-03-24 03:20:45
Size: 43304
Editor: www
Deletions are marked like this. Additions are marked like this.
Line 108: Line 108:
 * built new language-packs for Karmic and Lucid
 * triaged all 'New' bugs for language-selector
 * fixed a couple of bugs in language-selector
 * booked flight tickets for UDS

For minutes of previous meetings, please see DesktopTeam/Meeting.

Meeting Minutes


Main Meeting

  • Rick Spencer (rickspencer3) - chair
  • Alberto Milone (tseliot)
  • Arne Goetje (ArneGoetje)

  • Bryce Harrington (bryce)
  • Chris Cheney (ccheney)
  • Chris Coulson (chrisccoulson)
  • Didier Roche (didrocks)
  • Jonathan Riddell (Riddell)
  • Ken VanDine (kenvandine)

  • Martin Pitt (pitti)
  • Sebastien Bacher (seb128)
  • Till Kamppeter (tkamppeter)

Eastern Edition

  • Luke Yelavich (TheMuso)

  • Robert Ancell (robert_ancell)
  • Chris Rogers (RAOF)



  • Outstanding actions from last meeting
  • Partner Update
  • Kubuntu Update
  • x-update
  • mozilla update
  • Release Bugs/Release Status
  • Review activity reports
  • Any other business

Actions from this meeting

Outstanding actions from last meeting

Partner Update

Kubuntu Update

  • Beta release successful, hugs to keybuk and pitti and ev and cjwatson and the kubuntu community for helping lots there
  • However it caused milestoned bug count to multiply 6 fold, most seem fixable, count already down to < 20

  • Tracking bug 448705 akonadi server doesn't start at login, solutions seem to involve changes to mysql package
  • New logo needs deciding, I'll call a meeting of Kubuntu council and community. Awaiting "K" from design team.

x update

X Stabilization

-nouveau seems to have some stability problems. Not unexpected, we knew there would be regressions. People seem too shy about reporting them though.

-intel seems to have the usual scattered issues. We're collecting more bug reports but I don't think this means there are more bugs.

-ati seems really stable, few complaints, and rate of new bug reports seems light.

X bug triaging

This past week, myself, raof, and tseliot have divided out the bug triaging work between us. We've made some progress but there is still a lot of triaging to get caught up on.

  • Some stuff is marked off https://wiki.ubuntu.com/X/Projects

    • Some tasks will be ongoing, but we'll mark them grey when it looks like the task is under control.
  • I don't have a good feel of rate of bug upstreaming. I've been doing about 1 upstream per day myself, which I'd like to increase.

I made a graph showing just "lucid bugs needing developer attention": http://www2.bryceharrington.org:8080/X/Reports/ubuntu-x-swat/totals-lucid-workqueue.svg

Most of the growth has been coming from -intel, from the X freeze bugs being captured. Thanks to hard work from RAOF and Sarvat last night marking bunches of -intel dupes the graph is going down. Smile :-)

This raises a point though. The quantity of freeze bugs coming in is pretty high, high enough to sap some man power just reviewing and duping the bugs. It's not clear that we can flag the bugs as dupes mechanically; there is some intuition involved in identifying the dupes that is going to be hard to script. (It may get easier in lucid+1). But I wonder if we may be seeing a case of diminishing returns at this point. Presumably most of the existing freeze bugs have gotten reported by now, and we've got the info we need (or could just directly ask users to collect more).

Maybe we could turn the apport script off for now. We can turn it back on later, after we feel we've attended to most of the freeze bugs reported so far, or if we do any changes to drm and want more test results.

Big picture graph: http://www2.bryceharrington.org:8080/X/Graphs/totals.svg

mozilla update


  • Book travel
  • Blueprints, start thinking about what you what
  • Sponsorship, who should be sponsored?

Release Status


Other Business

Activity reports

Alberto Milone (tseliot)

  • Fix LP: #522969 call gmenucache trigger in all nvidia drivers.
  • Upload new fglrx driver compatible with Lucid's xserver.
  • Upload new Nvidia driver with a fix for the fan control bug.
  • Fix link to OpenCL.so library in nvidia-current.
  • Fix LP: #508890 Nautilus crashes when there are multiple XScreens.
  • Add missing link in fglrx which prevented the VA API from working.
  • Further packaging fixes for fglrx committed upstream.
  • Working on the implementation of plymouth's smooth transition in Kdm.

Arne Goetje (ArneGoetje)

  • built new language-packs for Karmic and Lucid
  • triaged all 'New' bugs for language-selector
  • fixed a couple of bugs in language-selector
  • booked flight tickets for UDS

Bryce Harrington (bryce)

  • X Stabilization
    • Plan out UDS X sessions
    • Write meeting minutes from X meeting
    • Write up status on X stabilization work
    • Add to X/Projects
    • Make procmail save bugs with tag 'lucid' to launchpad.lucid
    • Send thoughts about modesetting bug troubleshooting to kernel team
    • Triaging -ati bug reports
    • Upstream -ati bug #513956
  • Arsenal
    • Evaluate genshi to replace tt2
    • Reimplement milestone report in Arsenal
    • Write some scripting suggestions for tseliot for -nvidia
    • Plan work to improve process-new-bugs.py
    • Plan tasks to flesh out fix-titles.py more
    • Make a graph showing bugs per package that are tagged 'lucid'
    • Add nVidia Corporation G84 [10de:0407]
    • Set up chroot on blumonc that pulls in versions from debian and from prior releases
    • Set up package report cronjob
  • Bug Work
    • Run script on several packages to request retesting with new drm
    • Help keybuk try to diagnose a kms issue (bug #539730)
    • Troubleshoot why lshw causes screen to go corrupt (#527369)
  • Multitouch low level support
    • Test apw's kernel bits on lucid
    • Test touch screen support on new Fujitsu touchscreen laptop
  • Identify UI polish settings and create scripts to apply them
  • Test filing apport info from alternate machines
  • Review kees' procmail rules and scope out redoing my rules to be more concise
  • Arrange UDS travel

Chris Cheney (ccheney)

  • Uploaded openoffice.org 1:3.2.0-4ubuntu1 to lucid
  • Uploaded openoffice.org-l10n 1:3.2.0-4ubuntu1 to lucid
  • Uploaded openoffice.org-dictionaries 1:3.2.0~rc2-2ubuntu1 to lucid
  • Uploaded openoffice.org 1:3.2.0-4ubuntu0ppa1 to karmic ppa per support team
  • Uploaded openoffice.org-l10n 1:3.2.0-4ubuntu0ppa1 to karmic ppa per support team
  • Working on patches for next openoffice.org release to fix remaining major bugs
  • Working on revising patch for epiphany-browser to correct crash bugs

Chris Halse Rogers (RAOF)

  • X Stabilisation:
    • Drive through nouveau DDX bugs; all are upstream or awaiting info from reporters
    • Use upstream-testing section from NouveauEvaluation to start a Nouveau troubleshooting page

    • Pull some safe, small fixes from nouveau kernel tree for some small bugs we've filed. After testing the kernel, these will get sent to the kernel team.
    • Poke at Intel GPU hang list. With Sarvatt's help, we're down from more than 50 to 24, and we've identified a kernel patch that got accidentally dropped from the -16 kernel that should fix a large class of these.
    • Start on a patch to make VESA bail when kms is active; if VESA tries to drive a kms screen it results in pretty “blooming” rather than a useful display. Patch works, just requires some more autofoo & checks to ensure it's acceptable upstream.

  • F-Spot
    • Fixed the obnoxious problem where editing would make photos with orientation information mysteriously rotate by 90 degrees. As an added benefit, orientation information is now preserved when saving files.
    • Added the (hopefully!) final polish - handling when files appear in the directory f-spot's watching, and fixing the worst part of bug #175191 by preventing the incorrect timestamps being written to imported files.
    • Triaged some f-spot bugs while I was in the neighbourhood.
  • gnome-keyring-sharp
    • The keyring daemon in 2.30 no longer exports a socket interface, and gnome-keyring-sharp is totally dependent on that socket interface. Bug 536925 tracks this.

    • Poked at upstream, who are not particularly active.
    • There's a partial replacement by an upstream broken by the change on the bug, but that's unlikely to be acceptable to gnome-keyring-sharp upstream.
    • I'll keep prodding upstream & get an acceptable plan of attack

Didier Roche (didrocks)

  • Away on Wednesday and Thursday for Solution Linux at Canonical booth
  • WI related work:
    • gdm: continue working on setting gdm sound on and off at login. Encountered several pitfalls but on good track now.
  • Non WI related work:
    • UNE:
      • bug triaging
      • Merge outcomers contribution
      • review/setup a list of bug that should be close for beta 2
    • Prepare migration to clutter 1.2 (needed for GNOME 2.30). Needs updated introspection packages which involves work on:
      • clutter 1.2 (update)
      • clutter-gtk 0.10.4 (update)
      • gtk 2.0 (enabling and fix introspection build)
      • atk (enabling introspection)
      • pango (enabling and fix introspection build)

        -> will deliver that by Wednesday (or sooner) in the desktop team ppa

Jonathan Riddell (Riddell)

Ken VanDine (kenvandine)

Luke Yelavich (TheMuso)

Martin Pitt (pitti)

Assigned blueprints: All implemented


  • Bug fixing: apport, cdbs, udisks, policykit-default-privileges, gdm, gnome-settings-daemon, devicekit-disks, udisks, gnome-power-manager, jockey, libatasmart, media-player-info, libgphoto2, poppler, udisks
  • New versions: pm-utils, libgphoto2
  • Release management for Beta-1
  • CD install and upgrade testing
  • Debugged Kubuntu plymouth breakage
  • Job interview


  • None this week. I suck.

Robert Ancell (robert_ancell)

Sebastien Bacher (seb128)

  • desktop updates: libgpod gtksourceview totem-pl-parser gedit nautilus-sendot totem gtkmm telepathy-gabble
  • upgraded several machines to lucid beta1, did iso testing and new installations too
  • backported indicator-me fixes for status and focus issues
  • backported a gnome-desktop change to fix a gnome-settings-daemon crash
  • updated gnome-menus to fix some upgrade issues
  • debugged gnome-panel hardy upgrade issues, fix that and some other small bugs on the way
  • rebuild gir-repository with the gnome-menus binary since it's built from gnome-menus now
  • backported nautilus change to not display a frame around the desktop background
  • did a rhythmbox git fixes backporting sprint
  • resynced papyon on the debian version
  • debugged a telepathy-gabble crasher with upstream, got it fixed and the fix in lucid
  • emailed gnome-keyring and gnome list about concerning issues in the current versions,
    • got most of those fixed thanks to upstream rocking job
  • updated libgnome-keyring and gnome-keyring to git snapshots in lucid
  • backported papyon upstream changes to fix a crasher and add some debugging informations
  • reviewed and uploaded gtk changes from Cody to fix fileselector directory selection
    • issues and Michael to speed up software-center
  • backported libgnomekbd change to have the font color matching the theme in the keyboard indicator
  • cleaned evolution-indicator spamming session log
  • did another rhythmbox git fixes backporting sprint
  • looked at gdm issues with didrocks
  • update gnome-menus caching code to not translate things that should not be translated
  • updated gnome-session to undo string change
  • sponsoring: indicator-sound libindicator indicator-application ubuntu-mono

Till Kamppeter (tkamppeter)

  • Plug'n'Firmware: Worked out with mvo (update-notifier) a concept so that if a firmware-hungry printer (HP LaserJet 1000/1005/1018/1020/P100x/P150x) is connected but the needed firmware file is not installed on the computer hp-plugin (tool for firmware download and installation) is automatically popped up so that the user installs the missing firmware. This is triggered by update-notifier checking UDEV events. Uploaded new HPLIP package containing a wrapper for hp-plugin, so that the firmware installation works also without having Qt installed.

  • Ghostscript: Some code clean-up on the "cups" output device to avoid segfaults in PostScript/PDF error conditions (LP: #534525).
  • foo2zjs: Applied upstream fix to driver wrapper scripts to support custom page sizes via filter command line (LP: #539242), improved Debian patch so that downloaded color profiles and firmware files get installed at the correct place.
  • Answered and triaged printing-related bug reports.

Chris Coulson (chrisccoulson)

  • General:
    • Fixed a gnome-applets issue where the old keyboard switcher was not being cleaned from panel configs on upgrade
    • Acquired a 3G dongle and started investigating an issue some people are seeing where the dongle does not appear to be visible if connected during boot
    • Fixed a scrolling issue in gnome-terminal, introduced by my profile changes
    • Reviewed a nm-applet patch to correctly update existing notifications. Just waiting for this to be merged
    • Disabled the web UI by default again in transmission
    • Bug triage
    • Sponsoring: conkeror, gnome-web-photo, transmission
  • xulrunner 1.9.2 transition:
    • mozvoikko (done - needed a minor build-system change)
    • swt-gtk (done - needed quite a few source changes)
    • totem (done - dropped the xulrunner build depends, it's not needed)
    • xiphos (done - needed a fix to workaround a change in the GtkMozEmbed API)

    • chmsee (done - merged with Debian and applied a fix to load the 1.9.2 GRE)
    • gjs (test suite fails on i386. i spent some time trying to track this down, but not having much luck just yet)

IRC Logs

[16:31] <rickspencer3> ArneGoetje, bryceh, ccheney, chrisccoulson, didrocks, kenvandine, Nafai, pedro_, pitti, seb128, tkamppeter, tseliot
[16:31] <seb128> there
[16:31] <ArneGoetje> o/
[16:31] <tseliot> o/
[16:31] <rickspencer3> good morning, afternoon, evening ;)
[16:31] <Nafai> o/
[16:31] <didrocks> o/
[16:31] * kenvandine waves
[16:32] <rickspencer3> https://wiki.ubuntu.com/DesktopTeam/Meeting/2010-03-23/<<BR>> [16:32] <rickspencer3> shall we?
[16:32] <rickspencer3> let's go ...
[16:32] <rickspencer3> First is actions from last meeting
[16:32] <rickspencer3> this is bona fide empty, there were no actions to carry from last meeting
[16:32] <rickspencer3> next is partner update
[16:32] <rickspencer3> kenvandine, ?
[16:32] <kenvandine> yeah
[16:33] <kenvandine> sorry i didn't get to update the wiki yet
[16:33] <rickspencer3> kenvandine, understood
[16:33] <rickspencer3> you had more important things to do
[16:33] <rickspencer3> anything to report?
[16:33] <kenvandine> u1music store is in public beta now!
[16:33] <kenvandine> woot
[16:33] <rickspencer3> !
[16:33] <tseliot> \o/
[16:33] <kenvandine> please make sure to test/report bugs
[16:33] <Riddell> I've found three bugs already :)
[16:33] <kenvandine> Riddell, great... i found like 10 the first hour i used it a few weeks ago
[16:33] <kenvandine> so that is progress :-p
[16:34] <rickspencer3> :P
[16:34] <rickspencer3> wfm
[16:34] <kenvandine> also
[16:34] <kenvandine> OLS has 3 ffe bugs out there
[16:34] <kenvandine> bug 525803
[16:34] <ubottu> Launchpad bug 525803 in ubuntuone-client "Feature freeze exception: ubuntuone-client devices and services tabs" [High,In progress] https://launchpad.net/bugs/525803<<BR>> [16:34] <kenvandine> which i think is supposed to land today'ish
[16:34] <kenvandine> and bug 534707
[16:34] <ubottu> Launchpad bug 534707 in ubuntuone-client "[Freeze Exception] ubuntuone-client: Syncdaemon needs to auto-start/connect in Lucid" [High,Confirmed] https://launchpad.net/bugs/534707<<BR>> [16:34] <kenvandine> and bug 526084
[16:34] <ubottu> Launchpad bug 526084 in ubuntuone-client "[Freeze Exception] ubuntuone-client: Impossible to infer status of file syncing jobs" [High,In progress] https://launchpad.net/bugs/526084<<BR>> [16:35] <kenvandine> which are still in the works, i don't have status on those
[16:35] <kenvandine> we should have a bug fix for libu1 today as well, i need to check on that
[16:35] <kenvandine> on to DX, there is a ffe request for indicator-me
[16:36] <kenvandine> to disable the status entry when gwibber isn't available/configured
[16:36] <kenvandine> that is ready, just need approval
[16:36] * kenvandine doesn't have the bug handy this sec
[16:36] <kenvandine> bug 520932
[16:36] <ubottu> Launchpad bug 520932 in indicator-me "Status text field should be disabled if no broadcast accounts have been set up" [Medium,Fix committed] https://launchpad.net/bugs/520932<<BR>> [16:37] <kenvandine> that is all i have
[16:37] <rickspencer3> kenvandine, bug 53470 sounds like it would slow down boot?
[16:37] <ubottu> Launchpad bug 53470 in launchpad-foundations "Import fails, but sender doesn't get information about failed import (dup-of: 2038)" [Undecided,New] https://launchpad.net/bugs/53470<<BR>> [16:37] <ubottu> Launchpad bug 2038 in rosetta "Some failed uploads produce no error message" [Medium,Fix released] https://launchpad.net/bugs/2038<<BR>> [16:37] <rickspencer3> oops
[16:37] <rickspencer3> kenvandine, bug 534707 sounds like it would slow down boot?
[16:37] <ubottu> Launchpad bug 534707 in ubuntuone-client "[Freeze Exception] ubuntuone-client: Syncdaemon needs to auto-start/connect in Lucid" [High,Confirmed] https://launchpad.net/bugs/534707<<BR>> [16:37] <kenvandine> i think that is a lazy start
[16:37] <seb128> don't autostart that service, especially if it still takes some solid 15s of cpu use
[16:38] <pitti> I just commented on that
[16:38] <kenvandine> oh?
[16:38] <pitti> it should only autostart when you actually enable it, and with a solid delay
[16:38] <kenvandine> yeah
[16:39] <vish> seb128: hi.. regarding Bug 526560 , the option is not available now.. instead of it actually working
[16:39] <ubottu> Launchpad bug 526560 in gnome-keyring "Keyrings aren't unlocked if "Automatically unlock ..." is selected" [Low,Fix released] https://launchpad.net/bugs/526560<<BR>> [16:39] <kenvandine> right now it seems like the only way to make it connect is u1sdtool -c
[16:39] <kenvandine> in a terminal
[16:39] <tkamppeter> hi
[16:39] <kenvandine> which is not cool!
[16:39] <seb128> vish, hi, we are in a meeting right now
[16:39] <kenvandine> ok, that is all i have
[16:39] <seb128> vish, let's talk on an another channel?
[16:39] <vish> seb128: oh , oops sry..
[16:40] <rickspencer3> seb128, pitti, kenvandine can I assume that there will be no increase to boot time and put this out of my mind?
[16:40] <kenvandine> yes
[16:40] <pitti> rickspencer3: you can be that I'll make sure of it :)
[16:40] <rickspencer3> lol
[16:40] <kenvandine> rickspencer3, anything to slow the boot will not be accepted
[16:40] <pitti> s/be/bet/
[16:40] <rickspencer3> fair enough
[16:40] <rickspencer3> ok
[16:40] <seb128> same here
[16:40] <pitti> rickspencer3: as I said, I just commented on the bug with a question and a suggestion
[16:40] <rickspencer3> thanks kenvandine
[16:40] <seb128> I so don't want that to run on my config
[16:40] <seb128> I will make sure it doesn't if you don't use it ;-)
[16:40] <rickspencer3> moving on
[16:41] <rickspencer3> Riddell, Kubuntu status?
[16:41] <Riddell> * Beta release successful, hugs to keybuk and pitti and ev and cjwatson and the kubuntu community for helping lots there
[16:41] <Riddell> * However it caused milestoned bug count to multiply 6 fold
[16:41] <Riddell> * Good news is all the bugs seem fixable, we're already down to < 20
[16:41] <Riddell> * most worrying is bug 448705 akonadi server doesn't start at login, solutions seem to involve changes to mysql package
[16:41] <ubottu> Launchpad bug 448705 in kdepim "akonadi server doesn't start at login" [Unknown,Fix released] https://launchpad.net/bugs/448705<<BR>> [16:41] <Riddell> * New logo needs deciding, I'll call a meeting of Kubuntu council and community. Design team still owe us a "k"
[16:41] <rickspencer3> Riddell, could you please define "needs deciding"?
[16:42] <Riddell> rickspencer3: we came up with some options, we need to decide which to go with
[16:42] <rickspencer3> Riddell, who is "we" in this case?
[16:42] <pitti> Riddell: that sounds like a fun meeting; plan two hours :)
[16:42] <Riddell> Kubuntu people
[16:42] <Keybuk> Riddell: "k" - there you go
[16:42] <rickspencer3> Riddell, sorry, I am trying to understand if you are still blocked on the design team
[16:42] <Keybuk> or do you need a K ?
[16:43] * rickspencer3 kicks Keybuk under table
[16:43] <Riddell> rickspencer3: we're blocked on them by needing a properly rendered "k" in the new font
[16:43] <tseliot> :-D
[16:43] <Keybuk> rickspencer3: the comedy only works if you know that I have the new Ubuntu font installed here
[16:43] <Keybuk> so that actually looked right <g>
[16:43] <Riddell> they're blocked on us deciding the logo we want
[16:43] <rickspencer3> heh
[16:43] <rickspencer3> comedy gold
[16:43] <rickspencer3> Riddell, should I be helping or is it getting sorted?
[16:44] <Riddell> rickspencer3: it's getting sorted
[16:44] <rickspencer3> k
[16:44] <rickspencer3> thanks Riddell
[16:44] * rickspencer3 takes K and k from Keybuk
[16:44] <rickspencer3> moving on?
[16:44] <rickspencer3> bryceh, are you ready for the xorg update?
[16:44] * rickspencer3 display k and K on top of monitor
[16:45] <rickspencer3> hmmm
[16:45] <rickspencer3> maybe in Eastern Edition
[16:46] <rickspencer3> ccheney, mozilla update?
[16:46] * rickspencer3 tap tap tap
[16:46] <rickspencer3> is this thing on?
[16:46] <ccheney> i worked with asac on the patch and in the middle of getting the entry class rewritten
[16:47] <ccheney> was pulled off of it temporarily yesterday to get OOo updated as noted in the wiki
[16:47] <ccheney> but will be working on it again today
[16:47] <ccheney> we debugged the crashing in epiphany to the fact that callbacks were calling the wrong old functions and to fix it required redoing one of the classes that was backported
[16:49] <ccheney> OOo itself is in pretty good shape after the upload yesterday, i have two bugs still to fix and looking through the rest of the bug list to see if anything else major stands out
[16:49] <ccheney> openclipart can't be synced from Debian though due to what appears to be a bug in inkscape bug 529625
[16:49] <ubottu> Launchpad bug 529625 in inkscape "Sync openclipart 0.18+dfsg-9 (universe) from Debian unstable (main)" [High,Confirmed] https://launchpad.net/bugs/529625<<BR>> [16:50] <ccheney> inkscape causes the build log for even the current openclipart in lucid to go forever > 10GB anyway
[16:50] <ccheney> thats it for my status
[16:51] <rickspencer3> so in terms of mozilla, is this epiphany thing the last thing?
[16:51] <asac> rickspencer3: ccheney is working on hardy backport for epiphany
[16:51] <asac> the rest is here: https://wiki.ubuntu.com/DesktopTeam/Specs/Lucid/FirefoxNewSupportModel/xulrunner-list<<BR>> [16:51] <rickspencer3> k
[16:51] <asac> main for lucid is done (minus one package i think)
[16:52] <asac> we are ramping up getting as many universe things done by end of week
[16:52] <chrisccoulson> yeah, i think it's just packagekit now
[16:52] <asac> everything that didnt happen has to go
[16:52] <rickspencer3> ok
[16:52] <rickspencer3> so sounds on track
[16:52] <rickspencer3> this has been quite a project!
[16:52] <rickspencer3> shall we move on?
[16:52] <asac> heh
[16:53] <rickspencer3> before I turn the mic over to pitti, I wanted to bring up UDS just super briefly
[16:53] <rickspencer3> 1. you should have your flights booked
[16:53] <rickspencer3> if there is ambiguity about when to show up where, ping me out of channel
[16:53] <rickspencer3> 2. start thinking about blueprints
[16:53] <rickspencer3> I'll work with pitti and seb128 and assembling a list
[16:54] <rickspencer3> however, just think about it, focus should stay on Lucid
[16:54] * rickspencer3 hands mic to pitti
[16:54] * pitti clears throat
[16:54] <pitti> so, first a look at http://people.canonical.com/~pitti/workitems/canonical-desktop-team-ubuntu-10.04-beta-2.html<<BR>> [16:54] <pitti> this looks fairly on track and harmless right now
[16:55] <pitti> I'm a bit more concerned about our progress wrt. bugs
[16:55] <pitti> we've been really great at squashing a lot of bugs last week, but the release critical ones (https://wiki.ubuntu.com/DesktopTeam/ReleaseStatus) have been around for quite a while
[16:55] <pitti> so I'd like to discuss some of them in some detail in a bit
[16:56] <pitti> last week, seb128 introduced https://bugs.edge.launchpad.net/~canonical-desktop-team/+assignedbugs to us
[16:56] <pitti> (or rather, filled it up pretty well)
[16:56] <pitti> we got some done, but as another reminder, that's a good pool to take from if you run out of assigned RC bugs
[16:57] <dobey> hey pitti
[16:57] <pitti> in fact, https://wiki.ubuntu.com/DesktopTeam/Meeting/2010-03-23/Scratch/BugList has a slightly better sorted RC bug list
[16:57] <pitti> par odre de muftiWWWWRick kindly asked us to try and get this down by 10 this week
[16:58] <pitti> and it seems poor chrisccoulson earned half of them
[16:58] <pitti> bug 512615
[16:58] <chrisccoulson> yay \o/
[16:58] <ubottu> Launchpad bug 512615 in firefox "fonts are incorrectly rendered due to not using system cairo" [Unknown,Confirmed] https://launchpad.net/bugs/512615<<BR>> [16:58] <pitti> some thoughts about that
[16:58] <pitti> basically, we have built firefox with external cairo for years
[16:59] <pitti> is it really a trademark problem if we now just apply it to the local cairo copy in the firefox package?
[16:59] <pitti> asac is going to discuss it with upstream
[16:59] <asac> ack
[16:59] <pitti> but I wondered whether we can "just do" it for now
[16:59] <pitti> it should be a rather simple fix, and it keeps coming up
[16:59] <pitti> chrisccoulson: would that work?
[16:59] <asac> i dont care, just applying patches without getting approval might cause painful discussions
[17:00] <asac> laterone
[17:00] <chrisccoulson> i don't mind doing that if we all agree on that
[17:00] <chrisccoulson> i haven't seen the patch though ;)
[17:00] <asac> chrisccoulson: we can check that after meeting
[17:00] <pitti> asac: but we have shipped firefox with that patch (effectively) applied for many releases already?
[17:01] <asac> pitti: right. and using system-cairo was always something they disliked because they couldnt see all the patches.
[17:01] <asac> but yeah
[17:01] <asac> lets apply it and tell them: here is what we did ;)
[17:01] <pitti> asac: right, but it woulnd't change the status quo
[17:01] <pitti> and at least they can see it much better then
[17:02] <chrisccoulson> we're not the only distro with this issue are we
[17:02] <chrisccoulson> ?
[17:02] <pitti> it'd be very strange if we would be
[17:02] <asac> yeah. lets go for it. only problem i have with it, is that firefox updates system cairo regularly, so we dont really have much bake time for the rebase of this patch in security updates
[17:02] <asac> chrisccoulson: can you fire up fedora and try upstream build?
[17:03] <chrisccoulson> asac - yeah, will do that after the meeting
[17:03] <pitti> ok, thanks
[17:03] <pitti> another one I'd like to discuss is bug 532511
[17:03] <ubottu> Launchpad bug 532511 in gnome-terminal "terminal settings messed up after upgrade due to forced profile change" [High,Triaged] https://launchpad.net/bugs/532511<<BR>> [17:03] <asac> mozilla claimed that part of the font problems is our problem
[17:03] <pitti> rickspencer3, chrisccoulson: do you know what our requirements for this are?
[17:03] <pitti> certainly the new theme should be the default for new installs/new users
[17:03] <pitti> but do we have requirements for upgrades?
[17:03] <chrisccoulson> pitti - my understanding is that new installs / upgrades should get the new colours
[17:04] <chrisccoulson> but i tried to implement it in a way which didn't override existing settings
[17:04] <pitti> i. e. if we'd change it to not switch to the new theme if the user has a custom configuration, is that feasible, or would it cause trouble with sabdfl/design team/etc?
[17:04] <asac> ack. i can confirm as a "custom profile" users, that i didnt get migrated to the new profile.
[17:04] <rickspencer3> pitti, this is for the terminal?
[17:04] <pitti> yes
[17:05] <rickspencer3> I got the new profile, it was called Ambiance
[17:05] <rickspencer3> and I had to switch to it
[17:05] <pitti> after upgrade, you suddenly have huge text, scroll bars, and bad colors, etc.
[17:05] <rickspencer3> I did not have that
[17:05] <chrisccoulson> pitti - it's difficult to do though. the only key i've really switched is the one for the default profile
[17:05] <chrisccoulson> and most users don't customize that
[17:05] <rickspencer3> I had a pretty purple, transparent background with white text
[17:05] <pitti> chrisccoulson: right, but we could stop doing that if the default profile has customizations?
[17:05] <chrisccoulson> pitti - i'd have to think about how we could implement that
[17:06] <pitti> anyway, let's discuss implementation after meeting, but I was interested in the requirements
[17:06] <pitti> right now it seems stuck
[17:06] <chrisccoulson> an alternative would be for me to figure out how we could make the veraious VteTerminal bits themable
[17:06] <seb128> seems we have higher issues to fix
[17:06] <rickspencer3> hold it
[17:06] <rickspencer3> I agree with seb128
[17:06] <pitti> and this is a straight violation of the "golden rule"
[17:06] <rickspencer3> let's do something easy here
[17:06] <pitti> rickspencer3: I disagree
[17:06] <pitti> we tramp over the user's configuration
[17:06] <rickspencer3> right
[17:06] <rickspencer3> which we should not do
[17:06] <pitti> and I'd like to have a pretty strong reason to do so
[17:06] <seb128> pitti, right, let's do not touch any upgrade
[17:06] <rickspencer3> but we shouldn't get all fancy in not doing it
[17:07] <rickspencer3> just install the new profile and let users switch to it if they want
[17:07] <pitti> rickspencer3: that'd work for me
[17:07] <seb128> +1
[17:07] <chrisccoulson> rickspencer3, it wouldn't be default on new installs either then
[17:07] <pitti> rickspencer3: that'd be within the requirements from Mark, etc.?
[17:07] <rickspencer3> pitti, I saw no requirement concerning this
[17:07] <pitti> chrisccoulson: that sounds like a SMOP, though
[17:07] <seb128> chrisccoulson, how so?
[17:08] <pitti> implementation -> after meeting
[17:08] <seb128> ok
[17:08] <chrisccoulson> seb128 - the issue is that we install the profile and set it as the default. if we don't set it as the default, then new installs won't get it either
[17:08] <pitti> so, everyone's fine with "new theme on new install and upgrade only on unmodified profile"?
[17:08] <seb128> yes
[17:08] <seb128> well
[17:08] <rickspencer3> chrisccoulson, why does that stomp on the users' current profiles and cause the ugly things in the bug report?
[17:09] <seb128> I wouldn't even bother about "only on", just don't touch upgrade
[17:09] <rickspencer3> my experience was nothing like described in the bug report
[17:09] <pitti> next one is bug 456468 -- this seems to be an utterly hard thing to fix
[17:09] <ubottu> Launchpad bug 456468 in network-manager-applet "upgrade triggers nm-applet "resource not found" ... missing icon "nm-applet-device"" [Undecided,New] https://launchpad.net/bugs/456468<<BR>> [17:09] <seb128> the issue with gconf is that's it's often not easy to know if you profile is the default one or not
[17:09] <seb128> pitti, I don't think we can do anything about that one without changing hardy and karmic
[17:09] <pitti> and I wondered if we can get away with ignoring it?
[17:09] <chrisccoulson> yes, i've no idea what to do about that bug
[17:09] <ccheney> seb128: and if it is the default one, the intent, the user likes the current default or just wants whatever is default
[17:10] <pitti> at that time, apt should have done all the packages, and it doesn't break ethernet at all (just wifi, probably)
[17:10] <seb128> right, I did get disconnected there
[17:10] <seb128> didn't
[17:10] <chrisccoulson> and it shouldn't break system connections
[17:11] <pitti> ok, so I'll downgrade it
[17:11] <chrisccoulson> perhaps we could make nm-applet more robust this cycle though to stop it from happening again next cycle
[17:12] <seb128> chrisccoulson, I did that with my lazy init change
[17:12] <pitti> ok, thanks
[17:12] <chrisccoulson> seb128 - does that stop this from happening?
[17:12] <seb128> chrisccoulson, btw any news of getting the nm-applet changes it?
[17:12] <chrisccoulson> even when the theme is updated
[17:12] <seb128> chrisccoulson, yes, it uses a gtk stock icon as fallbacj when it can't load other icons
[17:12] <chrisccoulson> seb128 - i proposed a merge last week
[17:12] <pitti> bug implementation discussion> after meeting
[17:12] <pitti> rickspencer3: I'm done with release status
[17:12] <chrisccoulson> i don't know if asac saw the merge request though
[17:13] <dobey> pitti: please ping me after you're all done with the meeting :)
[17:13] <seb128> chrisccoulson, ok, let's check with asac later
[17:13] <rickspencer3> pitti, I am concerned about the number of release blockers on our list
[17:13] <rickspencer3> does it seem about normal, high, low, for this part of the cycle?
[17:14] <pitti> looks fairly normal to me
[17:14] <rickspencer3> k
[17:14] <rickspencer3> tseliot, does bug 474917 impact Lucid?
[17:14] <ubottu> Launchpad bug 474917 in nvidia-graphics-drivers-180 "nvidia drivers 185.xx compile into kernel 2.6.28 instead of 2.6.31 on update from jaunty to karmic" [Critical,In progress] https://launchpad.net/bugs/474917<<BR>> [17:15] <seb128> rickspencer3, I think it's just because I spent time fishing issues out of launchpad and putting some on our list which we don't do usually
[17:15] <seb128> I think we don't have an higher number of issues
[17:15] <tseliot> rickspencer3: nope, and it should be closed
[17:15] <seb128> we put perhaps less effort usually to get those milestoned
[17:15] <pitti> but we have some fairly bad X regressions
[17:15] <rickspencer3> tseliot, it's in progress, could you please adjust to get it off the list
[17:15] <rickspencer3> pitti, yes, we will be discussing xorg in the Eastern Edition
[17:15] <chrisccoulson> i've got a bad kernel regression here ;)
[17:16] <pitti> rickspencer3, tseliot: this is a parsing bug; this bug was closed ages ago
[17:16] <rickspencer3> tseliot, can I confirm that you working on bug 540177
[17:16] <ubottu> Launchpad bug 540177 in kdebase-workspace "KDM needs plymouth transition patch" [High,Triaged] https://launchpad.net/bugs/540177<<BR>> [17:16] <pitti> rickspencer3: fun, it affects bryceh's scripts as well
[17:16] <pitti> those also see this bug as in progress
[17:16] <tseliot> rickspencer3, pitti: ah, good to know
[17:16] <rickspencer3> pitti, ack
[17:16] <bryceh> hi
[17:17] <tseliot> rickspencer3: sure, it's just a matter of cleaning up my patch and waiting on Keybuk to upload his fix for plymouth (which he should do today)
[17:17] <rickspencer3> ok
[17:17] <rickspencer3> pitti, is it feasible to focus on the release blockers list, reduce it by say 10 bugs by eow?
[17:18] <pitti> that's what we should already be doing
[17:18] <rickspencer3> pitti, can we set a specific goal for certain bugs thoguh?
[17:18] <rickspencer3> or at least a numeric goal?
[17:18] <rickspencer3> for this week?
[17:18] <pitti> rickspencer3: bug 417009 should be fixed with latest OO.o, bug 460328 got fixed an hour ago, bug 474917 is invalid -- there, already 3 down :)
[17:18] <Keybuk> tseliot: within the next hour I expect
[17:19] <ubottu> Launchpad bug 417009 in openoffice.org "all openoffice apps die in 'com::sun::star::ucb::InteractiveAugmentedIOException' on armel in karmic" [Low,In progress] https://launchpad.net/bugs/417009<<BR>> [17:19] <ubottu> Launchpad bug 460328 in gnome-settings-daemon "Wrong keyboard settings when console-settings has multiple layouts" [High,Fix released] https://launchpad.net/bugs/460328<<BR>> [17:19] <ubottu> Launchpad bug 474917 in nvidia-graphics-drivers-180 "nvidia drivers 185.xx compile into kernel 2.6.28 instead of 2.6.31 on update from jaunty to karmic" [Critical,In progress] https://launchpad.net/bugs/474917<<BR>> [17:19] <tseliot> Keybuk: very good
[17:19] <pitti> ccheney: hm, why is 417009 still open? I thought it was applied?
[17:19] <pitti> rickspencer3: down by 10 sounds fine
[17:19] <rickspencer3> ok
[17:19] <ccheney> pitti: it was applied but there is question as to whether the build works on arm
[17:19] <rickspencer3> thanks pitti
[17:19] <ccheney> pitti: once its finished actually building ncommander is going to test it and verify its good (aiui)
[17:19] <pitti> ccheney: ah, so you wait with closing until it did? makes sense, thanks
[17:20] <ccheney> pitti: yea the patches for that bug have been through many rounds
[17:20] <rickspencer3> pl
[17:20] <rickspencer3> ok, even
[17:20] <rickspencer3> any toher business?
[17:20] * rickspencer3 shakes out finger
[17:20] <rickspencer3> s
[17:20] <rickspencer3> any other business
[17:20] <rickspencer3> ?
[17:21] <rickspencer3> ok
[17:21] <ccheney> if we have any inkscape experts looking at it would be nice, seems to have a lot of high heat crashers, which is also blocking openclipart :)
[17:21] <rickspencer3> Eastern Edition in 4.5 hours
[17:21] <rickspencer3> we'll get a good run down of xorg progress then, I htink
[17:22] <rickspencer3> ccheney, InkScape is crashy you say?
[17:22] <rickspencer3> are you looking for a specific action, or just just informational?
[17:23] <ccheney> rickspencer3: well fixing whatever is behind the failure to build openclipart is the most pressing for me, but after looking at its bug list the part about it being generally crashy is informational
[17:23] <ccheney> not sure if the solution would be to revert to the older version that used to work, or not
[17:23] <ccheney> some of the crasher bugs are from karmic timeframe so not sure about them
[17:23] <seb128> usually the way forward is to fix the crash not to downgrade
[17:24] <ccheney> yea
[17:25] <seb128> did you try to fix it?
[17:25] <ccheney> for unknown to me reasons most of the bugs are still marked private
[17:25] <ccheney> seb128: not yet, been too busy with other work to look into it
[17:25] <seb128> crashes are private until being reviewed and marked open by a triager
[17:25] <rickspencer3> will this block the release?
[17:25] <seb128> to avoid avoid passwords in a stacktrace going public or similar
[17:26] <rickspencer3> it seems that we need to find a bug report to stand in for the fix, and get that on the blocker list, and then get it fixed
[17:26] <ccheney> rickspencer3: no but users using clipart would be pretty annoyed since it is no longer installable in the archive, and we can't even rebuild the version in the archive due to the inkscape bug
[17:26] <bryceh> ccheney, we should be able to get fixes without too much trouble. What are the top 2-3 bug #'s we need fixed?
[17:26] <rickspencer3> ccheney, could you drive this please ^
[17:26] <ccheney> rickspencer3: ok
[17:27] <rickspencer3> ACTION: ccheney to determine bugs in inkscape on critical path to building open clip art
[17:27] <rickspencer3> any other business?
[17:27] <bryceh> ccheney, unfortunately I've not had much time to spend on inkscape lately, but I know folks who can help
[17:27] <pitti> dobey: pong (I'll take a quick break now, but will answer scrollback)
[17:27] <rickspencer3> ok
[17:27] <ccheney> bryceh: the issue behind bug 516771 is the one affecting openclipart
[17:27] <rickspencer3> thanks all
[17:27] <ubottu> Launchpad bug 516771 in openclipart "openclipart-openoffice.org cannot be installed in lucid as it wants to remove openoffice" [Medium,Triaged] https://launchpad.net/bugs/516771<<BR>> [17:27] <rickspencer3> see you in Eastern Edition


Back to DesktopTeam.

DesktopTeam/Meeting/2010-03-23 (last edited 2010-03-24 03:20:45 by www)