[06:40] <dupondje> xnox: hehe :) would like to try Firefox with Wayland support enabled. But seems a hard time to build it :)
[06:43] <didrocks> good morning
[06:53] <oSoMoN> good morning desktoppers
[07:06] <duflu> Salut didrocks, oSoMoN
[07:09] <oSoMoN> salut duflu, didrocks
[07:16] <didrocks> hey duflu, salut oSoMoN
[08:02] <Laney> yo
[08:03] <willcooke> morning Laney
[08:03] <Laney> what up willcooke
[08:03] <Laney> quite moist here today
[08:04] <didrocks> morning Laney, willcooke
[08:04] <willcooke> Dry here atm, but windy
[08:04] <Laney> yo didrocks
[08:04] <Laney> how's it going?
[08:04] <willcooke> Did you see the big storm headed k_envandine's way?  Could affect his flights to BRU
[08:04] <Laney> nest says it's 18.5, wonder if we'll get to the end of the month without it coming on
[08:04] <didrocks> same old, same old
[08:04] <Laney> oh yeah I heard about something like that on the radio
[08:06] <seb128> good morning desktopers!
[08:06] <willcooke> hi seb128 didrocks
[08:07] <seb128> hey willcooke, how are you?
[08:07] <duflu> o/  Laney  willcooke
[08:07] <duflu> And hi seb128
[08:07] <willcooke> afternoon duflu
[08:08] <willcooke> seb128, doing ok!  How about you?
[08:08] <seb128> hey duflu
[08:08] <seb128> willcooke, I'm good :)
[08:13] <oSoMoN> good morning Laney, willcooke, seb128
[08:13] <willcooke> hi oSoMoN!  Good weekend?
[08:13] <Laney> hey oSoMoN seb128 duflu
[08:13] <seb128> lut oSoMoN!
[08:14] <oSoMoN> yeah, I had a very good long week-end
[08:24] <darkxst> so mozjs60 tests pass on i386 when built with sse2
[08:25] <darkxst> is that something we could ship with, or too restrictive for old CPU's?
[08:25] <Laney> Simon is looking at fixing that
[08:26] <Laney> maybe have a word
[08:27] <darkxst> I replied to his comment on my debian bug for it
[08:30] <Laney> there was more in #debian-devel
[08:31] <Laney> think the thought was to skip those tests on i386
[08:33] <darkxst> which could all sorts of fun bugs, but then does anyone use i386 anymore?
[08:33] <darkxst> could cause all sorts
[08:36] <Laney> this discussion should stop being split across multiple channels imho
[08:36]  * Laney prefers to keep it all in #debian-gnome
[08:36] <Laney> thx for looking at the package though :-)
[08:43] <duflu> darkxst, a noticeable portion of bug reports still seem to be i386. But any i386 that lacks SSE2 support would be from last century. So Pentium 4 and the original Atom from this century are good. Only P3 and earlier would be a problem. And I think that's probably OK because Mesa dropped support for such GPUs around version 8 (IIRC)
[10:25] <Trevinho> Laney: hey, the infos for the FFe are enough now?
[10:25] <oSoMoN> kenvandine, it looks like I can't submit a MR from a +junk branch, but this might interest you: https://bazaar.launchpad.net/~osomon/+junk/thunderbird-snap-fix-theming/revision/13
[10:26] <Laney> Trevinho: Looks good to me, secretly hoping sil2100 or someone else looks (I don't really like reviewing desktop ones)
[10:26] <oSoMoN> I can't really confirm this works as expected, as when I run the snap I get only a completely black window, and lots of "[GFX1-]: Failed to lock new back buffer."
[10:27] <sil2100> I can has look
[10:27] <Laney> sil2100: (gnome-terminal)
[10:28] <Laney> 💋
[10:28] <ricotz> hey desktopers
[10:29] <oSoMoN> hey ricotz
[10:29] <oSoMoN> kenvandine, nevermind my last msg, the black window was due to not manually connecting the browser-sandbox interface
[10:30] <ricotz> oSoMoN, hi, I started to look at lo 6.1.1~rc2
[10:32] <oSoMoN> cool
[10:47] <Trevinho> sil2100: thanks, also if you can de-queue nautilus (SRU) again... :)
[10:56] <Trevinho> yay thanks sil2100... Laney your turn on packaging now :)
[10:56] <Trevinho> [now => when you've time, ofc]
[10:57] <Laney> subscribing ~ubuntu-sponsors in the meantime would be a smooth move
[11:14] <Trevinho> oSoMoN: ah, since I read that on TB... Is there a way you guys can make sure the launcher emblems to work within the snap?
[11:14] <Trevinho> oSoMoN: if it's using libappindicator it should go natively, but if it's not the case, you need to go for checking for a SNAP env end then use a different naming..
[11:14] <seb128> Trevinho, was it you who was talking to GNOME about getting a common API for those?
[11:15] <oSoMoN> Trevinho, ack, can we have a look at that together next week?
[11:15] <Trevinho> something like this https://code.launchpad.net/~3v1n0/libappindicator/snap-prefix-ignore-well-known-dirs/+merge/317330
[11:15] <Trevinho> seb128: yep, I had a talk with hadass on that, they were saying to use notifications api... But had no time to look at that in details yet
[11:16] <seb128> k, no hurry, our current solutions works :)
[11:16] <oSoMoN> Trevinho, the snap as it is today is built from a binary build downloaded from mozilla's servers, so no patching, unless we change it to build from source
[11:16] <Trevinho> mh, I see.. is that feature on ubuntu supported via patching if you know that?
[11:17] <Trevinho> [the change needed was this thoug] https://bazaar.launchpad.net/~indicator-applet-developers/libappindicator/trunk.16.10/revision/277
[11:17] <Trevinho> and in any case I guess they can accept this upstream in case
[11:19] <oSoMoN> I don't know
[11:19] <seb128> chrisccoulson is the one who knows about those details
[12:10] <Trevinho> jibel: hey, any update on https://code.launchpad.net/~3v1n0/ubiquity/+git/ubiquity/+merge/354135 ?
[12:36] <jibel> Trevinho, i'll have a look, sorry for the delay
[12:38] <Trevinho> jibel: no worries
[12:39] <jibel> Trevinho, to answer your question apparently ubuntu-artwork is needed for the distributor-logo that is displayed in partitioning step.
[12:40] <Trevinho> mh, ok... Well, this enforces even more my proposal of refactor that so that it depends on actual graphics though
[13:01] <k_alam> seb128: Latest merge for gnome-screenshot 3.30 tries to remove both the unity header patches and the save dialog......Can we discuss about this ?
[13:03] <seb128> k_alam, you want to talk to andyrock since he's the one working on that
[13:03] <oSoMoN> didrocks, I see in https://github.com/ubuntu/yaru/blob/master/build-helpers/prepare-build-snap that yaru's CI builds and pushes gtk-common-themes to edge, it would be useful if we had separate CI for gtk-common-themes, so maybe a separate branch for use by yaru's CI?
[13:04] <k_alam> seb128: Right. But is he here?
[13:04] <didrocks> oSoMoN: hum, unsure about what you are asking? We want each gtk-common-themes builds to refresh itself (and take previous yaru content), and we want as well every Yaru push to refresh gtk-common-themes
[13:04] <seb128> k_alam, type "and<TAB>" ;)
[13:05] <k_alam> Nice trick...:)
[13:06] <seb128> k_alam, btw is anyone porting unity-control-center away from the deprecated libnm libraries?
[13:07] <oSoMoN> didrocks, right, but today there's no automated build of gtk-common-themes when we push updates to the repo, only when there are yaru updates
[13:07] <didrocks> oSoMoN: that doesn't prevent gtk-common-theme to have its own Travis CI if that's your question
[13:07] <didrocks> oSoMoN: yeah, James was supposed to add one
[13:08] <k_alam> seb128: I already ported it but none to review.......It may require some changes since then
[13:08] <k_alam> https://code.launchpad.net/~khurshid-alam/unity-control-center/libnm-port-test
[13:08] <k_alam> https://code.launchpad.net/~khurshid-alam/unity-settings-daemon/libnm-port-test
[13:08] <oSoMoN> didrocks, and last week I added a yaru part to gtk-common-themes that will the theme from source, unsure if that renders your script obsolete?
[13:09] <seb128> k_alam, just land it :)
[13:09] <didrocks> oSoMoN: hum, wasn't the snapcraft pulling latest Yaru snap and unpack it?
[13:09] <oSoMoN> didrocks, ah I didn't know James was going to do it, Ken asked me to get it working asap so I'll get it
[13:09] <didrocks> oSoMoN: so, you renamed the communitheme entry?
[13:10] <oSoMoN> didrocks, gtk-common-themes didn't include yaru, only communitheme
[13:10] <didrocks> ah no, it's in addition
[13:10] <didrocks> phew ;)
[13:10] <didrocks> (otherwise, this would have broken bionic user :p)
[13:10] <didrocks> oSoMoN: so, basically, this looks good (I just opened the definition)
[13:10] <oSoMoN> hopefully we would have noticed by pushing to candidate first and testing :)
[13:10] <didrocks> well, you would have needed to test on bionic
[13:11] <didrocks> which is something you may have missed
[13:11] <didrocks> or that I would have :)
[13:11] <didrocks> so
[13:11] <didrocks> it looks good
[13:11] <didrocks> the issue is…
[13:11] <oSoMoN> no :) I'm still on bionic (don't tell anyone!)
[13:11] <didrocks> lalala
[13:11] <didrocks> there is no way in snapcraft.io to build from multiple branches as a trigger
[13:11] <didrocks> like telling "please trigger a build of gtk-common-themes if gtk-common-theme branch or Yaru branch changes"
[13:12] <didrocks> this is why I'm pushing from "the other side", leaving the rest untouched
[13:12] <didrocks> (but it needs to udpate as well to udpate the Yaru part now)
[13:12] <oSoMoN> ah
[13:12] <didrocks> also, you can only build from master in snapcraft.io
[13:12] <didrocks> which is a pb when people submit a PR
[13:13] <didrocks> which is where I want to build a snap for communitheme AND gtk-common-themes
[13:13] <didrocks> (if that makes sense to you?)
[13:13] <didrocks> like on: https://github.com/ubuntu/yaru/pull/452#issuecomment-389563519
[13:13] <gitbot> ubuntu issue (Pull request) 452 in yaru "Bright border color in maximized window mode" [Open]
[13:14] <didrocks> (see, there are 2 snaps generated for each PR, one being the snap for bionic, the other, being gtk-common-themes)
[13:14] <oSoMoN> I see
[13:14] <didrocks> so yeah, you are using snapcraft.io as CI, I guess?
[13:15] <oSoMoN> I don't think there's anything set up for gtk-common-themes yet, so we can use whatever we please
[13:15] <didrocks> the issue with Travis, is that it's amd64 only
[13:16] <didrocks> snapcraft.io has the advantage to build on all arch
[13:16] <didrocks> OR
[13:16] <didrocks> we can use launchpad, build on all arch
[13:16] <didrocks> and have a custom trigger (a service running somewhere), to trigger the build
[13:16] <didrocks> using launchpad API
[13:16] <didrocks> I just went with Travis as the snap is arch: all
[13:16] <oSoMoN> and the custom trigger could simply be a launchpad mirror of the github repo, and build with each new revision, which is already supported by LP
[13:17] <didrocks> yeah, but then, we are back to the same issue
[13:17] <didrocks> there are 2 branches
[13:17] <didrocks> (well, 2 repos)
[13:17] <didrocks> gtk-common-themes (or even other themes that we want to monitor)
[13:17] <didrocks> and yaru
[13:18] <oSoMoN> well we can keep the existing infrastructure for yaru, and add this other thing for gtk-common-themes, and the two things would live their own life independently
[13:18] <didrocks> yeah
[13:18] <didrocks> let me show you the definition for gtk-common-themes
[13:19] <didrocks> oSoMoN: https://github.com/ubuntu/yaru/blob/master/build-helpers/gtk-common-themes-parts.yaml
[13:19] <didrocks> as you can see, I'm fetching gtk-common-themes, strip the "communitheme" assets and have that in a separate part
[13:19] <didrocks> from the edge channel
[13:19] <didrocks> I guess something similar for Yaru is needed right now (stripping it, rebuilding)
[13:20] <oSoMoN> yes
[13:20] <oSoMoN> otherwise it will pick up the wrong version
[13:20] <k_alam> seb128: Everything works except for vlan (for new connection) we still need nm-editor......g-c-c is supposed to get new editor, but it never happened....  But if everyone agrees that we don't need vlan gui in u-c-c then I can propose a merge.
[13:21] <didrocks> oSoMoN: a little bit more complex for Yaru, as there is no snap, it needs to rebuild it when doing the PR
[13:21] <didrocks> or Travis CI stuff
[13:21] <didrocks> are you interested in doing it or should I?
[13:21] <oSoMoN> didrocks, I can do it
[13:21] <oSoMoN> I'll get my hands dirty after the weekly meeting, gotta prepare notes now
[13:21] <didrocks> sweet! :)
[13:22] <didrocks> thanks oSoMoN, good that you triggered that discussion
[13:22] <oSoMoN> yw!
[13:22] <didrocks> oSoMoN: I see that gtk-common-themes is arch: all btw, so you can use either snapcraft.io (but it will be restricted to master branch only) or Travis CI and I can help you there ;)
[13:29] <jibel> Hi everyone the weekly desktop meeting is about to start
[13:30] <didrocks> hey hey
[13:30] <jibel> #startmeeting Desktop Team Weekly Meeting - 2018-09-11
[13:30] <meetingology> Meeting started Tue Sep 11 13:30:25 2018 UTC.  The chair is jibel. Information about MeetBot at http://wiki.ubuntu.com/meetingology.
[13:30] <meetingology> Available commands: action commands idea info link nick
[13:30] <seb128> _o/
[13:30] <Nafallo> o/
[13:30] <jibel> Roll call:  andyrock, dgadomski, didrocks, duflu (out), heber, jbicha, jamesh (out), jibel, kenvandine (out), laney, oSoMoN, seb128, tkamppeter, trevinho, robert_ancell (out)
[13:30] <andyrock> o/
[13:31] <seb128> k_alam, nm-connection-editor is still available for those who need it, anyway your call but we are likely going to want to drop those deprecated libs next cycle, Debian already did
[13:31] <Laney> yo
[13:31] <jibel> Ok lets start
[13:31] <jibel> #topic andyrock
[13:31] <oSoMoN> o/
[13:31] <andyrock> - Update gnome-screenshot to 3.30.0 in cosmic (MP ready to be reviewed)
[13:31] <andyrock> - Update gnome-software to 3.30 in unstable (MP ready to be reviewed)
[13:31] <andyrock> - Add XUbuntuCancel to gnome-software (MP ready to be reviewed)
[13:31] <andyrock> - Propose branch to SRU fix for LP: #1787553
[13:31] <andyrock> - Working on LP: #1786344 (kind of blocked because the original bug was about
[13:31] <andyrock>   ibus not working at all on livecd, now it's more about using ibus by default)
[13:32] <jibel> Anything else andyrock ?
[13:32] <andyrock> nope
[13:32] <andyrock> EOW
[13:32] <jibel> Thank you
[13:32] <jibel> #topic dgadomski
[13:32] <dgadomski> hey
[13:32] <dgadomski> * reported bug #1789920 blocking verification of bug #1749289
[13:32] <dgadomski> * working on upstreaming bionic fixes (systemd done, mutter to go) for bug #1755490
[13:32] <dgadomski> * working on bug #1762391
[13:33] <dgadomski> eof
[13:33] <jibel> Thanks dgadomski
[13:33] <jibel> #topic didrocks
[13:33] <didrocks> * GNOME Updates:
[13:33] <didrocks>   - gnome-settings-daemon, gnome-control-center, evince, adwaita-icon-theme, gnome-shell (debian/ubuntu), mutter (debian, synced to ubuntu), gnome-shell-pomodoro sync, removal of other non compatible extensions, gdm3, gnome-session, getting-started-gnome-docs.
[13:33] <didrocks>   - network-mangager-applet merge from debian + update (with FFe acked). Backported some network-manager fix for n-m-applet FTBFS. Fix dh_translation (https://launchpad.net/ubuntu/+source/pkgbinarymangler/142) issue causing n-m-applet to FTBFS.
[13:33] <didrocks> * Yaru:
[13:33] <didrocks>   - Updated in cosmic
[13:33] <didrocks>   - Fix CI build issue due to snapcraft
[13:33] <didrocks>   - sync with core team, deciding on dark theme and gtk2 (pushing to have that merge before UIF)
[13:33] <didrocks> * Misc:
[13:33] <didrocks>  - lot of Marco's sponsoring for XUbuntuCancel on gnome-shell, gnome-calculator, mutter…
[13:33] <didrocks>  - installer next discussion and investigation
[13:33] <didrocks>  - spnosored GNOME Shell fix for bug #1790996
[13:33] <didrocks> .
[13:34] <jibel> Thank you didrocks
[13:34] <jibel> #topic duflu
[13:34] <jibel> * Gnome Shell performance:
[13:35] <jibel>   - NEW: Most of the week spent analyzing, learning and fixing app grid scrolling performance:
[13:35] <jibel>     . https://gitlab.gnome.org/GNOME/gnome-shell/merge_requests/224
[13:35] <gitbot> GNOME issue (Merge request) 224 in gnome-shell "st-box-layout: Avoid fullscreen relayout on scroll" [1. Feature, 5. St, 6. Performance, Opened]
[13:35] <jibel>     . Caveat: More efficiency does not necessarily reduce CPU usage. First it enables you to render at 60 FPS consistently where things might have been 30 FPS before. So the increased frame rate may cancel out any expected reduction in CPU load. But still, the higher frame rate looks sublime.
[13:35] <jibel>     . Bonus: the same bottleneck seemingly applies to dragging windows and the icon flock animation so I'll look into those soon.
[13:35] <jibel>   - BLOCKED waiting on upstream: https://gitlab.gnome.org/vanvugt/mutter/tree/super-smooth-v2 is effectively blocked due to multiple dependencies. And I can't rebase on multiple branches at once without making a mess. Awaiting either of these to land:
[13:35] <jibel>     . https://gitlab.gnome.org/GNOME/mutter/merge_requests/216
[13:35] <gitbot> GNOME issue (Merge request) 216 in mutter "cogl-winsys-glx: Fix frame notification race/leak" [1. Bug, 5. Cogl, Opened]
[13:35] <jibel>     . https://gitlab.gnome.org/GNOME/mutter/merge_requests/171
[13:35] <gitbot> GNOME issue (Merge request) 171 in mutter "Sync to the hardware refresh rate, not just 60.00Hz [performance]" [1. Feature, 5. Backend: Native, Opened]
[13:35] <jibel>   - BLOCKED waiting on upstream: https://gitlab.gnome.org/GNOME/mutter/merge_requests/73 due to multiple dependencies:
[13:35] <jibel>     . https://gitlab.gnome.org/GNOME/mutter/merge_requests/119
[13:35] <gitbot> GNOME issue (Merge request) 73 in mutter "WIP: renderer-native: CRTCs now hold front buffer refs." [1. Bug, 5. Backend: Native, 5. Performance, Opened]
[13:35] <gitbot> GNOME issue (Merge request) 119 in mutter "renderer-native: Reference count front buffers." [1. Feature, 5. Backend: Native, Opened]
[13:35] <jibel>     . https://gitlab.gnome.org/GNOME/mutter/merge_requests/140 (which is blocked waiting for !171 to land)
[13:35] <gitbot> GNOME issue (Merge request) 140 in mutter "WIP: renderer-native: Simplify swap notification" [1. Feature, 5. Backend: Native, Opened]
[13:36] <jibel>   - FUN FACT: A second person is reporting exactly 37.5% (same as my measurement) reduction in CPU usage from geometric (GPU-less) picking support (completed last week):
[13:36] <jibel>     . https://gitlab.gnome.org/GNOME/mutter/merge_requests/189
[13:36] <gitbot> GNOME issue (Merge request) 189 in mutter "Geometric (GPU-less) picking" [1. Feature, 5. Clutter, 5. Performance, Opened]
[13:36] <jibel> * Bug management:
[13:36] <jibel>   - Great progress continues for Firefox and Thunderbird this week.
[13:36] <jibel>   - Ubiquity and Nautilus backlogs are still quite bad and stopped improving a while ago.
[13:36] <jibel>   - https://docs.google.com/spreadsheets/d/e/2PACX-1vRDHPxGBHqM6XkT_S8ggtYfD0xchKSUD_z9PopNVE3G1rU05fVSnxDGcDsEstl7gu7N-tzCU6mLUp2V/pubchart?oid=254968654&format=interactive
[13:36] <jibel> EOW
[13:37] <jibel> #topic heber
[13:37] <heber> hey!
[13:37] <heber> * Submitted fixes for utah tests in MP comments, now it is merged.
[13:37] <heber> * Triage bugs reported as part of nvidia call for testing
[13:37] <heber> * Create a doc with a summary of nvidia testing
[13:37] <heber> * Clean jenkins jobs that are not used/needed anymore
[13:37] <heber> EOF
[13:37] <jibel> Thank you heber
[13:38] <jibel> #topic jamesh
[13:38] <jibel> snapcraft debug symbols:
[13:38] <jibel> * This is mostly stalled on feedback I hope to get from the Snapcraft folks next week in Brussels.
[13:38] <jibel> * I put together some sample code that can take the set of ELF build
[13:38] <jibel> IDs for files we don't have debug info and use the python-apt API to
[13:38] <jibel> determine which packages hold the debug info based on the "Build-Ids"
[13:38] <jibel> header.  I haven't haven't worked out a good way to integrate this,
[13:38] <jibel> since the existing stage-packages code just extracts all the
[13:38] <jibel> downloaded debs.  Perhaps this needs to be a separate apt cache then.
[13:38] <jibel> snapped gnome-shell search providers:
[13:38] <jibel> * There are two parts to this: having a snap install a
[13:38] <jibel> ".search-provider.ini" file, and allowing snaps to provide bus activatable services.
[13:38] <jibel> * I've been working on the second problem, since it is useful outside of search providers.
[13:39] <jibel> * mvo had made some dbus activation PRs in the past, but they weren't merged.  I've been using that as one reference for how to implement things.
[13:39] <jibel> * I've started by trying to extend snapd's daemon support to cover
[13:39] <jibel> daemons running on a systemd user session.  The plan is to write D-Bus
[13:39] <jibel> service files that will have systemd activate the service as a first
[13:39] <jibel> option with executing "snap run ..." for old non-systemd systems.
[13:39] <jibel> EOF
[13:39] <jibel> #topic jbicha
[13:40] <jibel> next
[13:40] <jbicha> • Uploaded lots of universe packages with language packs enabled
[13:40] <jbicha> • More GNOME 3.30 packaging, testing, bug reports, etc.
[13:40] <jibel> sorrt
[13:40] <jibel> y
[13:40] <jbicha> • Uploaded new Cantarell to unstable/cosmic. Did a bunch of fontmake related sponsoring.
[13:40] <jbicha> • Enabled build tests for gtksourceview packages & Debian's nautilus 3.30
[13:40] <jbicha> • Rejected bug 1789693
[13:40] <jbicha> • Figured out https://gitlab.gnome.org/GNOME/pango/issues/287 with help from mwhudson
[13:40] <gitbot> GNOME issue 287 in pango "ruby-gnome2 i386 test failure with pango built with meson" [Bugzilla, Opened]
[13:40] <jbicha> • Poked Debian ftp-masters to accept mozjs60 and pipewire from NEW
[13:40] <jbicha> • libdazzle update is blocked by https://gitlab.gnome.org/GNOME/libdazzle/issues/25
[13:40] <gitbot> GNOME issue 25 in libdazzle "test-application failure with 3.30.0" [Opened]
[13:40] <jbicha> 👣
[13:41] <seb128> jbicha, did you try to follow up with upstream on the libdazzle issue or is that stalled?
[13:41] <jbicha> it's stalled because upstream couldn't reproduce
[13:41] <seb128> :/
[13:42] <seb128> I guess someone needs to investigate, at least bisect the commit that added the regression
[13:42] <jbicha> I found the commit that broke stuff for us
[13:42] <jbicha> but I didn't know if it was a good idea to just revert it since I didn't really understand the consequences of that
[13:42] <seb128> maybe comment upstream to say that?
[13:42] <jbicha> "The test passes for me after reverting ace2d7a1"
[13:42] <seb128> that's a gentle reminder :)
[13:43] <seb128> ah
[13:43] <seb128> maybe comment asking if there would be any problem reverting it?
[13:44] <seb128> anyway
[13:44] <seb128> let's move on from that topic
[13:44] <seb128> can be dicussed after the meeting
[13:44] <jibel> thanks jbicha
[13:44] <jibel> #topic jibel
[13:44] <jibel> - Ongoing work on iso size indicators
[13:44] <jibel> - Reviewed and tested fixes for iso tests. Tests are back to green.
[13:44] <jibel> - Report on multi GPU systems.
[13:44] <jibel> - Maintenance of the QA testing infrastructure.
[13:44] <jibel> done
[13:45] <jibel> #topic kenvandine
[13:45] <jibel> * gtk-common-themes:
[13:45] <jibel>   - Fixed snapcraft deprecations
[13:45] <jibel>   - Added elementary gtk and icon themes
[13:45] <jibel> * snap-store:
[13:45] <jibel>   - Merged the latest upstream changes from gnome-software
[13:45] <jibel>   - Investigated launching snaps from the confined snap-store snap
[13:45] <jibel>   - Removed the launch button until we implement a way to launch the snaps via a dbus service
[13:45] <jibel> * Updated seeded snaps to 3.30
[13:45] <jibel> EOW
[13:45] <jibel> #topic laney
[13:46] <Laney> • fixed notify-osd tests for new dconf pkg
[13:46] <Laney> • updated dconf
[13:46] <Laney> • gtk update to 3.24
[13:46] <Laney> ∘ fixed autopilot-gtk tests for that, they assumed Ambiance was the default theme
[13:46] <Laney> ∘ Merged it again for a bug fix affecting vte
[13:46] <Laney> • new gdm failed to fall back to xorg from wayland for me (black screen), investigated that & reverted (in cosmic) the commit which broke it. Ray says it's going to be fixed in an upcoming MR
[13:46] <Laney> ∘ that's https://gitlab.gnome.org/GNOME/gdm/issues/419
[13:46] <gitbot> GNOME issue 419 in gdm "wayland -> X fallback is broken" [1. Bug, Opened]
[13:46] <Laney> • Looked at glib2.0 test failures from the recent SRU, retried a bunch & reported one upstream
[13:46] <Laney> ∘ https://gitlab.gnome.org/GNOME/glib/issues/1518
[13:46] <gitbot> GNOME issue 1518 in glib "/network-monitor/create-in-thread fails in (LXC) containers on glib-2-56" [Giomodule, Gnetworkmonitor, Test Failure, Opened]
[13:46] <Laney> ∘ then figured out why it was broken and submitted https://gitlab.gnome.org/GNOME/glib/merge_requests/313 to fix that
[13:46] <gitbot> GNOME issue (Merge request) 313 in glib "Some cleanups for the netlink GNetworkMonitor" [1. Bug, Gnetworkmonitor, Test Failure, Opened]
[13:46] <Laney> ∘ that SRU got promoted now
[13:46] <Laney> • Did a bit of GLib reviewing too
[13:46] <Laney> • Poked around on Debian porter boxes and talking to porter people to find out why one test started failing on armhf
[13:46] <Laney> ∘ https://salsa.debian.org/gnome-team/glib/commit/46ddaacdeddafc1df3e665432396fa523197024c, sponsored by jbicha: thanks! 💃
[13:46] <Laney> • sponsored mutter, gnome-shell in Debian & uploaded gnome-shell merge to cosmic
[13:46] <Laney> • New gdk-pixbuf, required some iterations to get tests passing on all arches but seems good now
[13:46] <Laney> • Just uploaded new gvfs to Debian, have a merged pkg, needs FFe to be approved
[13:46] <Laney> 🍂
[13:47] <jibel> Thank you Laney
[13:47] <Laney> you are welcome!
[13:47] <jibel> #topic oSoMoN
[13:47] <oSoMoN> hey
[13:47] <oSoMoN> This was a short week for me as I was off yesterday.
[13:47] <oSoMoN> • chromium
[13:47] <oSoMoN>   ∘ filed and investigated bug #1791307
[13:47] <oSoMoN>   ∘ continued focus on transitioning to a snap-only chromium
[13:47] <oSoMoN>   ∘ chromium 69 was promoted to the stable channel this week
[13:47] <oSoMoN>     - pushed update to cosmic
[13:47] <oSoMoN>     - prepared updates for xenial and bionic, they should hit -security sometime this week
[13:47] <oSoMoN>     - built, tested and published an update to the chromium-ffmpeg snap that third-party browser snaps use, also updated the corresponding documentation (https://forum.snapcraft.io/t/using-chromium-ffmpeg-in-third-party-browser-snaps/6545)
[13:47] <oSoMoN>   ∘ updated dev to 70.0.3538.9
[13:47] <oSoMoN>   ∘ reported a snapcraft regression in 2.43 that breaks the chromium snaps built from debs: https://forum.snapcraft.io/t/call-for-testing-snapcraft-2-43/7024/19
[13:47] <oSoMoN> • libreoffice
[13:47] <oSoMoN>   ∘ 6.1.0 migrated to cosmic
[13:47] <oSoMoN>   ∘ prompted by Rico, I discussed the builder disk space situation with launchpad folks and it appears we should be able to merge the two source packages now (libreoffice and libreoffice-l10n), I filed bug #1791943 to track the task (and a trello card: https://trello.com/c/U8LUIYN0/238-merge-libreoffice-and-libreoffice-l10n-source-packages)
[13:47] <oSoMoN> • snaps
[13:47] <oSoMoN>   ∘ filed https://github.com/snapcrafters/gtk-common-themes/issues/7 (Add Yaru theme) and submitted https://github.com/snapcrafters/gtk-common-themes/pull/8, merged
[13:47] <gitbot> snapcrafters issue 7 in gtk-common-themes "Add Yaru theme" [Closed]
[13:47] <gitbot> snapcrafters issue (Pull request) 8 in gtk-common-themes "Add Yaru theme (fixes #7)." [Closed]
[13:47] <oSoMoN>   ∘ discussed setting up separate CI for gtk-common-themes (currently bound to yaru's CI)
[13:47] <oSoMoN>   ∘ confirmed https://github.com/ubuntu/snapcraft-desktop-helpers/issues/152
[13:47] <gitbot> ubuntu issue 152 in snapcraft-desktop-helpers "Gtk3 application crashes due to "specified class size for type 'IBusIMContext' is smaller than the parent type's 'GtkIMContext' class size"" [Open]
[13:48] <oSoMoN>   ∘ looked into issues reported with the thunderbird snap and submitted a fix for the theming (https://forum.snapcraft.io/t/thunderbird-version/7146/16)
[13:48] <oSoMoN> 🐌
[13:48] <jibel> ^ what is it supposed to mean?
[13:48] <jibel> #topic seb128
[13:48] <oSoMoN> that I'm slow these days
[13:48] <jibel> :)
[13:49] <seb128> • desktop updates in debian & merges (gnome-desktop3, gnome-calculator, gnome-calendar, gnome-power-manager)
[13:49] <seb128> • desktop syncs (gnome-color-manager, gnome-maps, gnome-multi-writer, lcsm2)
[13:49] <seb128> • desktop scrits, added a report for the foundations set, some component updates, include git merge-requests
[13:49] <seb128> • reviewed some translation updates
[13:49] <seb128> • trello board review
[13:49] <seb128> • some preparation for topics to be discussed in brussels

[13:49] <jibel> Thanks seb128
[13:49] <jibel> #topic tkamppeter
[13:49] <tkamppeter> - ghostscript: FFe to update to version 9.24. Upstream is highly recommending this version for security reasons.
[13:49] <tkamppeter> - cpdb-libs: Set up access to porter boxes to debug autopkg test on arm64.
[13:49] <tkamppeter> - Google Summer of Code 2018: Continuing mentoring the students for completing their work.
[13:49] <tkamppeter> - Bugs.
[13:50] <jibel> Thank you tkamppeter
[13:50] <jibel> #topic trevinho
[13:51] <Trevinho> · Released libwnck 3.30.0 (and some maintancence)
[13:51] <Trevinho> · vte update to 0.54 and gnome-terminal to 3.30.0 (debian and ubuntu)
[13:51] <Trevinho>   - https://salsa.debian.org/gnome-team/vte2.91/merge_requests/3
[13:51] <Trevinho>   - https://salsa.debian.org/gnome-team/gnome-terminal/merge_requests/3
[13:51] <Trevinho>   - [FFe] https://bugs.launchpad.net/ubuntu/+source/gnome-terminal/+bug/1791373
[13:51] <Trevinho>   - [bileto ticket] https://bileto.ubuntu.com/#/ticket/3410 (git branches linked)
[13:51] <Trevinho> · Some terminal regex fun (since I was there...)
[13:51] <Trevinho>   - https://gitlab.gnome.org/3v1n0/gnome-terminal/merge_requests/1
[13:51] <gitbot> GNOME issue (Merge request) 3 in vte2.91 "debian/master: Update to 0.54.0" [Merged]
[13:51] <gitbot> GNOME issue (Merge request) 3 in gnome-terminal "debian/master: update to 3.30.0" [Merged]
[13:51] <gitbot> gnome-terminal issue (Merge request) 1 in gnome-terminal "Parse gitlab shortlinks" [Opened]
[13:51] <Trevinho> · Reviewed Andrea's XUbuntuCancel proposal for gnome-software
[13:51] <Trevinho> · Some debugging on https://gitlab.gnome.org/GNOME/mutter/issues/15
[13:51] <Trevinho> · Prepared SRU vcs branches for XUbuntuCancel support
[13:51] <Trevinho>   + Blocked on waiting for glib SRU (otherwise calculator might crash and nautilus)
[13:51] <Trevinho> · Added support to gettext to Javascript ES6 template strings (was causing issues to g-s)
[13:51] <Trevinho>   - https://savannah.gnu.org/bugs/?50920
[13:51] <Trevinho> · Update to gnome-calculator async search provider PR to use proper dbus errors
[13:51] <gitbot> GNOME issue 15 in mutter "gnome-shell crashes if a modal window closes whilst being dragged" [1. Crash, Closed]
[13:51] <Trevinho>   - https://gitlab.gnome.org/GNOME/gnome-calculator/merge_requests/10
[13:51] <Trevinho> 🌞
[13:51] <gitbot> GNOME issue (Merge request) 10 in gnome-calculator "Async and cancellable search provider" [Merged]
[13:51] <jibel> Thanks Trevinho
[13:51] <jibel> #topic robert_ancell
[13:52] <jibel> - Update gnome-software in cosmic to 3.30.0.
[13:52] <jibel> - Release wide scope search support in gnome-software (cosmic), prepare branch for bionic SRU.
[13:52] <jibel> - Rebased accountsservice patches for 0.6.50.
[13:52] <jibel> - Fixed build warnings in accountsservice.
[13:52] <jibel> - Fix gdbus-codegen from generating code that causes compile warnings.
[13:52] <jibel> - Setup CI configuration for accountsservice.
[13:52] <jibel> - Fix GTK+ bug with toolbars not handling the 'hexpand' property.
[13:52] <jibel> - Land various improvements to region panel in Settings.
[13:53] <jibel> - GNOME Software stakeholders meeting.
[13:53] <jibel> - Preparing for Brussels sprint.
[13:53] <jibel> eow
[13:53] <jibel> #topic rls-bb/cc-bugs
[13:53] <jibel> seb128, do you want to drive the bug review?
[13:53] <Trevinho> seb the bug review master
[13:54] <seb128> sure
[13:54] <seb128> http://reqorts.qa.ubuntu.com/reports/rls-mgr/rls-bb-incoming-bug-tasks.html
[13:54] <seb128> nothing for desktop on that one
[13:54] <seb128> http://reqorts.qa.ubuntu.com/reports/rls-mgr/rls-cc-incoming-bug-tasks.html
[13:54] <seb128> https://bugs.launchpad.net/ubuntu/+source/gnome-initial-setup/+bug/1790146
[13:55] <seb128> we should look at this issues
[13:55] <seb128> andyrock, do you have time to have a look and see what is needed?
[13:55] <didrocks> maybe we need design to provide us some content that can be reused over releases instead of going ad hoc with each release?
[13:56] <seb128> my understanding was that the "what's new" was to be a one time thing
[13:56] <seb128> but it's on the list of topics for brussels anyway
[13:56] <seb128> so we can discuss details there
[13:57] <didrocks> maybe it's a better time, indeed
[13:57] <seb128> :)
[13:57] <seb128> the other one on that list is bug #1791090
[13:57] <seb128> duflu investigated, seems xorg/kernel, i'm going to talk to Timo about it but no need to be in the meeting
[13:57] <seb128> http://reqorts.qa.ubuntu.com/reports/rls-mgr/rls-cc-tracking-bug-tasks.html
[13:57] <seb128> 1 assigned item there, good
[13:58] <seb128> http://reqorts.qa.ubuntu.com/reports/rls-mgr/rls-bb-tracking-bug-tasks.html
[13:58] <seb128> the list didn't seem to have changed much since previous week, I still need to review/try to move things for the backlog
[13:58] <seb128> most are xorg, I need to talk to Timo about those as well
[13:59] <seb128> bug #1789925 is for me, I'm going to assign it
[13:59] <seb128> does anyone want to do bug #1790609?
[13:59] <didrocks> I can do
[13:59] <seb128> thx
[14:00] <seb128> we should be good with bug reviews then
[14:00] <jibel> Thank you seb128
[14:00] <seb128> yw!
[14:00] <jibel> #topic aob
[14:00] <jibel> anyone anything?
[14:01] <jibel> seems not
[14:01] <jibel> Thank you everyone for attending and see you next meeting!
[14:01] <jibel> #endmeeting
[14:01] <didrocks> thanks!
[14:01] <meetingology> Meeting ended Tue Sep 11 14:01:36 2018 UTC.
[14:01] <meetingology> Minutes:        http://ubottu.com/meetingology/logs/ubuntu-desktop/2018/ubuntu-desktop.2018-09-11-13.30.moin.txt
[14:01] <andyrock> seb128: sure
[14:01] <seb128> thanks!
[14:01] <oSoMoN> thanks
[14:01] <seb128> andyrock, thank you
[14:01] <jbicha> jibel: do you have time to look into verifying the bionic evolution/-data-server SRUs?
[14:02] <jibel> jbicha, I'm afraid I won't have time this week
[14:20] <oSoMoN> didrocks, for gtk-common-themes CI, I’ve created https://code.launchpad.net/~ubuntu-desktop/+snap/gtk-common-themes and got a build running, let's see if it succeeds
[14:21] <oSoMoN> (the first build failed because I didn't specify the gnome 3.26 backports PPA)
[14:29] <didrocks> let's see :)
[14:33] <oSoMoN> fatal: unable to access 'https://git.launchpad.net/~ubuntu-desktop/gtk-common-themes/+git/master/': GnuTLS recv error (-110): The TLS connection was non-properly terminated.
[14:33] <oSoMoN> herm
[14:34] <andyrock> seb128: mtp also asked us to drop gnome-online-accounts page in gnome-intial-setup
[14:35] <andyrock> seb128: should I do it?
[14:40] <jbicha> I don't think users are generally too annoyed by entering their log in to multiple apps
[14:40] <jbicha> if people sign in to GOA with Google or Microsoft, it does provide useful integration in the default install (in Nautilus at least)
[14:55] <andyrock> also if we remove the welcome-page what should we show as first page?
[14:56] <seb128> andyrock, unsure, do you know why he wants to remove the online accounts page? where did he ask about that?
[14:56] <seb128> andyrock, we should probably keep that discussion/those changes for next week
[15:24] <andyrock> seb128: kk
[15:26] <andyrock> seb128: https://bugs.launchpad.net/ubuntu/+source/gnome-initial-setup/+bug/1762192/comments/3
[15:30] <Trevinho> sil2100: hey, have you seen my ping about the nautilus SRU? You remember there was that version thing, and then you said to reupload?
[15:31] <Trevinho> sil2100: also if you can check https://code.launchpad.net/~3v1n0/ubuntu/+source/gnome-calculator/+git/gnome-calculator/+merge/354330 where I merged with debian since there are very minor changes on debian/* side, while there's lots of clutter in the diff because of upstream tarball which some generated files.
[15:36] <Trevinho> before I check the code, seb128/didrocks: is there a way to override gsettings by adding something to the previous value (say, in a list, I want to add elements to the defaults).
[15:45] <oSoMoN> kenvandine, almost there with automated builds of gtk-common-themes, just need to sort out a build failure that appears to be an (expected) regression in snapcraft 2.43: https://code.launchpad.net/~ubuntu-desktop/+snap/gtk-common-themes
[15:49] <kenvandine> oSoMoN, i built it locally with 2.43.1
[15:49] <kenvandine> so regression in 2.43 that was fixed in 2.43.1?
[15:49] <kenvandine> oSoMoN, maybe we should move the source to gitlab and use CI there :)
[15:50] <kenvandine> or setup travis ci on github
[15:50] <oSoMoN> too many options :)
[15:50] <kenvandine> indeed
[15:51] <kenvandine> but if we built it in a container we'll get the updated snapcraft
[15:51] <oSoMoN> building locally with 2.43 to see if I can reproduce the failure
[15:53] <kenvandine> during the call for testing for 2.43.1 i built it on bionic
[16:04] <kenvandine> oSoMoN, thanks for the thunderbird fix
[16:04] <kenvandine> it's in edge now
[16:04] <kenvandine> works great
[16:06] <sil2100> Trevinho: hey! I can look at it in a moment ;) It's not my SRU day today so I was busy with other tasks
[16:06] <sil2100> But I'll get to it in a few
[16:06] <Trevinho> sil2100: ah, ok sorry :)
[16:06] <cyphermox> jbicha: ping; have you heard of issues with girepository? it looks to me like the FTBFS of glib-d might be an issue with either the gir for glib2.0 (worked with 1.56, fails starting 1.57) or the parsing of it by ldc
[16:09]  * jbicha forwards the ping to ricotz 😉
[16:10] <jbicha> glib 1.58 did break some stuff but I don't know the details
[16:18] <cyphermox> ack
[16:21] <seb128> Trevinho, not that I know for adding to/removing from a list
[16:41] <oSoMoN> kenvandine, can you please merge https://github.com/snapcrafters/gtk-common-themes/pull/11 ?
[16:41] <gitbot> snapcrafters issue (Pull request) 11 in gtk-common-themes "Use a branch tag instead of a tarball for sassc." [Open]
[16:42] <kenvandine> oSoMoN, oh, i can't merge it
[16:42] <kenvandine> popey, ^^
[17:07] <cyphermox> jbicha: ricotz: is gobject-introspection something that can be rolled-back? looks to me like just rebuilding glib-d against the g-i from bionic (no other changes) makes things happy
[17:11] <jbicha> cyphermox: no, we need gobject-introspection and glib to be the same version and we have some stuff depending on the new glib already
[17:14] <oSoMoN> thanks popey
[17:16] <popey> np
[17:22] <Laney> night!
[17:38] <oSoMoN> kenvandine, the build for gtk-common-themes succeeded, but it fails to upload it to the store: https://code.launchpad.net/~ubuntu-desktop/+snap/gtk-common-themes/+build/330527
[17:40] <kenvandine> oSoMoN, i'll look
[17:41] <oSoMoN> "Store upload failed: Cannot upload new revisions for name=gtk-common-themes series=16"
[17:41] <oSoMoN> not really clear what the problem is
[17:41] <kenvandine> oh, you aren't a collaborator?  maybe that's why
[17:41] <oSoMoN> I've configured the snap recipe to push to beta
[17:41] <oSoMoN> ah, could be
[17:42] <oSoMoN> but then it should have failed much earlier, when I configured the snap recipe to push to the store and it checked my 2fa credentials
[17:42] <kenvandine> oSoMoN, i just added you
[17:42] <kenvandine> yeah, i would have thought
[17:43] <kenvandine> but you weren't a collaborator, so that should have really failed
[17:43] <oSoMoN> kenvandine, thanks, trying the upload again
[17:44] <kenvandine> oSoMoN, it will fail
[17:44] <kenvandine> you haven't accepted the invitation yet
[17:46] <oSoMoN> huh
[17:46] <oSoMoN> now I have
[17:46] <oSoMoN> good that there is a retry button that doesn't involve rebuilding the snap every time
[17:46] <kenvandine> indeed
[17:47] <sil2100> Trevinho: hey! It seems that https://bugs.launchpad.net/ubuntu/+source/nautilus/+bug/1765776 is missing SRU information
[17:48] <kenvandine> oSoMoN, it's working
[17:48] <sil2100> Trevinho: I'm EODing now so I'll take a look at it tomorrow
[17:48] <sil2100> o/
[17:49] <oSoMoN> kenvandine, https://dashboard.snapcraft.io/snaps/gtk-common-themes/revisions/701/, want to publish in the candidate channel and start testing? gotta go prepare dinner
[17:49] <kenvandine> oSoMoN, i'll handle it
[17:49] <kenvandine> oSoMoN, thanks!
[17:50] <oSoMoN> alrighty, good night everyone
[17:50] <willcooke> night all
[20:41] <ricotz> cyphermox, hi, does it even use gobject-introspection? afaics it just uses the plain gir and the glib headers and chokes on them?
[20:54] <cyphermox> ricotz: if I build glib-d against gobject-introspection 1.56, it works. if I use >1.56, it doesn't.
[20:55] <Wimpress> !mumble
[20:55] <cyphermox> I don't know the runes behind that, but it sure sounds like something busted in the gir if you look at the error in glib-d's build log.
[20:58] <ricotz> cyphermox, I don't know what kind of magic is happening there, if it build with glib 1.58 headers and the old gir, it seems like a bug in gir-to-d?
[21:01] <jbicha> ximion: ^
[21:02] <ximion> cyphermox: you could report this as an error at https://github.com/gtkd-developers/gir-to-d
[21:03] <ximion> upstream is very responsive to these issues
[21:03] <ricotz> so again this glib-d and gir-to-d does not rely on any g-i library?
[21:03] <ximion> coincidentially that's also on my todo list, but so far I was tied up with other stuff and the libundead FTBFS
[21:04] <ximion> ricotz: no, they generate source code from the .gir XML files
[21:04] <ricotz> ximion, from the gir only or with support of the c-headers?
[21:04] <ximion> or rather, gir-to-d generates the code, glib-d only contains a bunch of definitions to make a good interface to GLib
[21:05] <ximion> from the gir only, AFAIK no C header parsing is involved anywhere
[21:05] <ximion> (or rather: I am certain no C header gets parsed)
[21:05] <ricotz> ok
[21:06] <ricotz> ximion, the error messages indicate some header parser if i am not mistaken though
[21:07] <ricotz> still this seems better to be addressed to upstream
[21:07] <ricotz> ... gir-to-d
[21:08] <ximion> ricotz: nah, the error is caused by gir-to-d generating invalid D code
[21:08] <ricotz> I see
[21:08] <ximion> that is pretty much confirmed by the previous gir-to-d version creating valid code
[21:08] <ximion> if you look at https://buildd.debian.org/status/package.php?p=glib-d, the arm64 build succeeded, as arm64 was a new architecture in the recent LDC release
[21:09] <ximion> and the only difference between the working arm64 build and the failing amd64 build is a newer gir-to-d (0.15 vs 0.16)
[21:11] <ricotz> allright