15:30 <willcooke> #startmeeting Desktop Team Weekly Meeting - 2017-06-20
15:30 <meetingology> Meeting started Tue Jun 20 15:30:42 2017 UTC.  The chair is willcooke. Information about MeetBot at http://wiki.ubuntu.com/meetingology.
15:30 <meetingology> 
15:30 <meetingology> Available commands: action commands idea info link nick
15:30 <oSoMoN> 🙋
15:31 <willcooke> Roll call: andyrock, dgadomski, didrocks, duflu (out), jbicha, jamesh (out), jibel, heber(out), kenvandine, laney, oSoMoN, seb128, tkamppeter, trevinho, robert_ancell (out)
15:31 <jbicha> ugh, Current topi
15:31 <kenvandine> o/
15:31 <andyrock> hey
15:31 <jibel> hi o/
15:32 <Laney> hi
15:32 <willcooke> jibel, good spot - I'll change the topic before I start next week, and change it back again when I'm done
15:32 <willcooke> s/jibel/jbicha
15:32 <Laney> it happened last week too
15:32 <Laney> just shrink it a bit after the meeting
15:32 <willcooke> kk
15:33 <oSoMoN> URL-shortening the link to trello should do the trick
15:33 <willcooke> door, sec
15:34 <willcooke> back
15:34 <willcooke> not Amazon
15:34 <willcooke> :(
15:34 <willcooke> right, let's go
15:34 <willcooke> #topic andyrock
15:34 <andyrock> # Lost some time trying to understand how to port snapd-glib to trusty (Rober alread did that, next time I'll make sure to ask before)
15:35 * willcooke makes a note that andyrock can be a snapd glib maintainer
15:35 <andyrock> # Working on adding livepatch status gui to update-manager (so list of CVEs + description)
15:35 <andyrock> # eow
15:36 <willcooke> thanks andyrock
15:36 <willcooke> #topic dgadomski
15:36 <dgadomski> hey
15:36 <dgadomski> Trying to find a way to disable WiFi Direct for iwlwifi devices. Looks like some enterprise AP hardware blocks them for security and most probably there's no way of doing that.
15:36 <dgadomski> eof
15:36 <willcooke> thanks dgadomski
15:36 <dgadomski> thanks
15:37 <willcooke> #topic didrocks
15:37 <willcooke> * Migrate unity session from ubuntu-session to a new unity-session package and all related changes in other packages due to systemd session.
15:37 <willcooke> * Snapcraft desktop helper time tracing for possible optimizations. Found that we can shove off 2 or 3s at first start if we rewrite it in Golang, but can't do way more (the longest time is waiting for update-mime-database to proceed which is a multiple seconds process)
15:37 <willcooke> * Worked on theme snaps (HO with seb). Draft some examples on github and wrote a google doc with needed changes and blockers on our existing snapd interface mechanism.
15:37 <willcooke> * some AA work for KDE packages.
15:37 <willcooke> * Merged some ubuntu make contributor PR.
15:37 <willcooke> * Prepared Stuttgart nearby travel for client doing snap demos and workshops.
15:37 <willcooke> * Finish a suite of blog posts to be published really (on ubuntu make vs classic snap experience).
15:37 <willcooke> * Answered on some desktop related questions on the snapcraft forum.
15:37 <willcooke> #topic duflu
15:37 <willcooke> * PulseAudio:
15:37 <willcooke> - SRU of A2DP Bluetooth audio fixes to xenial:
15:37 <willcooke> . Still waiting (4 weeks) to enter proposed (am I missing something?):
15:37 <willcooke> https://launchpad.net/ubuntu/xenial/+queue?queue_state=1&queue_text=pulseaudio
15:37 <willcooke> . The primary SRU bug (out of three) is here: https://bugs.launchpad.net/ubuntu/+source/bluez/+bug/1582213
15:37 <ubot5`> Ubuntu bug 1582213 in pulseaudio (Ubuntu Xenial) "[xenial] Bluetooth device doesn't play any sound in A2DP mode unless set to HSP/HFP first" [High,In progress]
15:37 <willcooke> * BlueZ:
15:38 <willcooke> - Release 5.45 to artful completed. Quicker than expected.
15:38 <willcooke> - Investigated the #1 hottest bluez bug for desktop. Thought I had a fix but came to a surprising conclusion; it's actually invalid: https://bugs.launchpad.net/bugs/1490349
15:38 <ubot5`> Ubuntu bug 1490349 in bluez (Ubuntu) "15:10 and 16.04: bluetoothd "Failed to start discovery: org.bluez.Error.NotReady" after bluetoothd restarted" [High,Invalid]
15:38 <willcooke> * Video acceleration:
15:38 <willcooke> - Got blocked in efforts to use Intel MSDK. Proprietary build dependencies and users having to rebuild their own video players are a roadblock. But hopeful that Intel fix this in future (they say they will).
15:38 <willcooke> - Succeeded elsewhere using a different plan: GStreamer has a VA-API decoder:
15:38 <willcooke> Result: 3% CPU to play 4K 60 FPS video on a Haswell desktop.
15:38 <willcooke> - Even better: 4K H.265 HEVC video is now playable (requires Sky Lake, or Kaby Lake for 10-bit). Until now even modern machines haven't been able to play those properly (software decoding).
15:38 <willcooke> - Documented instructions and future plans: https://wiki.ubuntu.com/IntelQuickSyncVideo
15:38 <willcooke> * Libinput:
15:38 <willcooke> - Spent time trying to find workarounds for non-responsive touchpads under Wayland/libinput. No success yet but did notice artful is using an old libinput release. We should update.
15:38 <willcooke> * Other Audio:
15:38 <willcooke> - Testing fixes for XPS 13 audio: https://bugs.launchpad.net/ubuntu/+source/alsa-driver/+bug/1654448
15:38 <willcooke> * Daily bug maintenance across gnome-shell, bluez, pulseaudio.
15:38 <willcooke> - In particular gnome-shell performance is becoming a common topic:
15:38 <ubot5`> Ubuntu bug 1654448 in linux (Ubuntu) "XPS 13 9360 and 9350, Realtek ALC3246, Headphone audio hiss" [Medium,In progress]
15:38 <willcooke> https://bugs.launchpad.net/ubuntu/+source/gnome-shell/+bugs?field.tag=performance
15:39 <willcooke> #topic jbicha
15:39 <jbicha> • Completed Evolution 3.24.3 transition
15:39 <jbicha> • Fixed several FTBFS resulting from vala 0.36 (although ricotz did most of the prep work, thanks!)
15:39 <jbicha> • Merged NetworkManager and nm-applet 1.8 from Debian unstable (nplan autopkgtest failures keeping it in -proposed)
15:39 <jbicha> • Merged dbus 1.10.10 from Debian stretch
15:39 <jbicha> • Released Tweak Tool 3.25.3
15:39 <jbicha> • Blogged about WebKit for Debian 9 release and new Tweak Tool release
15:39 <jbicha> eof
15:39 <willcooke> thanks jbicha
15:40 <willcooke> #topic jamesh
15:40 <willcooke> * produced a working proof of concept of a strict confined snap
15:40 <willcooke> talking to xdg-desktop-portal and wrote about how others could test it
15:40 <willcooke> on the Snapcraft forum.  No replies so far.  I also tried to restart
15:40 <willcooke> the discussion about per-user mounts, which is also waiting.
15:40 <willcooke> * started the process of upstreaming the snap plugin for
15:40 <willcooke> gnome-builder.  The maintainer is receptive to accepting it, but would
15:40 <willcooke> like some changes to how things are done (let the IDE directly call
15:40 <willcooke> configure/make to enable incremental builds).
15:40 <willcooke> * Started on gnome-software bug #1690280 (allow installation of
15:40 <ubot5`> bug 1690280 in gnome-software (Ubuntu) "Classic confined snaps don't install" [High,In progress] https://launchpad.net/bugs/1690280
15:40 <willcooke> classic snaps).  Using Robert's initial patch, and working on the UI
15:40 <willcooke> side.
15:40 <willcooke> #topic jibel
15:40 <jibel> * Continued work on desktop test plan and completed the part for gnome software
15:40 <jibel> * Investigating introspectability issues of gnome-software.
15:40 <jibel> * Merged job definition for running desktop tests on HW using testflinger.
15:40 <jibel> * Created document with highlights and definitions to start automating desktop tests.
15:40 <jibel> * Started implementing the foundations to run end user tests on gnome shell.
15:40 <jibel> eof
15:40 <willcooke> thanks jibel
15:40 <jibel> yw
15:40 <willcooke> #topic kenvandine
15:40 <kenvandine> * Finished snapping gedit using the gnome-3-24 platform snap, getting it into the store now
15:40 <kenvandine> * Created a topic on the snapcraft forum to discuss which snaps should be maintained by canonical vs community: https://forum.snapcraft.io/t/which-snaps-should-be-supported-by-canonical/1064
15:40 <kenvandine> * Found some issues with using the gnome-3-24 snap when an app (like gedit) needed stage-packages that aren't in the gnome-3-24 snap but also depend on libs in the gnome-3-24.  Those libs would still get duplicated.  I have a workaround to strip out all of those files in the gedit snap, but it's not pretty.  I started a discussion on the snapcraft forum about this.
15:41 <kenvandine> * Continued to followup with security on the gdm MIR, no progress yet but I'll keep nagging them from time to time
15:41 <kenvandine> * Getting started on the LightDM -> GDM transition
15:41 * kenvandine OUT
15:41 <willcooke> thanks kenvandine
15:41 <jbicha> lol
15:42 <willcooke> #topic Laney
15:42 <Laney> sec
15:42 <Laney> NO MORE SEC
15:42 <Laney> • gnome-software/asgen:
15:42 <Laney> ∘ Some prep packaging for 3.25. I said this would be uploaded 'this week' last week, but it hasn't happened yet, sry - still need to poke at a crash I saw
15:42 <Laney> ∘ proposed a MP to asgen to generate type=codec data for gstreamer packages so they show up nicely in gnome-software
15:42 <Laney> ∘ this work should all be good to go now once it's reviewed
15:43 <Laney> • poking at / testing a fix for a glib testsuite problem
15:43 <Laney> • much time discussing libtimezonemap change wrt India and reading about borders and stuff
15:43 <Laney> • some babysitting of autopkgtest runners
15:43 <Laney> • some upstream (GNOME) patch reviews
15:43 <Laney> 🛀
15:43 <willcooke> Is that... is that a bath tub?
15:43 <Laney> STEAMY!
15:43 <willcooke> lol
15:43 <kenvandine> lol
15:43 <Laney> actually it's not warm today
15:43 <willcooke> thanks for you help with libtimezonemap laney
15:43 <Laney> heh
15:44 <Laney> "n" "p"
15:44 <willcooke> #topic oSoMoN
15:44 <oSoMoN> • fixed bug #1697496
15:44 <ubot5`> bug 1697496 in chromium-browser (Ubuntu) "chromium 59.0.3071.86 crashes at startup on x86" [Critical,Fix committed] https://launchpad.net/bugs/1697496
15:44 <oSoMoN> • bisecting to find the upstream revision that caused bug #1696965, getting really close
15:44 <ubot5`> bug 1696965 in chromium-browser (Ubuntu) "First renderer process doesn't render page for chromium 59.0.3071.86 in KVM" [Critical,New] https://launchpad.net/bugs/1696965
15:44 <oSoMoN> • updated chromium beta to 60.0.3112.32, dev to 61.0.3128.3
15:44 <oSoMoN> • working on libreoffice 5.3.3 snap, switched to classic confinement to resolve a number of open issues (triaged all snap-specific bugs), should have something ready for wider testing by tomorrow
15:44 <oSoMoN> ⒺⓄⒻ
15:45 <willcooke> thanks oSoMoN
15:45 <seb128> oSoMoN, is there a tag or something to list the snap specific issues?
15:46 <willcooke> (sorry folks, another meeting going on)
15:46 <oSoMoN> seb128, https://bugs.launchpad.net/ubuntu/+source/libreoffice/+bugs?field.tag=snap
15:46 <seb128> great
15:46 <willcooke> #topic seb128
15:46 <seb128> • tried some co-working spaces, pondering subscribing to one of those
15:46 <seb128> • worked with Didier on detailing how snap themes could work, including writing examples
15:46 <seb128> • replied to a snapcraft-forum discussion about themes with details of what is working today
15:46 <seb128> • demoted a stack of unity components to universe
15:46 <seb128> • talked to the mir team about gtk-mir in artful
15:46 <seb128> • some binNEW reviews
15:46 <seb128> • usual bugs triaging, participated in list discussions, etc
15:46 <seb128> </week>
15:47 <willcooke> thanks seb128
15:47 <willcooke> #topic tkamppeter
15:47 <tkamppeter> - cups-filters: Several crash fixes and stability improvements in cups-browsed.
15:47 <tkamppeter> - CUPS: Crash bug got fixed upstream and feature to identify temporary CUPS queues got added upstream, so requirements to make cups-browsed work together with new CUPS are fulfilled.
15:47 <tkamppeter> - Packaging: Debian opened unstable after the last release, so several printing packages got uploaded there. Synced CUPS, cups-filters, ippusbxd, HPLIP, Gutenprint, and foo2zjs to Ubuntu.
15:47 <tkamppeter> - Google Summer of Code 2017: Student project coordination and mentoring.
15:47 <tkamppeter> - Bugs.
15:48 <willcooke> thanks tkamppeter
15:48 <willcooke> #topic Trevinho
15:48 <Trevinho> · Continued work in mutter and gnome-control-center
15:48 <Trevinho> · Review and landing of ubuntu-themes and compiz branches
15:48 <Trevinho> · Experimenting lighter Headerbar theme
15:48 <Trevinho> · Wrote a fractional scaling hackfest report (http://go.3v1n0.net/GNOME-Fractional-Report)
15:48 <Trevinho> - Got some user feedback, and updating Scaling wiki accordingly
15:48 <Trevinho> · Reviews of some Trello cards
15:48 <Trevinho> · Some improvements to my work environment (wrote some gnome snippets, more automated build environment...)
15:48 <Trevinho> --
15:49 <willcooke> thanks Trevinho
15:49 <willcooke> #topic robert_ancell
15:49 <willcooke> - Rebased xenial gnome-software on 3.20.5
15:49 <willcooke> - SRU snapd-glib 1.13 to zesty, xenial
15:49 <willcooke> - SRU snapd-glib 1.12 to trusty. Required a big patch to support glib 2.40
15:49 <willcooke> - Fix up and resubmit zesty gnome-software SRU
15:49 <willcooke> - Disable LightDM guest session in artful to match security fixes in stable releases (not likely to be fixed soon).
15:49 <seb128> Trevinho, feel free to share the work environment improvement tips :-)
15:49 <willcooke> - Sponsored fixes to dee, hud, unity-greeter
15:49 <willcooke> #topic AOB
15:50 <willcooke> 32 bit images.... I said I would get some indication of number of downloads - haven't got that yet
15:50 <seb128> I didn't find any info about apps which would stop 32 bit support and when
15:51 <gQuigs> I did some 32 bit image data from torrents a while ago.. I can look that up
15:51 <willcooke> Shall we punt it to next week again?
15:51 <seb128> +1
15:51 <willcooke> gQuigs, oh! that would be super useful, thanks
15:53 <willcooke> oki, if no objections I will move it to next week.
15:53 <willcooke> Anyone got any other business?
15:53 * oSoMoN keeps his slang to himself
15:54 <jbicha> jackpot51 tole me he wanted to discuss the encrypted-home thing again, but maybe he stepped out
15:54 <jbicha> he's got patches for review at LP: #1699216
15:54 <ubot5`> Launchpad bug 1699216 in gnome-initial-setup (Ubuntu) "Encrypted home support" [Undecided,Confirmed] https://launchpad.net/bugs/1699216
15:54 <jackpot51> I am here
15:55 <jackpot51> It is ready for testing - gnome-control-center and gnome-initial-setup have patches using a new accountsservice - look at the bug for details
15:55 <jackpot51> It means that the user creation in gnome-initial-setup and gnome-control-center allow you to select encrypted home folder
15:56 <Laney> jackpot51: the change to accountsservice is an API break, right?
15:56 <jackpot51> Yes, it is
15:56 <seb128> jackpot51, hey, what did you think about discussing those changes and gnome-initial-setup on ubuntu-devel@ as I suggested yesterday?
15:56 <jackpot51> A little bit
15:56 <Laney> I don't think we can do that
15:56 <Laney> what about making it a separate method?
15:57 <jackpot51> That is possible, I could make it a seperate function.
15:57 <jackpot51> The old function could become a stub that calls the new function with FALSE for the encrypt_home parameter
15:57 <jackpot51> This is meant to go along with another change - removing OEM mode from ubiquity and using gnome-initial-setup instead
15:58 <Laney> something like that
15:58 <jackpot51> I can probably complete that in an hour. I just need help to merge upstream, which is very important to me
15:59 <seb128> jackpot51, upstream bugs are on https://bugs.freedesktop.org that would be the first step, opening a bug describing what you need there
16:00 <Laney> indeed, having it upstream would be best
16:00 <Laney> the only problem is the --encrypt-home thing is Ubuntu specific?
16:00 <jackpot51> Yes, it appears to be. At the very least, having it upstream in Ubuntu is good
16:01 <jackpot51> Right now I am holding it in a PPA
16:01 <jbicha> that ext4 replacement might be really attractive to other distros when it's ready
16:01 <jackpot51> I agree. And at that point the patch can be offered to the most upstream place
16:01 <seb128> ext4 replacement?
16:02 <jackpot51> So - we are in agreement that such a patch, with the API compatability fixed, would go into upstream ubuntu easily?
16:02 <jackpot51> And may go further upstream if encrypt-home gets more wide support?
16:04 <jbicha> seb128: tyhicks said yesterday he'd like to switch to native ext4 instead of ecryptfs for encrypted home directories, but that's still an early WIP
16:05 <Trevinho> Laney: can you please add your sizes for https://wiki.gnome.org/Projects/GnomeShell/ScaleFactorTests as I'm not sure I wrongly typed it at the office
16:05 <jackpot51> He said the same --encrypt-home flag I am using would be used, right?
16:05 <Trevinho> for the XPS 13 we had
16:05 <jackpot51> Well, I will update my patches on that bug for whomever is interested
16:05 <tyhicks> jackpot51: yes, same adduser flag will continue to work
16:05 <jackpot51> cool
16:06 <seb128> jackpot51, you keep ignoring my hint about discussing it on the list btw so I'm going to stop mentioning it but I'm personally not interested to work on changes that haven't been discussed with the project
16:06 <jackpot51> Which list?
16:07 <seb128> the ones  I mentioned to you yesterday and earlier
16:07 <seb128> devel/desktop
16:08 <willcooke> https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
16:08 <willcooke> https://lists.ubuntu.com/mailman/listinfo/ubuntu-desktop
16:09 <Laney> Trevinho: I have to get a ruler and measure my screen?
16:10 <Trevinho> Laney: yep :-)
16:10 <Trevinho> Laney: but actually I'm mostly interested on what mutter / xrandr outputs
16:11 <Laney> Trevinho: k, what are the commands please?
16:11 <jackpot51> I subscribed seb128, I will wait for confirmation...
16:12 <seb128> jackpot51, great
16:12 <willcooke> kk, anymore for any more?
16:12 <seb128> not me
16:12 <Trevinho> Laney: xrandr --verbose (or without flag)
16:13 <willcooke> oki, I need to get on another meeting, so please carry on if needed
16:13 <willcooke> #endmeeting