15:30 #startmeeting 15:30 Meeting started Tue Jun 6 15:30:33 2017 UTC. The chair is seb128. Information about MeetBot at http://wiki.ubuntu.com/meetingology. 15:30 15:30 Available commands: action commands idea info link nick 15:30 Roll call: andyrock, dgadomski, didrocks, duflu, jbicha, jamesh, jibel/heber, kenvandine, laney, oSoMoN, seb128, tkamppeter, trevinho, robert_ancell (out) 15:30 o/ 15:30 Laney: confirmed, only flavors do it in scripts/casper-bottom/15autologin now 15:30 Oh seb.... it's unmeeeting! :_D 15:30 hey 15:30 nod 15:31 seb128: can you put me at the end of the list? still trying to record the video 15:31 :D 15:31 andyrock, sure 15:31 ok, let's get started 15:31 #topic dgadomski 15:31 dgadomski, hey 15:31 hey 15:31 * debugged a printer issue and reported bug #1695884, shared logs with tkamppeter 15:31 * back to investigating do-release-upgrade conflicting with landscape 15:31 bug 1695884 in cups (Ubuntu) "usb probing malforms Epson TM-T70 label printer output" [Undecided,Incomplete] https://launchpad.net/bugs/1695884 15:31 eof 15:32 thanks dgadomski 15:32 #topic didrocks 15:32 didrocks, hey 15:32 hey 15:32 * Amazon: 15:32 - UI changed drastically with 0.9, read their new user guide and content. 15:32 o/ 15:32 - Make the incoming snap multi-arch (was amd64 only previously). 15:32 - a lot of back and forth (mostly on our side) with amazon. Still blocked though. 15:32 * Snap: debugged theming issues. Found a new broken use case. Added a trello card to improve/hack around multiple themes potential issues. 15:32 o/ 15:32 * Removed unity from ubuntu seed (which was readded back and removed again now ;)). Tested on the ISO and found at least another package to temporary remove (ubuntu-session). We'll need to transition it and introduce an unity session. 15:32 . 15:32 back 15:33 EOF didrocks ? 15:33 yeah, that was my traditional "." :) 15:33 thanks didrocks 15:33 thx, just wanted to be sure 15:33 shall I drive seb128? 15:34 willcooke, what's the # to make you lead with the bot? 15:34 chair 15:34 #chair willcooke 15:34 Current chairs: seb128 willcooke 15:34 Laney, thanks 15:34 willcooke, there you go 15:34 #topic duflu 15:34 * PulseAudio: 15:34 - SRU of A2DP Bluetooth audio fixes to xenial: 15:34 . Minor bug updates but the SRU is not progressing (no sponsors yet) 15:34 . BLOCKED? 15:34 * BlueZ: Took over the 5.45 task since Konrad is busy/blocked: 15:34 - Surprisingly bluez git tags are quite different to their release tarballs of the same version. So many branches and trickery are required. 15:34 - Designed a maintenance plan and constructed lots of git branches/tags: https://git.launchpad.net/~bluetooth/bluez/ 15:34 (will document it soon) 15:34 - Prototyped bluez 5.45 packaging for artful 15:34 - Set up a PPA with 5.45: https://launchpad.net/~bluetooth/+archive/ubuntu/bluez 15:34 - Now trialling bluez 5.45 debs on artful. No problems yet. 15:34 * Video acceleration: 15:34 - Working out test cases (acceptance criteria) 15:34 - Reading lots about how the various architectures are meant to work 15:35 - Learning and reading, reading and learning. 15:35 * Daily bug maintenance across gnome-shell, bluez and pulseaudio. 15:35 * Mir: 15:35 - Canceled the 0.28 series and proposed a new changelog and project setup for 0.27.0 (almost entirely work completed many months ago before the reshuffle). 15:35 #topic jbicha 15:35 • Participated in Ubuntu Docs meeting organized by GunnarHJ 15:35 • Started ML discussion with GNOME Docs team about introducing a Docs String Freeze for improved translations. 15:35 • Worked on getting webkit2gtk 2.16 in to Debian stretch (to follow something like how Ubuntu 16.04 LTS has handled updating this package) 15:35 • There's some resistance to both the Docs and webkit2gtk proposals but I'm still hopeful 15:35 • GNOME accessibility packaging in Debian's VCSs, forwarded Unity7 Orca patches to GNOME (don't know if they want them) 15:35 • Completed evolution 3.24 packaging (waiting in artful new queue). The conversion to cmake made this update challenging. LP: #1685683 15:35 Launchpad bug 1685683 in evolution-data-server (Ubuntu) "Update evolution to 3.24.1" [Wishlist,Fix committed] https://launchpad.net/bugs/1685683 15:35 • Began dropping UOA support from evolution-data-server and unity-control-center (LP: #1695928) 15:35 Launchpad bug 1695928 in gnome-control-center-signon (Ubuntu) "Please remove obsolete UOA packages" [Undecided,New] https://launchpad.net/bugs/1695928 15:35 eof 15:35 thanks jbicha 15:36 #topic jamesh 15:36 I put together a PPA of gnome-builder with the Snapcraft plugin, and 15:36 published some instructions for people to use it on the Snapcraft 15:36 forum. 15:36 I am in the process of putting together a proof of concept prototype 15:36 of xdg-desktop-portal communicating with snaps. 15:36 In the coming week, I will continue with the xdg-desktop-portal work, 15:36 and try to upstream the gnome-builder changes (which will probably 15:36 involve some changes before it is accepted). 15:36 #topic jibel 15:36 heber said he would be out today, are you around jibel? 15:36 willcooke, heber just sent an email with a summary to us 15:37 ah 15:37 saying he has IRC issues 15:37 * willcooke checks mail 15:37 * Investigated https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1694531 15:37 * Investigated https://bugs.launchpad.net/ubuntu/+source/gnome-shell/+bug/1695212 15:37 * Ubiquity tests are most stable, although with the incoming desktop iso gnome-shell we will need to fix some things. 15:37 * Smoke tests fail for server issues (Timeouts, network connection errors, etc.) Seems that the server is still overloaded. Need to take actions on that. 15:37 * Started work on automated installation of desktop images on HW for testing using MaaS and Testflinger. 15:37 Ubuntu bug 1694531 in main-menu (Ubuntu Artful) "17.10 Install Fails to Start" [Critical,Fix released] 15:37 Ubuntu bug 1695212 in gnome-shell (Ubuntu) "[Desktop 20170602] gnome-shell fails to start. Black screen on boot." [Critical,Fix released] 15:37 thanks heber 15:37 #topic kenvandine 15:37 * gnome-3-24 platform snap is now published in the store, edge channel for now. Once I update some of my app snaps to use it for more testing, I'll publish it in the stable channel. 15:37 * Shared the poll results and my initial thoughts with the GNOME engagement team. We'll be getting a call arranged with the GNOME design team and the gnome-shell maintainer to discuss. 15:37 * Prepared a blog post to publish the results, it'll be published on insights later this week. 15:37 * The mir folks would like to keep the mir gtk backend enabled for now, so I'm working on re-enabling that. I'd like to re-enable it and still build without content-hub, working on a patch for that. 15:37 * Re-added unity to the desktop seed until we can get the gdm3 MIR approved, turns out it still tried to load the gdm schema which is provided by libgdm1. 15:37 EOF 15:38 thanks kenvandine 15:38 #topic Laney 15:38 • Merged debhelper from experimental 15:38 ∘ There were a few regressions that I debugged / fixed, and submitted the patches upstream. All got merged now, I think. Turns out it was quite a risky release (experimental, yeesh) 15:38 • All that ↑ was so that I could sync fwupd 15:38 • ubuntu-dev-tools - finished the bzr → git migration and uploaded to experimental, synced 15:38 • asgen/gnome-software: 15:38 ∘ Some meson / buildsystem fixes 15:38 ∘ Implemented type=codec generation for gstreamer packages so they look nice once we use GS instead of sessioninstaller 15:38 ∘ Codec installation to have stopped working in gnome-software since I last tested - need to get to the bottom of that before making a PR for asgen 15:38 • Siloed and uploaded fixes to move the Amazon launcher from Unity to GNOME Shell 15:38 • Internal stuff (travel planning and work priority discussions) 15:38 🙊 15:38 thanks Laney 15:39 #topic oSoMoN 15:39 • upgraded to artful 15:39 • updated chromium dev to 60.0.3112.7 then 60.0.3112.10 15:39 • updated chromium beta to 59.0.3071.83 then 59.0.3071.86 15:39 • chromium 59.0.3071.86 got promoted to stable last night, currently building in PPA and will hand over to ch_risccoulson when ready to publish 15:39 • fixed chromecast bug #1621753 15:39 bug 1621753 in chromium-browser (Ubuntu) "Google Cast no longer finds Chromecast device" [Medium,Fix committed] https://launchpad.net/bugs/1621753 15:39 • working on a script to update copyright file for chromium-browser (not updated since 2012!) 15:39 🔚 15:39 thanks oSoMoN 15:40 BTW, oSoMoN - re: Netflix - I think you should just be able to sign up for a free month's trial, would be easier than me setting it up I think (for me anyway). If that doesnt work though, I'll get an account sorted for you 15:40 ok, I’ll try that 15:40 ta 15:40 #topic seb128 15:40 (one day off, monday was a national holiday) 15:40 • debugged a bit gnome-shell not starting on the daily iso after the unity->GNOME changes (due to missing gdm schemas) 15:40 • tested gnome 3.24 snaps, debugged some issues with Didier and Ken 15:40 • reviewed libreoffice update from Olivier 15:40 • disabled langpack builds for ubuntu touch 15:40 • HR reviews 15:40 • some planning discussions and travel org 15:40 15:40 thanks seb128 15:40 #topic tkamppeter 15:42 tkamppeter, ping when you're back and we'll come back to you 15:42 - system-config-printer: Several patches to improve on recognition of discovered devices coming from the same physical device, conection type list entries in new-printer wizard, also more debug output. Patches submitted upstream, no answer from maintainer yet. 15:42 - CUPS: Mike Sweet rejected filtering network-discovered printers in CUPS. Filtering needs to be done on print dialog level. Concept will be then that the CUPS backend for the print dialog will have option to turn off listing of DNS-SD-discovered printers and then admin can control listing via cups-browsed. 15:42 - Google Summer of Code 2017: Worked with the students on a D-Bus interface to connect print dialog backends with the print dialog itself. 15:42 - Bugs. 15:42 heh, solved. thanks tkamppeter 15:42 #topic robert_ancell 15:43 - Made gnome-shell work without libgdm (more work required - https://bugs.launchpad.net/bugs/1695212) 15:43 - Investigation of ODRS (https://odrs.gnome.org/) for use in Ubuntu 15:43 - Trello board tidying 15:43 - Writing Snapcraft.io forum posts about outstanding desktop issues 15:43 - Simple Scan sponsoring 15:43 Ubuntu bug 1695212 in gnome-shell (Ubuntu) "[Desktop 20170602] gnome-shell fails to start. Black screen on boot." [Critical,Fix released] 15:43 - Short week (Out sick and holiday) 15:43 #topic andyrock 15:43 andyrock, you need more time? 15:43 sooo I failed to record the video 15:43 ack 15:43 #topic AOB 15:43 So, what to do with sessions, usernames, U7 etc 15:43 Me? 15:43 willcooke: ? 15:44 whaaa 15:44 how did that happen 15:44 #topic Trevinho 15:44 * Trevinho feels ignored, alone in a bench in Taipei... 15:44 :-D 15:44 · Fractional scaling in gnome-shell and mutter work (studying code, small fixes...) 15:44 · Looking at fixing st-label (gnome shell toolkit) scaling on scaled-framebuffer based mutter 15:44 · Joined the Fractional Scaling Hackfest (have you heard about it?) 15:44 · Implemented an algorithm to define the closest allowed scale for each sub-integer value we want (8 per integer at the moment) 15:44 · Working in supporting fractional scaling in gnome-control-center 15:44 · Lots of other things related to fractional scaling (like looking at bugs) 15:44 More at http://go.3v1n0.net/FactorialScalingJournal 15:44 😇 15:44 Thanks Trevinho 15:44 so, I had a question... 15:45 If everything is scaled up before being re-scaled for the display its on, is that going to cause lag? 15:45 Seems like n+1 compositing stages? 15:45 No really... 15:46 https://usercontent.irccloud-cdn.com/file/sMgQi8mK/out-5.ogv 15:46 The thing is things aren't scaled up until this is not needed... When it's needed it's just a GL transformation in the compositor, so It scales Up and Down again and it will be 1:1 in 4k, while in normal monitors it will be more expensive, but really nothing you see... 15:46 at least... 15:47 with an intel driven T440s + a 4k monitor, things are pretty fast anyway 15:47 got it, I think 15:47 thanks! 15:47 Great to read the journal 15:47 exciting stuf 15:47 f 15:47 #topic andyrock 15:47 andyrock: thanks for including the commentary! :-D 15:48 ahahah removing the audio was too complicated at this point 15:48 willcooke: it's just few lines, we'll fill better :-) 15:48 so I've working on finishing up the canonical-livepatch panel 15:48 and all the mocks+ 15:48 + 15:48 I need to cleanup a little bit as thare are few corner cases still out there 15:49 and I need to run the mock online 15:49 nice! 15:49 than I'm done with this part 15:49 eow 15:49 awesome, thanks andyrock 15:49 sorry for the audio and for the xxx directory 15:49 :D 15:49 I'll cut the audio out and re-share 15:50 nothing bad inside 15:50 :D 15:50 andyrock: the xxx directory is more important I think... :-D 15:50 andyrock: I believe you! :) 15:50 C:\windows\system32\systemconfiguration.dll # totally not a secret zip file 15:50 but at least it *looks* tthat you pay your taxes... ;-D 15:51 too much bulling 15:51 :) 15:51 :) thanks andyrock 15:51 oki, so back to... 15:51 #topic AOB 15:52 Session names, Unity 7, etc etc 15:52 Who wants to kick us off? 15:52 Depending on what we do (but if we ship any extensions/theme change), I would suggest to use 3 sessions: 15:52 * kenvandine agrees 15:52 - ubuntu -> GNOME with our tweaks (theme/extensions) if any 15:52 - gnome -> traditional upstream GNOME, even maybe with default theme 15:52 - we add an unity session for unity7 15:52 (was ex ubuntu) 15:53 thoughts? 15:53 (default installations will only have "ubuntu") 15:53 don't forget about Wayland/X 15:53 why? This is an implementation detail for our users 15:53 GNOME upstream has "GNOME" and "GNOME on X" (Debian stretch & Ubuntu GNOME switched that to "GNOME on Wayland" and "GNOME") 15:53 i don't think our users should see that 15:54 would make sense for technical users 15:54 less for others 15:54 users shouldn't care if it's Wayland or X 15:54 If they can't run Wayland for whatever reason, can we invisibily move them on to X? 15:54 or without fallback mechanism 15:54 well, that's what we did with unity3d/2d 15:54 that was 2 sessions, and we fallbacked depending on capability at runtime 15:54 assuming we do Wayland by default, you're not going to give an easy option to revert to non-Wayland? 15:55 well, they won't see the session name 15:55 but they still can switch for more advanced users 15:55 on the login screen, Ubuntu GNOME has 2 separate choices (3 if you count GNOME Classic) 15:55 is that something many users are going to want/need? if so why? 15:55 I would think silent fallback is the best 15:56 +1 if we can 15:56 there's a nasty bug in 17.04 (probably mostly fixed by the libgweather SRU) that crashes GNOME Shell 15:56 if GNOME Shell crashes when you are running Wayland, you lose your session and all your work, but if you are running GNOME on X it should magically restart gnome-shell if it can (it usually can) 15:57 well, sounds like more wayland bugs that needs to be fixed, I don't think anyone wants that in the long term 15:57 Having many options at the login screen is too confusing. By default we have GNOME (our implementation), if they have also installed U7, that's there, and if we do have some packages for a more vanilla GNOME and they choose to install that, ok. But by default I think we should just have the one session available from the login screen. 15:57 or everyone would stay on X forever… 15:58 there are a few apps that do not run on our current implementation of GNOME on Wayland (synaptic, gparted, etc.) because they never switched to using policykit 15:58 doesn't prevent to have an option for advanced users somewhere triggering the fallback on boot 15:58 (s/boot/login/) 15:58 jbicha, will those apps run under xwayland? 15:59 but yeah, nothing in the face of our users IMHO 15:59 willcooke: no, I mean there are a couple hacks we could do if we really wanted to but I think we actually do not want to weaken the security improvement 16:00 can we get by without gparted? Does disks do enough to replace it? 16:01 btw, gparted is currently on the live iso but removed after install (while GNOME Disks stays) 16:02 ah, right 16:02 anyway, that's 2 reasons for why users might want to run X instead 16:03 the only other concern is implementing a single session since upstream split it in two 16:03 we can either split it (but it means 2 sessions on the login screen), or do an option in some settings which triggers the fallback at boot that we'll need either way 16:03 I thought I switched synaptic and gparted to pkexec a while ago? 16:05 mdeslaur: I'll investigate and get back with you on that 16:05 jbicha, would you mind checking on that ^ ? 16:06 seb128, Laney, jbicha - So did we reach a conclusion here? 16:07 Don't think so 16:07 hum 16:07 not really no 16:07 Seems to hinge on this fallback thing 16:08 Who can investigate that some more? didrocks, is your knowledge of the 2d/3d stuff useful here? 16:08 GNOME already only offers Wayland if it thinks it's supported 16:09 in gnome-session 16:09 willcooke: for the implementation of the fallback mecanism, yes, but on knowing what to detect for wayland, I'm unsure 16:09 it TryExecs it 16:09 ah 16:09 I don't know if that then falls back to X though 16:09 that's to show in the greeter 16:09 oh right, they do it in gdm 16:09 before 16:09 AFAIK! 16:09 so more work in LightDM needed? Hummm 16:10 Laney: I think it does something more complicated than a simple TryExec 16:10 ok, where's that then? 16:10 I see a TryExec in the .desktop file here 16:10 willcooke: well, we will need a lightDM greeter running on wayland and one on X 16:10 I see a TryExec=gnome-shell which doesn't tell you about wayland 16:10 so basically, lightdm/greeter will need to do that detection 16:11 ? 16:11 it's in /usr/share/wayland-sessions/ 16:12 anyway, I never read the code to find out if that is what happens :) 16:13 seb128, Laney - could you get us to a point that we can talk more on Thursday, and try and get this resolved one way or the other by Friday EOD? 16:13 I think part of it is that GDM runs with Wayland by default 16:13 but if GDM is running as X then it does not provide a Wayland option 16:14 seb128: Laney: don't forget my point about the lightdm greeter in both flavors is needed (wayland and X), so I guess doing the right session presentation is needed at an earlier stage that we had 16:14 didrocks, do we need them at the same time? 16:14 or just fallback to start a X one if wayland fails? 16:15 seb128: "same time", being? 16:15 willcooke, should be doable, we can continue that discussion after meeting/tomorrow 16:15 thanks 16:15 ok, anyone got more AOB before we close the meeting? 16:15 didrocks, "in both flavors" you mean having option to use it with wayland or X or having one of each on different VTs? 16:16 the other question was if unity7 would be autoremovable for upgraders 16:16 willcooke, I have one that can wait next week (rediscuss i386 iso status) 16:16 seb128: no, I mean, lightdm needs to be smart to start an X version of the greeter if the wayland one doesn't work 16:16 didrocks, right, agreed 16:16 (and so, we need a wayland based greeter and an X one) 16:17 are those different greeters? or just lightdm being smart enough to know what server to use and the same greeter code? 16:17 anyway to discuss with Robert I guess 16:17 oki, lets speak about auto remove U7 then 16:17 don't know, yeah, Robert should know more 16:17 thanks for poiting it out didrocks 16:17 yw! 16:17 I'm going to email Robert and Cc you/L_aney/j_bicha 16:17 and Will 16:17 thanks seb128 16:18 thx 16:19 Auto remove U7 - very tricky. I don't really have a strong feeling. It's not that we're offering up GNOME Shell as a preview here, it is the new desktop. So removing it seeeeems to be ok. 16:19 we didn't remove gnome classic when switching to unity 16:19 But it will be surprising to a few people, no doubt 16:19 I'm a bit balanced on that as well 16:19 I don't think it's a policy to remove things for people who upgrade (but we change default session with our new things) 16:19 maybe to revist once we have some feedback about what users think of the new desktop 16:19 I would suggest we keep unity, just move it to this new session for people who want to reswitch 16:19 but yeah what didrocks says 16:19 LP: #1686081 is a nasty bug if a user has Unity7 installed 16:19 Launchpad bug 1686081 in xorg (Ubuntu) "If -synaptics is installed, GNOME Mouse & Touchpad Settings doesn't work" [High,Confirmed] https://launchpad.net/bugs/1686081 16:20 don't want to shock the users 16:20 currently, there are zero people maintaining unity7 16:20 jbicha keeps nagging about that one 16:20 unity7 FTBFS right now and I'd like if someone could help me on it :) 16:20 (test failures) 16:21 Trevinho, andyrock, ^ can you have a look? 16:21 jbicha, do you have a bug open about that? 16:21 would be useful to assign etc 16:21 I'll probably open a bug, but currently it was briefly discussed at https://code.launchpad.net/~jbicha/unity/drop-uoa-recommends/+merge/325097 16:21 it's eod for me 16:21 please open a bug and give me the ref 16:21 I can take a look tomorrow 16:22 I see with andyrock tomorrow 16:22 andyrock, yeah tomorrow is fine, thanks 16:23 I would be happier with Unity7 if it could be ported to use unity-settings-daemon, unity-control-center but that needs more work to reimplement stuff that GNOME upstream dropped 16:23 you mean gnome-settings-daemon? 16:23 oops, yes 16:24 well the oem team would like to backport libinput support for xenial ideally 16:24 so maybe that specific synaptic issue is resolved by that work 16:24 well "backport", rather add and backport 16:24 yes, that would fix that 16:25 on removing let's maybe revisit in another meeting or discuss on the list? 16:25 +1 16:25 we quite overrun already 16:25 let's get the blockers looked at in the meantime too 16:25 yeah 16:25 some GNOME apps really want GNOME Online Accounts to be available but that's messy with having 2 control-centers in Unity7 16:26 I will add that bug to the Trello board and we can get someone assigned 16:26 well uoa is deprecated so it wouldn't be 2 in artful 16:26 willcooke, thanks 16:26 right, UOA is gone but GOA is a gnome-control-center panel 16:26 right :-/ 16:27 let's do another meeting about unity7 16:27 it's too much to discuss today 16:27 maybe meanwhile list issues on a wiki or bugtag them 16:27 it is possible to run Unity7 without unity-settings-daemon but a few things don't work (some media-keys shortcuts, gnome-control-center's displays panel, etc.) 16:28 ok, going to close this meeting now. Seb, let's talk i386 next week. We can carry on this discussion in a moment ^ 16:29 seb128 is going to email Robert 16:29 cool 16:29 We need to look at that libinput stuff again 16:29 and then we can decide what to do with U7 16:29 Is that right? 16:30 yes 16:30 :) thanks jbicha 16:30 in which case... 16:30 #endmeeting