Hey Friends, Here are the minutes and log of today's meeting.
Cheers, -- sebas http://www.kde.org | http://vizZzion.org
Plasma Team meeting, 20-302017 Present: notmart, Sho, bshah notmart: * [Qt] eaten events qt patch https://codereview.qt-project.org/#/c/185946/ * [Qt] Add NOTIFY signals for WheelArea https://codereview.qt-project.org/#/c/189158/ MERGED * [Qt] Environment variable enabling the workaround FBO readback bug https://codereview.qt-project.org/#/c/185259/ MERGED * MERGED better layout and animate keyboard in sddm * MERGED better layout and animate keyboard in lockscreen * Multiscreen fix: corner case of primary screen at 0,0 that moves the other (another case in which would be nice to test with 3 or more) https://phabricator.kde.org/D5323 Diff 5323 "Correctly handle when a new primary screen displaces the old" [Closed] https://phabricator.kde.org/D5323 * Kirigami2: bugfixes, ** in ScrollablePage whith multiple children of which only one is the scrollable ** support for high dpi icons in the desktop version ** better at opening the drawers with tap ** MERGED ApplicationItem AbstractApplicationItem, Gemini starting using them * had to rebuild whole stack, wasted a couple of days Sho: * [FV] Fixed the actions overlay icons sometimes appearing in the drag pixmap attached to the cursor during DND, needs review * [FV] Fixed a regression causing popups to close when dragging into them via spring-loading, needs review * [FV] Some code simplification/cleanups * [Input] On Martin's request, wrote a many-page overview over the state of complex text input in Plasma and what needs to be done (especially for Wayland): https://mail.kde.org/pipermail/plasma-devel/2017-April/068978.html to kick off discussion ** Participated in said discussion ** Pointed people (e.g. the ibus maintainer) at this discussion via IRC and email poking and https://blogs.kde.org/2017/04/06/complex-text-input-plasma *** Happy to say we got lead devs of both fcitx and ibus involved * [Input] Helped csslayer test some kimpanel changes to better autostart our ibus panel frontend * [Input] Talked with notmart and csslayer about how we can start integrating Input Method Panel and System tray to provide consistent layout and behavior and eliminate redundant icon hiding systems, which is one of the lower-hanging fruit in the input story * [Launchers] Working on some Simple Menu UI changes/experiments * [Launchers] Me, Marco and Ivan decided together to delay per-activity faves until 5.11 - particularly the migration story is not fully cooked yet, and we can no longer merge it early enough in the cycle to get sufficient testing * [Launchers] Instead we're working to expose "most used" data better in the launchers for 5.10 bshah: - 1st thing I was working on the archlinux rootfs for plasma mobile - I am now able to run kwin/wayland and plasmashell but.. problem is xwayland is not started if session/kwin is started by simplelogin resulting in kwin_wayland quitting if I start kwin_wayland manually it starts fine - I will investigate that over next week - Next thing is After Canonical stopped development of ubuntu touch, ubports team is looking forward with closer collaboration on base stack Basically, android part of the Plasma Mobile/Ubuntu Touch or Mer romangg: - Xwayland research and working on direct scanout - plan on doing more this week kbroulik: - Did some fixes here and there, plasma-windowed and Folder View - Ported kwin scripts importer to new KPackage stuff so it actually works (D5318 few minor issues need addressing before it can go in) Diff 5318 "[KWin Scripts KCM] Restore "import" option" [Needs Revision] https://phabricator.kde.org/D5318 - We now have KWin desaturate windows of frozen apps on X \o/ go check it out kill -SIGSTOP something and click the X button ;) - Animated quick tiling outline (++beauty) D5262 - Diff 5262 "Animate quick tiling outline" [Needs Review] https://phabricator.kde.org/D5262 - Some pending patches D5333, D5331, D5330, D5321 Diff 5333 "[Plasma Windowed] Disable standard actions on SNI" [Needs Review] https://phabricator.kde.org/D5333 Diff 5331 "Whitelist all Keyboard Layout Switcher shortcuts" [Needs Review] https://phabricator.kde.org/D5331 Diff 5330 "[QDBusMenuBar] Don't wait for (un)register call to finish" [Needs Review] https://phabricator.kde.org/D5330 Diff 5321 "[Notifications Applet] Don't crash when containment() disappears" [Needs Review] https://phabricator.kde.org/D5321 - Had a look at d_ed's browser extension work and started working on a KRunner that finds open browser tabs (e.g. "open windows runner" for browser tabs) and allows switching to them. Will continue on that, I guess. [12:00:16] <sebas> hey [12:00:37] <Sho_> yo sebas [12:01:19] <notmart> oh, 12:00 [12:01:21] <notmart> meeting? [12:01:23] <sebas> right :) [12:01:24] <Sho_> yup [12:01:25] <sebas> yup [12:01:28] <sebas> Who's here for it? [12:01:32] <Sho_> me me me [12:01:33] -*- sebas obviously! [12:01:35] <notmart> i'm here [12:01:55] <sebas> kbroulik? romangg? perhaps [12:02:05] <bshah> me me me [12:02:16] <Sho_> Riddell and d_ed are other options [12:02:58] <sebas> We can start I guess :) [12:03:01] <sebas> notmart! [12:03:19] <notmart> ok [12:03:22] <kbroulik> here [12:03:23] <notmart> * [Qt] eaten events qt patch https://codereview.qt-project.org/#/c/185946/ [12:03:24] <notmart> * [Qt] Add NOTIFY signals for WheelArea https://codereview.qt-project.org/#/c/189158/ MERGED [12:03:26] <notmart> * [Qt] Environment variable enabling the workaround FBO readback bug https://codereview.qt-project.org/#/c/185259/ MERGED [12:03:27] <notmart> * MERGED better layout and animate keyboard in sddm [12:03:29] <notmart> * MERGED better layout and animate keyboard in lockscreen [12:03:30] <notmart> * Multiscreen fix: corner case of primary screen at 0,0 that moves the other (another case in which would be nice to test with 3 or more) https://phabricator.kde.org/D5323 [12:03:30] <sKreamer> Diff 5323 "Correctly handle when a new primary screen displaces the old" [Closed] https://phabricator.kde.org/D5323 [12:03:32] <notmart> * Kirigami2: bugfixes, [12:03:33] <notmart> ** in ScrollablePage whith multiple children of which only one is the scrollable [12:03:35] <notmart> ** support for high dpi icons in the desktop version [12:03:36] <notmart> ** better at opening the drawers with tap [12:03:38] <notmart> ** MERGED ApplicationItem AbstractApplicationItem, Gemini starting using them [12:03:39] <notmart> * had to rebuild whole stack, wasted a couple of days [12:03:50] --> Schrostfutz (~Schrostfu@131.234.199.69) has joined #plasma [12:04:07] <notmart> so, of the 3 painful qt patches, 2 are merged, https://codereview.qt-project.org/#/c/185946/ is still being ignored [12:04:11] <notmart> will pound more on that [12:05:19] <notmart> on the multiscreen, people test more on that, seems to fix issues for kbroulik, would like to hear also from subdiff as had a similar issue [12:06:04] <Sho_> notmart: on multiscreen i keep getting reports about the "tm forgets windows on screen res changes and monitor replugs" thing we tried to fix (in both libtm and the containment not signalling screengeo change) [12:06:12] <Sho_> i'm out of things i can do with pure audit since i don't have hw [12:06:17] <Sho_> so it needs other people to debug at this point [12:06:46] <Sho_> (happy to help obviously) [12:06:49] <notmart> yeah, is not a problem i ever had (well, except for one particular case we long fixed...) [12:07:17] <Sho_> i have a vague hope that the remaining reporters are on the wrong combo of plasma and frameworks or something since the fix was distributed [12:07:20] <Sho_> but not really sure [12:08:02] <sebas> that's more a question of triaging, not testing, right? [12:08:41] --> Axm__ (~sailfish@223.180.13.200) has joined #plasma [12:08:44] <Sho_> well let's put it this way, when we get reports against 5.9.4 it's unlikely that the same distro has an ancient frameworks [12:08:46] <-- Axm___ (~sailfish@223.180.11.14) has quit (Read error: Connection reset by peer) [12:09:38] <sebas> but if you're unsure, you got to ask the reporter, I don't see a way around that...? [12:10:05] <-- knarfy (~knarf@2a01:6600:2000:8804:a1c6:7ea7:11a0:e172) has quit (Ping timeout: 252 seconds) [12:10:08] <Sho_> sure [12:10:27] <sebas> Alright, let's move on :) [12:10:33] <sebas> Sho_: your turn :) [12:10:42] <Sho_> * [FV] Fixed the actions overlay icons sometimes appearing in the drag pixmap attached to the cursor during DND, needs review [12:10:44] <-- Axm__ (~sailfish@223.180.13.200) has quit (Read error: Connection reset by peer) [12:10:44] <Sho_> * [FV] Fixed a regression causing popups to close when dragging into them via spring-loading, needs review [12:10:45] <Sho_> * [FV] Some code simplification/cleanups [12:10:47] <Sho_> * [Input] On Martin's request, wrote a many-page overview over the state of complex text input in Plasma and what needs to be done (especially for Wayland): https://mail.kde.org/pipermail/plasma-devel/2017-April/068978.html to kick off discussion [12:10:48] --> Axm___ (~sailfish@223.180.23.127) has joined #plasma [12:10:48] <Sho_> ** Participated in said discussion [12:10:50] <Sho_> ** Pointed people (e.g. the ibus maintainer) at this discussion via IRC and email poking and https://blogs.kde.org/2017/04/06/complex-text-input-plasma [12:10:51] <Sho_> *** Happy to say we got lead devs of both fcitx and ibus involved [12:10:53] <Sho_> * [Input] Helped csslayer test some kimpanel changes to better autostart our ibus panel frontend [12:10:54] <Sho_> * [Input] Talked with notmart and csslayer about how we can start integrating Input Method Panel and System tray to provide consistent layout and behavior and eliminate redundant icon hiding systems, which is one of the lower-hanging fruit in the input story [12:10:56] <Sho_> * [Launchers] Working on some Simple Menu UI changes/experiments [12:10:57] <Sho_> * [Launchers] Me, Marco and Ivan decided together to delay per-activity faves until 5.11 - particularly the migration story is not fully cooked yet, and we can no longer merge it early enough in the cycle to get sufficient testing [12:10:59] <Sho_> * [Launchers] Instead we're working to expose "most used" data better in the launchers for 5.10 [12:11:54] <sebas> No questions ... ? [12:11:56] <sebas> bshah then [12:12:07] <bshah> okay [12:12:52] <bshah> - 1st thing I was working on the archlinux rootfs for plasma mobile [12:13:07] <bshah> - I am now able to run kwin/wayland and plasmashell [12:13:09] <bshah> but.. [12:13:29] <bshah> problem is xwayland is not started if session/kwin is started by simplelogin [12:13:36] <bshah> resulting in kwin_wayland quitting [12:13:47] <bshah> if I start kwin_wayland manually it starts fine [12:13:58] <bshah> - I will investigate that over next week [12:14:08] <bshah> - Next thing is [12:14:46] <bshah> After Canonical stopped development of ubuntu touch, ubports team is looking forward with closer collaboration on base stack [12:15:04] <bshah> Basically, android part of the Plasma Mobile/Ubuntu Touch or Mer [12:15:09] <Sho_> it might be helpful to mention what ubports is (i don't really know) [12:15:13] <sebas> the latter is interesting, definitely something we should work on this week [12:15:37] <bshah> UBports, is community project to port ubuntu touch to devices [12:15:44] <sebas> the kwin_wayland stuff : awesome as usual bshah :) [12:16:21] <bshah> So yeah I will send detailed email to plasma-devel soon about this [12:16:32] <Sho_> did ubports exist before the canonical desicion? [12:16:35] <Sho_> or is this newly formed [12:16:41] <bshah> Sho_: it was [12:16:51] <-- [Enrico] (~chiccoroc@gentoo/contributor/Enrico) has quit (Ping timeout: 245 seconds) [12:16:52] <bshah> ubports.com [12:16:57] <sebas> seemed to have been there for quite some time no? [12:16:59] <notmart> already existed for porting to devices not supported to canonical i think [12:17:03] <bshah> yes [12:17:04] <Sho_> ah ok [12:17:12] <sebas> cyanogenmod of ubuntu touch kind of? [12:17:21] <bshah> no not really [12:17:21] --> stikonas (~gentoo@wesnoth/translator/stikonas) has joined #plasma [12:17:25] <Sho_> do they have any involvement from (now former?) canonical staff or all external? [12:17:41] <bshah> Sho_: no mostly community persons [12:17:51] <bshah> "UBports is a team of developers and a meeting place for developers that wish to port Ubuntu Touch to as many devices as possible, this is a place where developers can talk to other developers, learn from each other and help push Ubuntu to more devices as a team, or by yourself but with community support if you wish." [12:17:57] <bshah> description from their website [12:18:01] <Sho_> aye [12:18:12] <notmart> that's why they are still around and not disappeared the minute after the announcement :p [12:18:16] <bshah> this is basically attempt to kill the fragmentation we have in terms of mobile operating system [12:18:38] <bshah> and in my opinion we should be part of it [12:18:54] <Sho_> i would agree, it doesn't really make sense to do enablement work redundantly [12:19:09] <sebas> we should at least talk about vision, direction, plans and stuff [12:19:10] <notmart> yeah [12:19:17] <acheronuk> talk to Mario maybe? he has been on kubuntu podcast about ubiports a couple of times [12:19:30] <bshah> I am in already talks with them.. [12:19:36] <acheronuk> :) [12:19:36] <sebas> Or maybe someone who manages / runs ubports directly? [12:19:39] <Sho_> ubports might have to make some sort of commitment that the base enablement part isn't all geared toward the ut shell [12:19:41] <bshah> and will attend meeting today at 15 UTC.. [12:19:54] --> Axm__ (~sailfish@106.209.204.83) has joined #plasma [12:19:54] <bshah> sebas: yes, mariogrip is main maintainer AFAIK [12:19:57] <sebas> I mean, "someone who talks about someone" is not better than the people who are now left alone? [12:19:57] <Sho_> also in terms of public messaging, because i wouldn't want PM to be treated as alternative [12:20:07] --> HoloIRCUser4 (~holo...@user-94-254-131-28.play-internet.pl) has joined #plasma [12:20:12] <sebas> Sho_: +1 [12:20:29] <bshah> Sho_: this collaboration is not about userspace at all [12:20:36] <bshah> It ends at libhybris builds [12:20:39] <sebas> bshah: 15UTC today, what kind of meeting? Where? [12:20:52] <bshah> sebas: its hangout most likely [12:20:54] <sebas> I'm interested and probably can make it then [12:20:59] <bshah> if you want I can invite you [12:21:03] <Sho_> bshah: i know, i'm just saying if they e.g. get you to contribute to ubports repositories, then the public side should represent your work properly [12:21:03] <sebas> yes please [12:21:23] <notmart> y may be there as well if needed tough would be mostly lurking [12:21:43] <Sho_> i can't do a/v because coffeeshop but i could do text [12:22:07] <silviof> How can I build a complete plasma mobile image with latest sources? Some documentation around? (bshah) [12:22:22] <-- Axm___ (~sailfish@223.180.23.127) has quit (Ping timeout: 240 seconds) [12:22:27] <Sho_> silviof: mind waiting 20 mins till our meeting is over? :) [12:22:29] --> Axm___ (~sailfish@106.209.202.136) has joined #plasma [12:22:30] <Sho_> sorry [12:22:30] <notmart> silviof: we're in a meeting at the moment, just wait a bit [12:22:42] <silviof> hups - sorry! [12:22:47] <Sho_> np [12:22:58] <sebas> I'm last anyway [12:23:01] <bshah> so.. well yeah, that's it from my side [12:23:08] <bshah> any questions? otherwise next [12:23:10] <-- Axm__ (~sailfish@106.209.204.83) has quit (Read error: Connection reset by peer) [12:23:25] <sebas> Not much happened on Plasma side, but I'm pretty much done moving and have settled into the new place now, feels really good. [12:23:50] <-- llucas (~holo...@host-89-230-172-232.dynamic.mm.pl) has quit (Ping timeout: 255 seconds) [12:23:52] <sebas> My desk is full of the stuff that I didn't manage to lose over the move, so I'll clean up the more urgent things this week [12:24:02] <Sho_> does your new lair have a divable underground cave system in the basement? [12:24:06] <Sho_> because that's how i envision it [12:24:15] <Sho_> :) [12:24:29] <kbroulik> was it already my turn? sorry, been distracted [12:24:38] <notmart> also known as cellar with broken plumbing? :p [12:24:38] <sebas> Also: my laptop shows signs of disk death, and my workstation is entirely dead, no bleep even. I can't backup my laptop to have it repaired until I fixed the workstation (new power supply, which I hope is the culprit is on my desk already) [12:24:57] --> HoloIRCUser6 (~holo...@host-89-230-172-232.dynamic.mm.pl) has joined #plasma [12:24:58] <Sho_> excuses to buy new hw! [12:25:04] <sebas> Sho_: sort of, my bed room is souterrain, and I'm planning to equip it with deep wreck diving photos :) [12:25:07] --> HoloIRCUser (~holo...@user-94-254-130-222.play-internet.pl) has joined #plasma [12:25:27] <sebas> it's not filled with water, but may be at some point, I'm 150m from the Rhine [12:25:33] <romangg> your new home is over a magnetized iron-stream 8) [12:25:49] <Sho_> trivia: koreans don't have bathtubs or shower cabins, isntead the entire bathroom is built as a wet cell so you use it straight as shower cabin [12:25:53] <sebas> romangg: that, or my moving skills are a little to hard-handed [12:25:53] <Sho_> potentially divable [12:25:57] <sebas> could really be either [12:26:16] <sebas> Sho_: yes, it's water, that makes it enticing :) [12:26:28] <romangg> (sorry, I have to leave again, but there wasn't much stuff I did last week besides some Xwayland research and working on direct scanout - plan on doing more this week) [12:26:35] <sebas> also: dangerous with currents and traffic and low vis, must be awesome ;) [12:26:42] <Axm___> hi everyone, just a short question, i have been working on the mycroft plasmoid project but am facing a blocker with plasma shell crashing randomly when using objectmodel to dynamicly add components, after a gdb backtrace i end up with memory related corruptions in libqml and libqt so should i be filing a bug for this in Plasma or Upstream at Qt ? [12:26:44] <sebas> romangg: aye, cool [12:26:50] <sebas> kbroulik: next up is you [12:27:07] <sebas> Axm___: could you give us a few minutes until we finished this meeting? [12:27:13] <kbroulik> Did some fixes here and there, plasma-windowed and Folder View [12:27:14] <Sho_> Axm___: drop the backtrace here after we get through the meeting [12:27:19] <kbroulik> Ported kwin scripts importer to new KPackage stuff so it actually works (D5318 few minor issues need addressing before it can go in) [12:27:20] <sKreamer> Diff 5318 "[KWin Scripts KCM] Restore "import" option" [Needs Revision] https://phabricator.kde.org/D5318 [12:27:27] <kbroulik> We now have KWin desaturate windows of frozen apps on X \o/ go check it out kill -SIGSTOP something and click the X button ;) [12:27:38] <kbroulik> Animated quick tiling outline (++beauty) D5262 [12:27:38] <sKreamer> Diff 5262 "Animate quick tiling outline" [Needs Review] https://phabricator.kde.org/D5262 [12:27:41] <sebas> Axm___: sounds like you want to do a minimal test case and report against Qt, we can't fix this kind of problems in Plasma [12:27:49] <kbroulik> Some pending patches D5333, D5331, D5330, D5321 [12:27:50] <sKreamer> Diff 5333 "[Plasma Windowed] Disable standard actions on SNI" [Needs Review] https://phabricator.kde.org/D5333 [12:27:50] <sKreamer> Diff 5331 "Whitelist all Keyboard Layout Switcher shortcuts" [Needs Review] https://phabricator.kde.org/D5331 [12:27:50] <sKreamer> Diff 5330 "[QDBusMenuBar] Don't wait for (un)register call to finish" [Needs Review] https://phabricator.kde.org/D5330 [12:27:50] <sKreamer> Diff 5321 "[Notifications Applet] Don't crash when containment() disappears" [Needs Review] https://phabricator.kde.org/D5321 [12:27:54] <romangg> btw bshah: If Xwayland still doesn't work for you we can try to solve it later this afternoon again. [12:28:00] <kbroulik> Had a look at d_ed's browser extension work and started working on a KRunner that finds open browser tabs (e.g. "open windows runner" for browser tabs) and allows switching to them. Will continue on that, I guess. [12:28:11] <kbroulik> next [12:28:34] <Sho_> oh i thought you were lead on the browser stuff [12:28:47] <-- HoloIRCUser4 (~holo...@user-94-254-131-28.play-internet.pl) has quit (Ping timeout: 252 seconds) [12:28:47] <-- aspotashev (~aspotashe@31.173.49.33) has quit (Ping timeout: 252 seconds) [12:29:20] <-- HoloIRCUser6 (~holo...@host-89-230-172-232.dynamic.mm.pl) has quit (Ping timeout: 252 seconds) [12:29:26] <sebas> "We're KDE, when someone asks us who the leader is, we point elsewhere." [12:29:28] <Sho_> kbroulik: mind re-reviewing my two open FV phabs later if you get a moment? [12:29:38] <-- ltinkl (~ltinkl@2a00:1028:83cc:7d32:79a4:34bf:148d:3a35) has quit (Remote host closed the connection) [12:29:44] <kbroulik> Sho_: sure [12:29:48] <Sho_> thx <3 [12:30:04] <kbroulik> Sho_: well, I am, but d_ed did some protocol-marshalling work stuff so I have a good infrastructure "plugin like" to work on :) [12:30:18] <kbroulik> just doing praise where it's due [12:30:22] --> ltinkl (~ltinkl@2a00:1028:83cc:7d32:79a4:34bf:148d:3a35) has joined #plasma [12:30:27] <Sho_> team work ftw [12:30:34] <sebas> cool [12:30:43] <sebas> Anybody else got something to report? [12:30:53] <Sho_> i have a small follow-up on the multiscreen thing earlier [12:31:04] <Sho_> i checked all the reports, and i see two reports against f 5.31 [12:31:08] --> fbeutel (~thunde...@wl-pool2-ont-074.uni-muenster.de) has joined #plasma [12:31:10] <Sho_> frameworks that is [12:31:11] <Sho_> which has the patch [12:31:13] <Sho_> so, still buggy [12:31:20] <sebas> Meh ;) [12:31:24] <Sho_> yeah :( [12:31:33] <Sho_> sorry to end on a downer ;) [12:31:47] <notmart> Sho_: what patch/what bug? [12:31:47] <Sho_> bonus spring pic for better feelings: https://www.instagram.com/p/BSlBaE_F5K1/ [12:32:04] <Sho_> notmart: 2c69d22546639, the "containment doesn't signal screen geo change" thing [12:32:10] <bshah> I've one thing to mention [12:32:12] <bshah> Akademy CFP deadline is today! [12:32:13] <Sho_> the bug is https://bugs.kde.org/show_bug.cgi?id=373075 [12:32:13] <sKreamer> KDE bug 373075 in plasmashell (Task Manager) "Changing Resolution hides programs in task manager" [normal,] https://bugs.kde.org/show_bug.cgi?id=373075 [12:32:16] <notmart> ok, the tasks disappearing then [12:32:35] <Sho_> (also, indoor amusement park! https://www.instagram.com/p/BSlgkYVF4e6/) [12:32:38] <sebas> ow fuck ... now already? [12:32:51] <notmart> Sho_: so, it's probably a very similar scenario, but happening at a slightly different codepath from what the patch is adressing [12:32:52] <sebas> (akademy deadline) [12:33:02] <bshah> sebas: yessir [12:33:08] <bshah> and yes.. I am in same panic mode [12:33:10] <bshah> :p [12:33:16] <sebas> Should we close the meeting though and call the rest breakouts or normal chattery? [12:33:25] <sebas> bshah: PANIC IS GOOD [12:33:30] <notmart> anybody else? [12:33:32] <Sho_> i know rdieter from fedora is working on a talk submission [12:33:35] <Sho_> fwiw [12:33:42] <sebas> except when it kills you, then death being bad trumps panic being good [12:33:54] <notmart> i don't think i have particular things i want to talk about this year [12:34:03] <sebas> So, meeting closed, thanks everybody, will send minutes later.
