Week 14 of 2020 Plasma meeting notes from 2020-03-30:
Hope you're all doing well and staying at home.
DavidRedondo
============
* I am continuing work for coloring icons in the correct colors
* Patch for breeze kdecoration now up on phab:
https://phabricator.kde.org/D28330
* IconItem is fixed already
* Somebody reported to me a bug in the systray because they thought it
was also coloring bug
* Turns out our sni implementation was bad
* The data engine does complicated icon logic which breaks and has
weird interaction with plasma iconitem
* Patch on phab to expose only the raw properties in the engine and do
the rendering in the actual implementation
https://phabricator.kde.org/D28208
Q/A:
<d_ed> dataengines are supposedly stable API
<kbroulik> d_ed: yeah that was my main concern also but I think we all
involved agreed it was an acceptable break for the common good™
<DavidRedondo> Yeah that's why I didn't actually remove the properties
for now
<d_ed> ok, cool
<DavidRedondo> So we could keep them around if we want to
<DavidRedondo> And just ignore them in the applet
<d_ed> nah, I'm fine with killing all dataengines with fire :)
<d_ed> as long as we're aware of everything
<DavidRedondo> We asked latte if they use it and they just use the
systray applet
d_ed
====
* I've been working on two big things
* plasma getting a systemd powered startup
* This is most going all well, all the long groundwork moving
plasma-session meant it's all gone pretty smoothly
* Also I've been doing some big changes to ksysguard
* last year we introduced plugins for processes, which worked quite well
* so I've done the same for the global stats
* with a new API and all
* these are some big reviews, so we need some volunteers on that
* In terms of my usual frustrated rant:
* This bug list of "very high priority bugs"
https://bugs.kde.org/buglist.cgi?cmdtype=dorem&list_id=1729437&namedcmd=VHI-priority%20Plasma%20bugs&remaction=run&sharer_id=208174
* seems to not have moved at all
* like last week
* we had a "security" ish, issue last week with .desktop files having an
Exec line being used in KRun(someUrl)
* whilst we're doing the Application/Command job porting and looking
through usages of KRun it's worth checking any that use the
KRun(someUrl) constructor and possibly making them not execute things
* my other thing I wanted to discuss - if lots of people are on Corona
lockdown it might be a good opportunity for some onboarding
* so I was thinking we could organise some "bug day" or day-long
virtual sprint or something...
Q/A:
<DavidRedondo> I think the breeze gtk thing is moving
<kbroulik> I've got stuck on the pinned apps moving
<kbroulik> I'm pretty sure I know what's going on and what's causing it,
fixing it is annoying
<d_ed> ok, that's good to hear
<d_ed> make sure to keep bugs up to date with findings
<kbroulik> I wonder if Bug 418466 is also a side-effect of that ghns
changedentrieschanged emissions bug
<sKreamer> KDE bug 418466 in systemsettings (kcm_desktoptheme) "Intalled
Global themes not showing up in Setting's Global theme section."
[normal,] https://bugs.kde.org/show_bug.cgi?id=418466
<d_ed> I hereby task you with finding out !
<kbroulik> ok
[...]
<DavidRedondo> d_ed: but that's the same with clicking on said file right
<kbroulik> DavidRedondo: when clicking on it in dolphin it asks
<kbroulik> recent docs runner would just blatantly execute
<d_ed> KRun(someUrl) also asks
<d_ed> if not executable
<kbroulik> also, if possible use service explicitly with URL
<d_ed> yeah, that too
<kbroulik> e.g. that's what I added KServiceAction for
applicationlauncherjob, so we don't exec a command but invoke the
"proper service"
<d_ed> DavidRedondo: so you're right it wasn't super critical or easy to
exploit, but it's still worth locking down what we can
[...]
<DavidRedondo> Speaking of corona, because of sprints being canceled I
had the idea maybe we could arange topic calls or something if needed?
<d_ed> DavidRedondo: ack. I think we need a wayland call at some point
after that long ML thread
<bshah> Virtual sprint? :D
bshah
=====
* I have been mostly busy with:
* "[libnotificationmanager] forward notifications to lockscreen" [Needs
Review] https://phabricator.kde.org/D28427
* Diff 28428 "[screenlocker] listen to notifications from the
notification server" [Needs Review] https://phabricator.kde.org/D28428
* There's some more code but it is not polished enough to be public :P
* Apart from it my last week hasn't been that much productive.
Q/A:
<d_ed> is that stuck on anything needing discussion?
<bshah> d_ed: nope, not really
<ahiemstra[m]> how generic is that?
<ahiemstra[m]> as in, would it also support displaying notifications on
the normal lockscreen?
<kbroulik> not very iirc
<kbroulik> we've discussed making it generic but we kinda coujldn't
agree on a scope and security model
<bshah> ahiemstra[m]: it could, yes. But I am test-driving it on mobile
greeter first
<ahiemstra[m]> right
<bshah> then someone can add it in desktop greeter ;)
<ahiemstra[m]> sure
<bshah> there's bit we (I and Kai) discussed about notification handler
<bshah> which is kinda neat, but then we didn't find a right place to
put this as API
<bshah> and then my nice plan was killed :'(
<kbroulik> yeah I would have loved a generic notification proxy thing
also for kde connect
<kbroulik> but whenever we started discussing it, the scope went out of
hand and then it ended up nowhere for being too large
<kbroulik> so I'd say we focus on this specific thing first, make it
good, and then see what we need in the future
<bshah> (I still have some notes about handler stuff locally, I can
potentially start discussion on p-d)
<ahiemstra[m]> TODO: far in the future ;)
<kbroulik> bshah: yes, please do
<kbroulik> also make sure to include kdeconnect in the mix :)
ahiemstra
=========
* I finished up the shadowedtexture last week, that's ready to be landed
* Diff 28222 "Add ShadowedImage and ShadowedTexture items" [Accepted]
https://phabricator.kde.org/D28222
* I also did a follow up to have separate corner radii
* Diff 28314 "Allow setting separate corner radii for
ShadowedRectangle" [Needs Review] https://phabricator.kde.org/D28314
* I've done some small updates on the ksysguard library work that
supports d_ed's daemon
* last thing I wanted to mention is that I'm looking into
* https://bugreports.qt.io/browse/QTBUG-49646 "Kerning incorrect in
some applications"
notmart
=======
# Kirigami:
* fixed a freeze while flicking on mobile
* finished overlaysheet refactor
# Plasma Bigscreen:
* last fixes for the beta
* beta released!
* wrote a couple of blog posts about it (second i'll publish today)
https://notmart.org/blog/2020/03/plasma-bigscreen
* started making mycroft fully optional
let's see if i'll be able of doing some more plasma-plasma things now
Q/A:
<kbroulik> I've did an experiment of using that color detector thing
from bigscreen to colorize my task bar entries by app icon
<kbroulik> it sounded nicer in my head than it now actually looks :)
<notmart> eh, it's the typical cool thing that has to be used very with care
<kbroulik> yeah
<veggero> That's interesting though, maybe it could be worth a VDG
discussion
<notmart> with great power comes great current squared times resistance
<kbroulik> veggero: I just did a crappy "colorize the svg thing". I
recall icontasks in kde4 had a nice glow and stuff where the color
looked really nice
<kbroulik> maybe works better for icontasks.. anyway i'd love that color
detector somewhere in a usable lib :P
<notmart> i wouldn't like a clone of the windows7 taskbar look with that
silly colorized gradient on mouse over
<ahiemstra[m]> what they do in win10 is even worse, it's basically that
effect but only on an outline
fvogt
=====
* would only ask about the state of Qt 5.15 support anyway
Q/A:
<kbroulik> fvogt: what about it?
<kbroulik> apart from deprecation porting
<d_ed> fvogt: As I understand it everything is all ok
<d_ed> I should upgrade my laptop to latest Qt5.15 branch to be sure
there's been nothing last minute
<fvogt> There were some reports about severe instability on bugzilla,
but openQA couldn't confirm that
ismailof
========
* @kbroulik asked me to discuss a bit about the kwinrules kcm?
Q/A:
<kbroulik> is there something you're stuck on with your rules kcm?
<d_ed> ismailof[m]: you've done a great job on that btw
<kbroulik> I've been poking some devs occasionally to keep the momentum
going
<ismailof[m]> not really stuck, just being conservative about when "wip"
ends :D
<d_ed> ismailof[m]: given it's an advanced user KCM, I think in terms of
the UI/UX you're hitting diminishing returns with changes
<ahiemstra[m]> yeah it's really cool
<kbroulik> just be mindful of the 5.19 freezes early may
<kbroulik> so perhaps merge soon-ish and then clean it up until the beta? :)
<kbroulik> I just want to avoid someone going a day before the Beta "oh
this is so cool, can we have this in" so i'd rather merge sooner than later
<ismailof[m]> alright! I'll take today for some polishing maybe, and set
it ready for review
kbroulik
========
# Notifications:
* Continued discussion about inline-reply spec on XDG...
* Added `KNotificationJobUiDelegate`, please review: D28295
* Diff 28295 "Introduce KNotificationJobUiDelegate" [Needs Review]
https://phabricator.kde.org/D28295
* Allows to use KNotification for job error reporting rather than a
modal message box
* Useful for the new `ApplicationLauncherJob` replacement for `KRun` so
apps launched from Plasma use a nice Plasma error notification
# plasma-firewall:
* Made `KSortFilterProxyModel` `invalidateFilter` public
* Eases creation of QML-only custom view filtering
* Fixed size propagation bug in `FormLayout`
* Did you know QtQml `Timer` default `interval` is 1000? :)
* Some minor UI improvements
# Other:
* Fixed virtual desktop pager not updating rows when only rows changed
but not the number of desktops
* Made a bodge fix for wallpaper thumbnails not loading: D28420
* Diff 28420 "[Image Wallpaper] Fix thumbnail generation when model is
reloaded in-flight" [Needs Review] https://phabricator.kde.org/D28420
* Caused by GHNS Button signalling a change when there isn't. See
commit description in the phab for what's going on
* Added support for launching `KServiceActions` in `ApplicationLauncherJob`
* This will give us a proper API for launching jump list actions rather
than just running the action `Exec` as command
* Patch for Task Manager D28286 and services/shell runner D28347
* Diff 28286 "[Task Manager] Port backend to ApplicationLauncherJob"
[Needs Review] https://phabricator.kde.org/D28286
* Diff 28347 "Port services and shell runner away from KRun" [Needs
Review] https://phabricator.kde.org/D28347
[11:49:29] <kbroulik> meeting in 10'
[12:00:53] <DavidRedondo> Hello
[12:01:51] <d_ed> yo
[12:03:28] <bshah> helo
[12:03:46] <kbroulik> hello
[12:05:59] <kbroulik> nicofee: jgrulich fvogt ahiemstra[m]
[12:06:12] <kbroulik> ok, let's go
[12:06:15] <kbroulik> DavidRedondo:
[12:06:19] <ahiemstra[m]> yes hi hello
[12:06:28] <notmart> here-ish
[12:06:52] <fvogt> Moin!
[12:07:08] <DavidRedondo> I am continuing work for coloring icons in the correct colors
[12:07:31] <d_ed> the window decos? or more?
[12:07:38] <DavidRedondo> Patch for breeze kdecoration now up on phab
[12:07:54] <DavidRedondo> IconItem is fixed already
[12:08:40] <DavidRedondo> Somebody reported to me a bug in the systray because they thought it was also coloring bug
[12:09:02] <DavidRedondo> Turns out our sni implementation was bad
[12:09:39] <DavidRedondo> The data engine does complicated icon logic which breaks and has weird interaction with plasma iconitem
[12:10:23] <DavidRedondo> Patch ono phab to expose only the raw properties in the engine and do the rendering in the actual implementation
[12:10:28] <DavidRedondo> https://phabricator.kde.org/D28208
[12:10:28] <sKreamer> Diff 28208 "Move sni icon handling logic from data engine to applet" [Needs Review] https://phabricator.kde.org/D28208
[12:10:52] <DavidRedondo> Question ?
[12:10:54] <DavidRedondo> * Questions ?
[12:11:09] <d_ed> dataengines are supposedly stable API
[12:11:30] <kbroulik> d_ed: yeah that was my main concern also but I think we all involved agreed it was an acceptable break for the common good™
[12:11:42] <DavidRedondo> Yeah that's why I didn't actually remove the properties for now
[12:11:59] <d_ed> ok, cool
[12:12:02] <DavidRedondo> So we could keep them around if we want to
[12:12:10] <DavidRedondo> And just ignore them in the applet
[12:12:14] <d_ed> nah, I'm fine with killing all dataengines with fire :)
[12:12:24] <d_ed> as long as we're aware of everything
[12:12:52] <DavidRedondo> We asked latte if they use it and they just use the systray applet
[12:13:15] <kbroulik> good stuff :) alright, d_ed is next
[12:14:04] <d_ed> I've been working on two big things
[12:14:24] <d_ed> plasma getting a systemd powered startup
[12:14:44] <d_ed> This is most going all well, all the long groundwork moving plasma-session meant it's all gone pretty smoothly
[12:15:05] <d_ed> Also I've been doing some big changes to ksysguard
[12:15:15] <d_ed> last year we introduced plugins for processes, which worked quite well
[12:15:25] <d_ed> so I've done the same for the global stats
[12:15:30] <d_ed> with a new API and all
[12:15:41] <d_ed> these are some big reviews, so we need some volunteers on that
[12:16:40] <kbroulik> cool cool, questions=?
[12:16:45] <d_ed> In terms of my usual frustrated rant:
[12:16:55] <d_ed> This bug list of "very high priority bugs"
[12:16:57] <d_ed> https://bugs.kde.org/buglist.cgi?cmdtype=dorem&list_id=1729437&namedcmd=VHI-priority%20Plasma%20bugs&remaction=run&sharer_id=208174
[12:16:57] <sKreamer> KDE bug 208174 in Planet KDE (general) "broken Subject in planetkde" [normal,FIXED] https://bugs.kde.org/show_bug.cgi?id=208174
[12:17:02] <d_ed> seems to not have moved at all
[12:17:07] <d_ed> like last week
[12:17:43] <DavidRedondo> I think the breeze gtk thing is moving
[12:18:00] <kbroulik> I've got stuck on the pinned apps moving
[12:18:13] <kbroulik> I'm pretty sure I know what's going on and what's causing it, fixing it is annoying
[12:18:23] <d_ed> ok, that's good to hear
[12:18:45] <d_ed> make sure to keep bugs up to date with findings
[12:18:59] <kbroulik> I wonder if Bug 418466 is also a side-effect of that ghns changedentrieschanged emissions bug
[12:18:59] <sKreamer> KDE bug 418466 in systemsettings (kcm_desktoptheme) "Intalled Global themes not showing up in Setting's Global theme section." [normal,] https://bugs.kde.org/show_bug.cgi?id=418466
[12:19:22] <d_ed> I hereby task you with finding out !
[12:19:36] <kbroulik> ok
[12:19:50] <d_ed> and I'm done
[12:21:22] <kbroulik> alright, I have bshah next
[12:22:44] <bshah> moment
[12:23:19] <bshah> I have been mostly busy with : https://phabricator.kde.org/D28427 and https://phabricator.kde.org/D28428
[12:23:19] <sKreamer> Diff 28427 "[libnotificationmanager] forward notifications to lockscreen" [Needs Review] https://phabricator.kde.org/D28427
[12:23:19] <sKreamer> Diff 28428 "[screenlocker] listen to notifications from the notification server" [Needs Review] https://phabricator.kde.org/D28428
[12:23:38] <bshah> There's some more code but it is not polished enough to be public :P
[12:24:00] <bshah> Apart from it my last week hasn't been that much productive.
[12:24:13] <d_ed> is that stuck on anything needing discussion?
[12:24:14] <bshah> questions?
[12:24:24] <bshah> d_ed: nope, not really
[12:24:28] <ahiemstra[m]> how generic is that?
[12:24:33] <kbroulik> not very iirc
[12:24:40] <kbroulik> we've discussed making it generic but we kinda coujldn't agree on a scope
[12:24:43] <kbroulik> and security model
[12:24:45] <ahiemstra[m]> as in, would it also support displaying notifications on the normal lockscreen?
[12:25:07] <bshah> ahiemstra[m]: it could, yes. But I am test-driving it on mobile greeter first
[12:25:11] <ahiemstra[m]> right
[12:25:12] <bshah> then someone can add it in desktop greeter ;)
[12:25:38] <ahiemstra[m]> sure
[12:25:57] <bshah> there's bit we (I and Kai) discussed about notification handler
[12:26:09] <bshah> which is kinda neat, but then we didn't find a right place to put this as API
[12:26:19] <bshah> and then my nice plan was killed :'(
[12:26:23] <bshah> :P
[12:26:36] <kbroulik> yeah I would have loved a generic notification proxy thing also for kde connect
[12:26:51] <kbroulik> but whenever we started discussiont it, the scope went out of hand and then it ended up nowhere for being too large
[12:27:07] <kbroulik> so I'd say we focus on this specific thing first, make it good, and then see what we need in the future
[12:27:37] <bshah> (I still have some notes about handler stuff locally, I can potentially start discussion on p-d)
[12:27:39] <ahiemstra[m]> TODO: far in the future ;)
[12:27:53] <kbroulik> bshah: yes, please do
[12:28:01] <kbroulik> also make sure to include kdeconnect in the mix :)
[12:28:11] <kbroulik> alright, ahiemstra[m] is next
[12:28:32] <ahiemstra[m]> ok
[12:29:18] -*- fvogt goes to lunch now, but would only ask about the state of Qt 5.15 support anyway
[12:29:23] <ahiemstra[m]> I finished up the shadowedtexture last week, that's ready to be landed (https://phabricator.kde.org/D28222)
[12:29:23] <sKreamer> Diff 28222 "Add ShadowedImage and ShadowedTexture items" [Accepted] https://phabricator.kde.org/D28222
[12:29:41] <kbroulik> fvogt: what about it?
[12:29:46] <kbroulik> apart from deprecation porting
[12:29:54] <d_ed> fvogt: As I understand it everything is all ok
[12:30:18] <ahiemstra[m]> I also did a follow up to have separate corner radii: https://phabricator.kde.org/D28314
[12:30:18] <sKreamer> Diff 28314 "Allow setting separate corner radii for ShadowedRectangle" [Needs Review] https://phabricator.kde.org/D28314
[12:30:18] <d_ed> I should upgrade my laptop to latest Qt5.15 branch to be sure there's been nothing last minute
[12:30:22] <fvogt> There were some reports about severe instability on bugzilla, but openQA couldn't confirm that
[12:31:14] <ahiemstra[m]> that's all pretty much done, which means we can look at improving kirigami's card after that
[12:31:32] <kbroulik> ahiemstra[m]: I always love the grouped property apis you do :) too bad one can't have a variant property where you can set an int or address individual properties inside
[12:31:42] <ahiemstra[m]> I've done some small updates on the ksysguard library work that supports d_ed's daemon
[12:31:55] <ahiemstra[m]> yeah
[12:32:32] <kbroulik> ok, cool, notmart is next
[12:32:50] <ahiemstra[m]> :p
[12:32:51] <ahiemstra[m]> I wasn't done yet
[12:32:54] <kbroulik> sorry
[12:33:10] <ahiemstra[m]> last thing I wanted to mention is that I'm looking into https://bugreports.qt.io/browse/QTBUG-49646
[12:33:25] <kbroulik> fuck yes!
[12:33:28] <ahiemstra[m]> which I hope may be less work than I initially thought
[12:33:46] <kbroulik> pretty sure VDG people will worship you when you pull that off
[12:34:15] <notmart> ok
[12:34:37] <notmart> oh, indeed
[12:34:56] <ahiemstra[m]> since it only occurs with `horizontalAlignment: Center`
[12:35:12] <kbroulik> qRound(x()), done? :D
[12:36:07] <ahiemstra[m]> in theory :D
[12:36:13] <ahiemstra[m]> ok now I am done ;)
[12:36:24] <kbroulik> ok, thanks, notmart now
[12:36:53] <notmart> # Kirigami
[12:36:54] <notmart> * fixed a freeze while flicking on mobile
[12:36:56] <notmart> * finished overlaysheet refactor
[12:36:57] <notmart> # Plasma Bigscreen
[12:36:59] <notmart> * last fixes for the beta
[12:37:00] <notmart> * beta released!
[12:37:02] <notmart> * wrote a couple of blog posts about it (second i'll publish today) https://notmart.org/blog/2020/03/plasma-bigscreen
[12:37:03] <notmart> * started making mycroft fully optional
[12:37:15] <notmart> let's see if i'll be able of doing some more plasma-plasma things now
[12:38:09] <kbroulik> I've did an experiment of using that color detector thing from bigscreen to colorize my task bar entries by app icon
[12:38:16] <kbroulik> it sounded nicer in my head than it now actually looks :)
[12:38:44] <notmart> eh, it's the typical cool thing that has to be used very with care
[12:38:48] <kbroulik> yeah
[12:38:59] <veggero> <kbroulik "I've did an experiment of using "> That's interesting though, maybe it could be worth a VDG discussion
[12:39:24] <notmart> with great power comes great current squared times resistance
[12:39:48] <kbroulik> veggero: I just did a crappy "colorize the svg thing". I recall icontasks in kde4 had a nice glow and stuff where the color looked really nice
[12:40:02] <kbroulik> maybe works better for icontasks.. anyway i'd love that color detector somewhere in a usable lib :P
[12:40:37] <kbroulik> alright, anyone else wants to add something, otherwise I'll go and wrap up the meeting?
[12:41:11] <kbroulik> ok, my notes:
[12:41:15] <kbroulik> Notifications:
[12:41:15] <kbroulik> * Continued discussion about inline-reply spec on XDG...
[12:41:15] <kbroulik> * Added `KNotificationJobUiDelegate`, please review: D28295
[12:41:15] <kbroulik> * Allows to use KNotification for job error reporting rather than a modal message box
[12:41:15] <kbroulik> * Useful for the new `ApplicationLauncherJob` replacement for `KRun` so apps launched from Plasma use a nice Plasma error notification
[12:41:16] <sKreamer> Diff 28295 "Introduce KNotificationJobUiDelegate" [Needs Review] https://phabricator.kde.org/D28295
[12:41:22] <kbroulik> plasma-firewall:
[12:41:22] <kbroulik> * Made `KSortFilterProxyModel` `invalidateFilter` public
[12:41:22] <kbroulik> * Eases creation of QML-only custom view filtering
[12:41:22] <kbroulik> * Fixed size propagation bug in `FormLayout`
[12:41:22] <kbroulik> * Did you know QtQml `Timer` default `interval` is 1000? :)
[12:41:22] <kbroulik> * Some minor UI improvements
[12:41:28] <kbroulik> Other:
[12:41:28] <kbroulik> * Fixed virtual desktop pager not updating rows when only rows changed but not the number of desktops
[12:41:28] <kbroulik> * Made a bodge fix for wallpaper thumbnails not loading: D28420
[12:41:28] <kbroulik> * Caused by GHNS Button signalling a change when there isn't. See commit description in the phab for what's going on
[12:41:31] <kbroulik> * Added support for launching `KServiceActions` in `ApplicationLauncherJob`
[12:41:31] <kbroulik> * This will give us a proper API for launching jump list actions rather than just running the action `Exec` as command
[12:41:31] <kbroulik> * Patch for Task Manager D28286 and services/shell runner D28347
[12:41:37] <sKreamer> Diff 28420 "[Image Wallpaper] Fix thumbnail generation when model is reloaded in-flight" [Needs Review] https://phabricator.kde.org/D28420
[12:41:46] <sKreamer> Diff 28286 "[Task Manager] Port backend to ApplicationLauncherJob" [Needs Review] https://phabricator.kde.org/D28286
[12:41:46] <sKreamer> Diff 28347 "Port services and shell runner away from KRun" [Needs Review] https://phabricator.kde.org/D28347
[12:41:49] <notmart> i wouldn't like a clone of the windows7 taskbar look with that silly colorized gradient on mouse over
[12:42:43] <d_ed> yes, I have two more things I want to discuss
[12:42:46] <ahiemstra[m]> what they do in win10 is even worse, it's basically that effect but only on an outline
[12:42:58] <d_ed> That I just remembered
[12:43:17] <kbroulik> d_ed: go ahead
[12:43:18] <d_ed> we had a "security" ish, issue last week with .desktop files having an Exec line being used in KRun(someUrl)
[12:43:56] <d_ed> whilst we're doing the Application/Command job porting and looking through usages of KRun it's worth checking any that use the KRun(someUrl) constructor and possibly making them not execute things
[12:43:57] <DavidRedondo> d_ed: but that's the same with clicking on said file right
[12:44:24] <kbroulik> DavidRedondo: when clicking on it in dolphin it asks
[12:44:31] <kbroulik> recent docs runner would just blatantly execute
[12:44:32] <d_ed> KRun(someUrl) also asks
[12:44:43] <d_ed> if not executable
[12:45:19] <kbroulik> also, if possible use service explicitly with URL
[12:45:29] <d_ed> yeah, that too
[12:45:34] <kbroulik> e.g. that's what I added KServiceAction for applicationlauncherjob, so we don't exec a command but invoke the "proper service"
[12:45:51] <d_ed> DavidRedondo: so you're right it wasn't super critical or easy to exploit, but it's still worth locking down what we can
[12:46:10] <kbroulik> ok, we're done now I guess?
[12:46:28] <d_ed> my other thing I wanted to discuss - if lots of people are on Corona lockdown it might be a good opportunity for some onboarding
[12:46:31] <ismailof[m]> @kbroulik asked me to discuss a bit about the kwinrules kcm?
[12:46:32] <d_ed> making the best of a bad situation
[12:46:49] <ismailof[m]> https://phabricator.kde.org/D28152
[12:46:49] <sKreamer> Diff 28152 "[WIP] KWinRules KCM Redesign" [Needs Review] https://phabricator.kde.org/D28152
[12:46:54] <d_ed> so I was thinking we could organise some "bug day" or day-long virtual sprint or something...
[12:46:56] <notmart> indeed
[12:47:00] <d_ed> I haven't really thought through past that
[12:47:15] <kbroulik> d_ed: sounds good
[12:47:18] <DavidRedondo> Speaking of corona, because of sprints being canceled I had the idea maybe we could arange topic calls or something if needed?
[12:47:35] <d_ed> DavidRedondo: ack. I think we need a wayland call at some point after that long ML thread
[12:47:57] <bshah> Virtual sprint? :D
[12:48:08] <d_ed> so we have two somewhat distinct things - calls to replace our real sprint, and onboarding type "sprints"
[12:48:10] <notmart> eh, let's di it
[12:48:16] <DavidRedondo> dfaure is the expert for that I think :D
[12:48:20] <kbroulik> :D
[12:48:43] <kbroulik> ismailof[m]: hello :)
[12:48:51] <kbroulik> is there something you're stuck on with your rules kcm?
[12:48:58] <d_ed> ismailof[m]: you've done a great job on that btw
[12:49:02] <kbroulik> I've been poking some devs occasionally to keep the momentum going
[12:49:36] <ismailof[m]> not really stuck, just being conservative about when "wip" ends :D
[12:50:09] <d_ed> ismailof[m]: given it's an advanced user KCM, I think in terms of the UI/UX you're hitting diminishing returns with changes
[12:50:09] <ahiemstra[m]> yeah it's really cool
[12:50:13] <kbroulik> just be mindful of the 5.19 freezes early may
[12:50:38] <kbroulik> so perhaps merge soon-ish and then clean it up until the beta? :)
[12:51:49] <kbroulik> I just want to avoid someone going a day before the Beta "oh this is so cool, can we have this in" so i'd rather merge sooner than later
[12:52:10] <ismailof[m]> alright! I'll take today for some polishing maybe, and set it ready for review
[12:52:18] <kbroulik> lovely
[12:52:21] <kbroulik> good job
[12:52:23] <kbroulik> ok, I think we're done now
