freemangordon | Wizzup: do we have any idea what needs to be done to fix qt white background? | 09:05 |
---|---|---|
freemangordon | uvos: ^^^ | 09:05 |
Wizzup | I don't really know | 10:17 |
freemangordon | ok. it seems QAbstractScrollArea has wrong pallette | 12:12 |
Wizzup | freemangordon: super great if you figure it out | 12:19 |
freemangordon | I am on it | 12:19 |
freemangordon | Wizzup: drawControl does not get called when style == maemo5 | 14:29 |
freemangordon | trying to understand why | 14:29 |
Wizzup | freemangordon: maybe a function signature mismatch or override missing? | 14:33 |
Wizzup | http://man.hubwiz.com/docset/Qt_5.docset/Contents/Resources/Documents/doc.qt.io/qt-5/qtwidgets-widgets-styles-example.html this seems to have an example for it | 14:33 |
freemangordon | yes, Q_DECL_OVERRIDE is missing | 14:33 |
freemangordon | trying with it ATM | 14:34 |
freemangordon | nope | 14:36 |
freemangordon | hmm, QFrame::paintEvent is not called | 14:38 |
Wizzup | is our paintevent virtual or real? | 14:45 |
freemangordon | we don;t havew one | 14:46 |
freemangordon | it does not get to here https://codebrowser.dev/qt5/qtbase/src/widgets/widgets/qabstractscrollarea.cpp.html#956 | 14:46 |
Wizzup | that is a weird looking case stmt | 14:47 |
freemangordon | because of the local variable | 14:48 |
freemangordon | but that's not the issue | 14:48 |
Wizzup | does it get to the paintevent part at least | 14:48 |
freemangordon | no | 14:48 |
freemangordon | I guess we override something in such a way that we stop the event | 14:48 |
freemangordon | will start commenting the code to see what breaks it | 14:49 |
Wizzup | hm | 14:50 |
Wizzup | could be event filter | 14:51 |
freemangordon | mhm | 14:51 |
freemangordon | where it is? | 14:51 |
Wizzup | it is in gtk2, not mameo5, so probably not | 14:52 |
freemangordon | ok | 14:52 |
Wizzup | I read online somewhere that repaint() will call paint event, and update() might not | 14:52 |
Wizzup | not sure if this is it | 14:52 |
freemangordon | hmm, I think I made some progress | 14:52 |
freemangordon | gimme some time to see what exactly I found | 14:52 |
Wizzup | ok | 14:52 |
Wizzup | oh, ignore my thing about update vs repaint | 14:53 |
freemangordon | commenting the code https://github.com/maemo-leste/qtstyleplugins/blob/master/src/plugins/styles/maemo5/maemo5style/qmaemo5style.cpp#L511 makes event being generated | 14:55 |
freemangordon | and drawControl called | 14:55 |
freemangordon | but background is wrong | 14:56 |
freemangordon | so it is not that | 14:57 |
freemangordon | sorry, false alarm, missing call to drawControl is because maemo forces no frame | 15:00 |
freemangordon | Wizzup: this https://github.com/maemo-leste/qtstyleplugins/blob/master/src/plugins/styles/maemo5/maemo5style/qmaemo5style.cpp#L148 | 16:26 |
freemangordon | for some reason it seems qlistview is using standard palette | 16:27 |
Wizzup | hmmmm | 16:28 |
Wizzup | where is this getTextColorWidget called? | 16:28 |
freemangordon | from gtk style | 16:32 |
freemangordon | Wizzup: there was one more issue actually, QMaemo5StylePrivate::applyCustomPaletteHash was not called, because applyCustomPaletteHash in parent class was not virtual | 16:33 |
freemangordon | but even after fixing it, it still does not work properly | 16:34 |
freemangordon | albeit it does qApp->setPalette(gtkWidgetPalette("HildonPannableArea.GtkTreeView"), "QAbstractScrollArea"); | 16:34 |
Wizzup | should that be GtkListView instead of GtkTreeView? probably does not matter | 16:36 |
Wizzup | sounds like you're getting closer actually | 16:36 |
freemangordon | yes, it does not | 16:36 |
freemangordon | mhm | 16:36 |
freemangordon | for some reason no custom palette is applied to QAbstractScrollArea | 16:37 |
freemangordon | Wizzup: hmm, actually it seems to be a bug in gtk style plugin | 16:48 |
freemangordon | gtkWidgetPalette does not set QPalette::Base | 16:49 |
Wizzup | arg... | 16:52 |
freemangordon | with that fixed, it seems to work properly, lemme make sure | 16:53 |
freemangordon | YAY! | 17:08 |
freemangordon | lemme push the fixes | 17:08 |
freemangordon | phew, this one was driving me nuts for long time | 17:21 |
Wizzup | great.... :) | 17:33 |
Wizzup | I will really try to have news post this weekend | 17:33 |
Wizzup | we should also let pavel know | 17:36 |
Wizzup | he asked @ fosdem | 17:36 |
Wizzup | :p | 17:36 |
freemangordon | 'know'? | 17:45 |
freemangordon | what about? | 17:45 |
Wizzup | freemangordon: god this is so much better | 19:29 |
Wizzup | freemangordon: pavel asked about the qt theming bug | 19:29 |
Wizzup | freemangordon: in calendar the textarea is still white on black, as opposed to black on white (go to tasks or notes and then use new note or new task) | 19:55 |
aczid | hi, I'm not sure if this is the right place to ask such a noob question, but: I really miss the keyboard shortcut ctrl+backspace to go to the overview of running apps. was it removed intentionally or am I running in to a bug, and how do I get it back? | 20:04 |
aczid | otherwise as a user after not playing with maemo-leste for almost 2 years it's really impressive how much better everything has gotten! respect | 20:05 |
Wizzup | is this on the droid 4, or n900? | 20:05 |
aczid | droid4 | 20:05 |
Wizzup | I think the reason that is doesn't work is because it is mapped to the 'home' icon on your touch screen | 20:06 |
Wizzup | and we probably don't allow multiple keys to be mapped to one action | 20:06 |
Wizzup | (just a guess) | 20:06 |
aczid | where does that shortcut mapping happen anyway? | 20:07 |
Wizzup | it is /usr/share/hildon-desktop/shortcuts.ini.leste | 20:08 |
Wizzup | well, without the .leste | 20:08 |
Wizzup | there is also this btw (WIP) http://maemo-leste-manual.motionlibre.org/index.html | 20:08 |
Wizzup | http://maemo-leste-manual.motionlibre.org/chapters/ui.html these are more or less finished | 20:08 |
aczid | thanks I figured it would be hildon but I only grepped around in /etc/ for it | 20:09 |
aczid | ah great | 20:09 |
aczid | didn't find that anywhere yet (github/wiki) so much appreciated. looks great | 20:10 |
sicelo | aczid: did it ever work on Droid 4? or you're an ex-N900 user? | 20:11 |
Wizzup | yeah the doc is still wip and not really publicized | 20:11 |
aczid | sicelo: it did work on the droid4 with a quite old version of leste before I dist-upgraded, and yes I used the n900 before | 20:11 |
aczid | could very well be that that version didn't have the on screen buttons yet, I first noticed them after | 20:12 |
aczid | so I'm optimistic that Wizzup's intuition is right | 20:12 |
arno11 | Wizzup: did you see my messages about 44100/48000hz ? | 20:17 |
arno11 | (because we should keep the default 44100hz rate) | 20:19 |
arno11 | and i think we can use only one asound.state for the moment | 20:20 |
arno11 | the custom one works well with calls, video and music with no troubles (but at 44100hz) | 20:22 |
arno11 | ATO i didn't find a way to use the earphone... | 20:24 |
Wizzup | arno11: I will have to check re 44.1 | 20:34 |
Wizzup | but yes, we can never have one asound state | 20:34 |
Wizzup | this is why we need ucm | 20:34 |
Wizzup | sphone will already switch to HiFo or VoiceCall profiles if they are defined | 20:35 |
Wizzup | HiFi* | 20:35 |
arno11 | ah cool | 20:35 |
arno11 | and did you try 2G calls (just to compare with my device) ? | 20:36 |
Wizzup | not yet | 20:39 |
Wizzup | at least I think I was on 3G | 20:39 |
arno11 | ok | 20:39 |
Wizzup | I agree with keeping default rate, so we do not need 48000 at all? | 20:40 |
arno11 | exactly | 20:40 |
arno11 | thanks to soxr resampling | 20:40 |
Wizzup | so I guess there are few things remaining: | 20:54 |
Wizzup | - your work on cmtspeech to use the right VoiceCall state | 20:55 |
Wizzup | - UCM | 20:55 |
Wizzup | - module loading (either loading it at the right time or fixing the module to probe later) | 20:55 |
Wizzup | for now we can make an init script that loads the module | 20:55 |
Wizzup | the UCM is more tricky I think | 20:55 |
arno11 | yes agree | 21:06 |
freemangordon | Wizzup: ok, will have a look | 21:41 |
freemangordon | Wizzup: sorry, I don;t understand what the issue is | 21:43 |
freemangordon | oh | 21:44 |
freemangordon | got it | 21:44 |
Wizzup | freemangordon: yeah, for the larger than one line text field | 21:45 |
freemangordon | but was it ok before my changes? | 21:47 |
Wizzup | I don't remember | 21:50 |
Wizzup | let me see if I can figure it out | 21:50 |
Wizzup | looks like it, yes | 21:51 |
Wizzup | at least on my beowulf bionic | 21:51 |
Wizzup | let me check my chimaera n900 | 21:51 |
freemangordon | hmm | 21:51 |
Wizzup | hum, wait | 21:53 |
Wizzup | I am not sure if I can still reproduce it | 21:53 |
Wizzup | it looks fine on my d4 with latest packages now (?) | 21:53 |
Wizzup | what the | 21:53 |
freemangordon | I can | 21:53 |
freemangordon | which is weird | 21:54 |
Wizzup | could it be some graphical glitch? | 21:57 |
Wizzup | restarted it, now I see it black again | 21:57 |
freemangordon | no, I can repro on both d4 and VM | 21:57 |
Wizzup | (it = qalendar) | 21:58 |
freemangordon | mhm | 21:58 |
freemangordon | weird | 21:58 |
freemangordon | maybe a bug in qtcalendar | 21:59 |
Wizzup | could be, but I don't think it happened before | 21:59 |
freemangordon | because the same widget looks fine in separate project (QTextEdit) | 21:59 |
Wizzup | maybe it depends on the widgets used/loaded by a qt project | 22:00 |
freemangordon | or the order | 22:00 |
Wizzup | yeah | 22:02 |
freemangordon | but it is steady reproducible here | 22:02 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!