=== racarr|desert is now known as racarr | ||
Mirv | if any Unity8 devs around, filed a stack publishing preventing bug https://bugs.launchpad.net/unity8/+bug/1220057 | 05:51 |
---|---|---|
ubot5 | Launchpad bug 1220057 in Unity 8 "Failing autopilot test EphemeralNotificationsTests.test_urgency_order" [Critical,New] | 05:51 |
dednick | mzanetti: ping | 07:18 |
mzanetti | dednick: good morning | 07:18 |
dednick | mzanetti: howdy. | 07:18 |
dednick | mzanetti: do you know much about object ownership changes passing between c++ & qt? | 07:19 |
mzanetti | dednick: you mean QML? | 07:19 |
dednick | sorry, yes | 07:19 |
dednick | mzanetti: as in passing a (new QObject) back into qml. | 07:20 |
mzanetti | dednick: there are some QML components which take ownership of the objects if they don't have another parent yet | 07:20 |
dednick | mzanetti: is that only with QMLComponent, or QQuickItem as well? | 07:20 |
mzanetti | QQuickItems as well. | 07:21 |
mzanetti | dednick: so if you just do a new QObject() and put it somewhere into a view, the view will set itself as parent | 07:21 |
dednick | mzanetti: QObject as well? | 07:22 |
dednick | doesnt seem to be happening for me :( | 07:23 |
dednick | not even with a QQuickItem | 07:23 |
dednick | mzanetti: do you know it it works with qml properties ? | 07:26 |
dednick | mzanetti: like "property Item action: menu.getAction(...)" | 07:27 |
dednick | where getAction returns new object | 07:27 |
mzanetti | dednick: no. in that case I don't think its reparented | 07:28 |
dednick | mzanetti: hm. ok | 07:28 |
dednick | mzanetti: thanks. guess that would be my problem then :) | 07:28 |
larsu | dednick: morning :) | 07:28 |
dednick | larsu: hey :) | 07:28 |
dednick | larsu: just talking about your code. | 07:29 |
larsu | dednick: yep, just read that | 07:29 |
mzanetti | dednick: you need it to be reparented? | 07:29 |
larsu | dednick: so the problem is basically that I'm parenting the actions to the model? | 07:29 |
dednick | larsu: yep. but also the way we work it in qml using a property doesnt change the owner of the object | 07:30 |
dednick | mzanetti: yep | 07:30 |
larsu | dednick: but surely it does js-style ref counting? | 07:30 |
dednick | larsu: apparently not. | 07:31 |
dednick | larsu: i removed the parent in the function, but they're not being deleted. | 07:31 |
mzanetti | dednick: you really sure about that? I mean... if c++ code does a "new"... you really want it to "get lost" in javascript? | 07:31 |
dednick | larsu: the js counting only works within js functions i think | 07:31 |
larsu | dednick: ah, this is what I'm trying right now as well. Sigh. | 07:32 |
dednick | mzanetti: yeah, i would prefer that we left delete responsibility to what creates them, but it sticks around forever. | 07:32 |
mzanetti | dednick: well. if its about the cleanup you could do something like this: MyQMLItem { Componet.onDestruction: myCppObj.destroy() } | 07:33 |
mzanetti | which will call deleteLater() on QObject iirc | 07:33 |
dednick | mzanetti: the property can change as well. so those will be lost. | 07:34 |
larsu | dednick: I guess we need something like this: https://qt-project.org/doc/qt-5.0/qtqml/qqmlengine.html#setObjectOwnership | 07:36 |
mzanetti | interesting... | 07:37 |
dednick | larsu: http://qt-project.org/wiki/SharedPointersAndQmlOwnership . but i dont think it applies in this case | 07:37 |
dednick | larsu: as far as i've read Q_INVOKABLE funcions have the ownership set to JavaScriptOwnership, but maybe that isnt the case... | 07:39 |
larsu | dednick: right, so we shouldn't have to call setOwnership explicitely | 07:40 |
dednick | larsu: from my understanding | 07:41 |
dednick | larsu: guess we could give it a go and see if it changes things | 07:41 |
dednick | larsu: maybe using them on a property setter is behaving differently under this circumstance. | 07:41 |
larsu | dednick: I'm trying right now | 07:41 |
larsu | dednick: nope, same problem | 07:43 |
dednick | larsu: ok, i think having the index property may be the best option. Then we dont have to worry about ownership. | 07:44 |
dednick | since we can just create the object in qml | 07:44 |
dednick | larsu: the index will be bound to the modelIndex, so changes to that will need to sync down to the registration. | 07:45 |
dednick | stupid custom actions. | 07:49 |
larsu | dednick: stupid non-existent ref-counting :P | 07:50 |
greyback | mzanetti: hey, using the app api in unity-api, found something that's problematic with it. Question: do you ever call the get(index) method to get a raw ApplicationInfo object? | 08:24 |
mzanetti | greyback: I don't think I need it, no | 08:25 |
greyback | mzanetti: because I'm wondering why we have that ApplicationInfo class at all. I'm thinking it should be an implementation detail, not something that needs to be given to QML in any way (which get() does) | 08:25 |
mzanetti | greyback: why that? | 08:25 |
greyback | mzanetti: well get() returns the ApplicationInfo object. The only place that's needed is for focusApp() and stopApp(). Why not instead just always use appId as the unique identifier, not at times using a appId, other times using ApplicationInfo* | 08:26 |
mzanetti | greyback: so if you want to pass stuff on to somewhere else you need to pass all the 5 strings separately | 08:27 |
greyback | mzanetti: do you need to do that? As I don't | 08:28 |
mzanetti | greyback: actually I intended to pass the ApplicationInfo object to the launcher plugin and have all the data in there. without the need of starting to parse .dekstop files again | 08:28 |
mzanetti | greyback: not really the get() method, but as parameter in applicationFocused | 08:28 |
greyback | mzanetti: ok I see what you're doing | 08:28 |
nic-doffay | mzanetti, I'm not really sure what you mean by this: * it breaks the narrowMode decision. Try opening the search field and drag it smaller/bigger and compare that to trunk | 08:28 |
greyback | mzanetti: yes then you find it useful. Okay, will think a bit more | 08:29 |
mzanetti | greyback: not that I would _really_ need it. but it feels like a good idea to minimize parsing the file system | 08:30 |
mzanetti | nic-doffay: just try it :) | 08:30 |
greyback | mzanetti: no true. Having 1 object that has all the data from the desktop file ready internally, that can be shared around, makes sense | 08:31 |
mzanetti | greyback: which still leaves one question open | 08:32 |
mzanetti | greyback: I'd need to link the applicationinfo implementation in the launcher plugin | 08:32 |
mzanetti | not really sure yet how to do that | 08:32 |
greyback | mzanetti: yep. Is do-able, will see once implementation done & working. | 08:37 |
sil2100 | bregma: hi! | 08:37 |
sil2100 | bregma: we have a serious unity AP problem that's blocking the unity from releasing: https://bugs.launchpad.net/unity/+bug/1219636 | 08:38 |
ubot5 | Launchpad bug 1219636 in Unity "unity autopilot tests fails to move mouse, leading to infinite test run" [Critical,New] | 08:38 |
sil2100 | bregma: it most probably doesn't happen all the time, but frequently | 08:38 |
larsu | dednick: ah there you are. I'm almost done rewriting that patch. | 08:54 |
dednick | larsu: cool | 08:54 |
nic-doffay | mzanetti, any idea who would know more about those icon themes? | 09:47 |
nic-doffay | I'd like to get to the bottom of why certain icons aren't available for me. | 09:47 |
larsu | dednick: something's totally off now, UnityMenuAction instances aren't even created for me anymore :( | 09:50 |
larsu | nic-doffay: it looks like the gicon image provider has been removed?! | 09:50 |
* larsu only judges by all the warnings he's seeing | 09:51 | |
dednick | larsu: can you push your branch to junk or something? | 09:51 |
nic-doffay | larsu, there are so many image related warnings I lose track >_< | 09:51 |
larsu | nic-doffay: right :( Let's fix them! | 09:51 |
nic-doffay | larsu, some icons are still available though... | 09:51 |
nic-doffay | eg "search" is available. | 09:52 |
larsu | dednick: I'm doing a clean rebuild now, maybe I messed something up. Will push after that | 09:52 |
nic-doffay | larsu, any idea who to talk to about this? | 09:54 |
larsu | nic-doffay: I'm afraid that would be me | 09:54 |
larsu | I'll be looking into this after I fixed that issue for dednick | 09:54 |
nic-doffay | larsu, cool :P | 09:56 |
mzanetti | nic-doffay: was in a meeting... | 10:00 |
nic-doffay | mzanetti, no worries | 10:00 |
mzanetti | nic-doffay: well... what does dpkg -L ubuntu-mobile-icons give you? | 10:00 |
nic-doffay | mzanetti, that's the thing a lot of icons. | 10:00 |
nic-doffay | mzanetti, majority of those work. | 10:02 |
nic-doffay | filters didn't though. | 10:02 |
nic-doffay | (ive tried skype and other random picks before) | 10:02 |
mzanetti | nic-doffay: and is the "filter" icon listed in there? | 10:02 |
nic-doffay | mzanetti, it's not. | 10:03 |
mzanetti | dednick: way better :) | 10:03 |
nic-doffay | But I was wondering how you got it? | 10:03 |
larsu | dednick: I pushed it to lp:~/qmenumodel/use-full-action-name. | 10:03 |
mzanetti | nic-doffay: apt-cache policy ubuntu-mobile-icon | 10:03 |
mzanetti | nic-doffay: apt-cache policy ubuntu-mobile-icons | 10:03 |
dednick | larsu: ta | 10:03 |
nic-doffay | mzanetti, mine's 13.04+13.10.20130812-0ubuntu1 | 10:04 |
larsu | dednick: there might be some bugs in there still, I can 't test it because I'm getting that qml error: | 10:04 |
dednick | mzanetti: you mean the checkbox? | 10:04 |
larsu | file:///home/lars/src/unity8/unitymenumodel.messages/builddir/plugins/Unity/Indicators/Messaging/MessageMenuItemFactory.qml:136: TypeError: Cannot read property 'valid' of null | 10:04 |
dednick | larsu: ok | 10:04 |
mzanetti | dednick: yes :) | 10:04 |
mzanetti | nic-doffay: please paste both, the complete output of "apt-cache policy ubuntu-mobile-icons" and "dpkg -L ubuntu-mobile-icons" | 10:06 |
mzanetti | nic-doffay: and your code where you try to use the Icon with the filter icon | 10:06 |
nic-doffay | mzanetti, the code is simple | 10:06 |
nic-doffay | I just reference use the icon name. | 10:06 |
nic-doffay | Without the type suffix. | 10:06 |
nic-doffay | That works for the other icons in that package. | 10:07 |
mzanetti | nic-doffay: I know its simple. yet you must have a mistake somewhere.... otherwise it would work | 10:07 |
larsu | nic-doffay: what are the exact icons uris you are using? | 10:07 |
mzanetti | larsu: the Icon {} component has some magic... it has a property "name" which refers to an icon name, withouth e.g. image:// and file type extension | 10:08 |
nic-doffay | mzanetti, here's everything. https://pastebin.canonical.com/96734/ | 10:09 |
larsu | mzanetti: I know, but it seems to be doing the right thing afaics | 10:09 |
nic-doffay | mzanetti, filter is there. | 10:09 |
mzanetti | nic-doffay: what happens if you remove the color property? | 10:10 |
nic-doffay | mzanetti, nothing. | 10:10 |
mzanetti | but I've changed the code here to be exactly this. and it works | 10:11 |
nic-doffay | mzanetti, changed it to exactly what? | 10:13 |
larsu | nic-doffay: QIcon::fromTheme() returns an empty icon for "filter" | 10:14 |
mzanetti | nic-doffay: http://paste.kde.org/p8a71429e | 10:14 |
larsu | nic-doffay: so it's not a problem with your code at all | 10:14 |
nic-doffay | larsu, yeah as I suspected. | 10:15 |
nic-doffay | But what differs between myself and mzanetti's setup? | 10:15 |
nic-doffay | This is what I'd like to find out. | 10:15 |
nic-doffay | larsu, I mean the filters icon is clearly visible after dpkg -L ubuntu-mobile-icons | 10:16 |
nic-doffay | And I've used other icons there as a test. | 10:16 |
larsu | nic-doffay: are you on the desktop or on the phone? Most likely, the theme name is wrong. | 10:16 |
larsu | ya, it works if I set the theme name manually | 10:17 |
larsu | unity8 should set this if the platform abstraction thingy doesn't | 10:18 |
larsu | dednick: are you seeing the same error? | 10:18 |
dednick | larsu: nope | 10:18 |
dednick | just testing on phone now | 10:19 |
larsu | dednick: hm. Any idea why I could be running into that? It makes it really hard to test :) (and there's no other error about not being able to load the UnityMenuAction type) | 10:19 |
dednick | larsu: you maybe left the UnityMenuActions as getAction in the messaging menu factor | 10:19 |
nic-doffay | larsu, currently off the desktop. | 10:19 |
larsu | dednick: no, I triple-checked and rebuilt the whole thing from scratch | 10:20 |
larsu | nic-doffay: there's your problem, then :) | 10:20 |
dednick | larsu: revert changes to unity8 and apply http://pastebin.ubuntu.com/6058425/ | 10:21 |
larsu | nic-doffay: changing your icon theme won't work though, because unity8 doesn't pick it up | 10:21 |
dednick | larsu: then ./build && ./run from source | 10:21 |
dednick | s/source/source dir | 10:21 |
larsu | nic-doffay: basically, someone needs to call QIcon::setThemeName("ubuntu-mobile") at some point | 10:22 |
larsu | dednick: ya, this is what I have as well | 10:22 |
dednick | larsu: hm. | 10:22 |
larsu | dednick: does sending replies work for you now? | 10:24 |
dednick | larsu: not sure how to test that without phone. which i'm still updating at the moment. | 10:25 |
dednick | larsu: it's getting through to the muxer at least. | 10:25 |
larsu | dednick: and you're not getting those error messages? /me is thoroughly confused | 10:26 |
dednick | larsu: but the UnityAction.isValid doesnt seem to be working. | 10:26 |
greyback | mzanetti: hey, I need your opinion. In shell, I want to use QSortFilterProxyModel to filter the running applications list to get list of running sidestage apps. If I want to focus one of those side-stage apps, right now I need its Application pointer. | 10:26 |
greyback | mzanetti: To get that I need to use get(i) where i is the index in the original model, not the filtered model. But I don't think we've implemented anything like the mapToSource() method on AbstractProxyModel | 10:27 |
greyback | mzanetti: Or I could change the api in unity-apis, so that focus uses the appId, not the Application pointer | 10:28 |
greyback | mzanetti: or I could add a role to the model which returns the Application pointer | 10:28 |
larsu | dednick: oh, so the UnityMenuAction _is_ being created, but the replyAction property is still null | 10:28 |
nic-doffay | mzanetti, can you clarify one of you review comments for me? | 10:29 |
mzanetti | greyback: roles that return pointers might be problematic | 10:30 |
greyback | mzanetti: yeah, it's my least favourite option | 10:30 |
dednick | larsu: i think your fullActionName may be acting up. the toUtf8() is returning a QByteArray, which i think is being destroyed after the call to constData(), so the mem is no longer valid. | 10:30 |
mzanetti | greyback: changing all api's to use appId except having a get(appId) to get the appInfo pointer sounds ok... | 10:30 |
larsu | dednick: ugh, very good point | 10:31 |
mzanetti | greyback: while I once did expose something like mapToSource() in another project and it still works fine, I think it's not very nice | 10:31 |
nic-doffay | mzanetti, with this comment: * the animation feels weird. It covers everything except the searchfield. Please check with design. | 10:31 |
mzanetti | nic-doffay: yeah... | 10:31 |
larsu | dednick: but that shouldn't lead to that error, should it? | 10:31 |
greyback | mzanetti: I wasn't looking forward to trying it, I'll admit that :) | 10:31 |
nic-doffay | Do you mean it's strange that the animation affects everything but the search bar? | 10:31 |
dednick | no | 10:31 |
mzanetti | nic-doffay: there is this thing coming down fro the top.. it goes below the textfield but above everything else | 10:31 |
mzanetti | nic-doffay: feels/looks really weird imho | 10:32 |
dednick | can you pastebin your unity8 source diff? | 10:32 |
greyback | mzanetti: appId it is. Thanks | 10:32 |
nic-doffay | mzanetti, right gotcha. | 10:32 |
nic-doffay | Having a chat about it now. | 10:32 |
dednick | larsu: ^ | 10:32 |
larsu | dednick: it's exactly what you pastebinned earlier | 10:32 |
larsu | dednick: the UnityMenuAction constructor is called, but none of the setters. Something must delete the object during construction... | 10:40 |
dednick | larsu: any other errors? | 10:40 |
larsu | dednick: a bazillion warnings about icons and the hud, but nothing about this | 10:41 |
dednick | larsu: my line 136 is "appIcon: menu && menu.ext.xCanonic..." | 10:41 |
larsu | dednick: because of your patch. It's line 139 now | 10:42 |
larsu | dednick: I get these errors for every occurence of replyAction | 10:43 |
dednick | larsu: try putting a Component.onCompleted with log message in the replyAction | 10:45 |
larsu | dednick: already did, doesn't get called | 10:45 |
dednick | larsu: bleh. hmm. ok, try just creating a QMenuMode.UnityMenuAction in the Factory root item. maybe it will log out an error if you create like that | 10:46 |
dednick | larsu: not using a property i mean | 10:46 |
larsu | dednick: ah, good idea | 10:46 |
dednick | larsu: but i really dont get why it's working on mine :/ | 10:46 |
greyback | mzanetti: http://pastebin.ubuntu.com/6058490/ is what will suit me. What do you think of line 30-31 of the diff, would that make sense? It would remove idea of model index entirely, so appId is the only unique identifier | 10:49 |
dednick | larsu: reply works on device now | 10:50 |
larsu | dednick: MessageMenuItemFactory.qml:35:5: Cannot assign object to list | 10:50 |
larsu | dednick: when try to just add a UnityMenuAction directly | 10:51 |
larsu | this is beyond weird | 10:51 |
dednick | larsu: yeah, sorry, put it in the loader | 10:51 |
greyback | as aside, this is interesting article on working around carriers not pushing OS updates: http://arstechnica.com/gadgets/2013/09/balky-carriers-and-slow-oems-step-aside-google-is-defragging-android/ | 10:51 |
larsu | dednick: ah, default property magic? | 10:51 |
dednick | larsu: there's some dodgeyness going on in the sdk listview items where you cant child qobjects | 10:51 |
dednick | larsu: yep | 10:52 |
larsu | dednick: no additional error :( | 10:53 |
dednick | no onCompleted log? | 10:54 |
dednick | larsu: is the destructor being called? | 10:54 |
dednick | really strange | 10:55 |
mzanetti | greyback: +1 on lines 30-31 | 10:55 |
greyback | mzanetti: ta. MR on it's way | 10:56 |
mzanetti | greyback: you sure its a good idea to remove the return value from startApplication? | 10:56 |
larsu | dednick: onCompleted is called, the destructor is not called | 10:56 |
mzanetti | greyback: might make sense to retain a bool? | 10:56 |
dednick | larsu: try reinstalling archive qmenumodel and see if it complains about the index property | 10:56 |
greyback | mzanetti: oops, that should keep a bool yes | 10:57 |
mzanetti | greyback: rest looks quite good I'd say | 10:57 |
larsu | dednick: it does | 10:58 |
dednick | larsu: does it call onCompleted? or say it cant create the UnityMenuAction? | 10:59 |
dednick | larsu: nevermind. just tried myself. doesnt create | 11:00 |
larsu | dednick: right, it doesn't even have the messagemenuitemfactory type now | 11:00 |
larsu | I guess it treats the non-existent property like a syntax error | 11:01 |
dednick | larsu: yeah | 11:01 |
dednick | larsu: when you added to the factory, did you create an empty one UnityMenuAction {}, or with properties? | 11:02 |
larsu | dednick: with properties, I copied the one from the replyAction | 11:03 |
dednick | larsu: sigh. i dunno wtf is going on... | 11:03 |
dednick | larsu: is it not complaining about the replyAction in TextMessage? line 100? | 11:05 |
larsu | dednick: yes it is. It complains about all occurrences of UnityMenuAction | 11:06 |
dednick | ah | 11:06 |
larsu | I'm building without the patch now, lets see if that still works | 11:06 |
larsu | dednick: anyway, I'm hungry. You said it does work for you now, right? | 11:16 |
dednick | larsu: yeah | 11:17 |
larsu | dednick: okay, going to lunch then :) | 11:18 |
larsu | dednick: I just pushed a fix for the fullActionName thing | 11:18 |
dednick | larsu: ta. | 11:22 |
=== hikiko is now known as hikiko|lunch | ||
dednick | larsu: right, so i approved in terms of a review, but i'm hesitent to do global until we get it working for you :( | 12:01 |
=== alan_g is now known as alan_g|lunch | ||
dednick | larsu: you on qt5.0 or 5.1? | 12:03 |
=== dednick is now known as dednick|lunch | ||
larsu | dednick|lunch: 5.0.2 | 12:14 |
mzanetti | seb128: hey, did the settings override for the launcher already land in the image? | 12:19 |
seb128 | mzanetti, http://people.canonical.com/~ogra/touch-image-stats/20130903.changes | 12:19 |
seb128 | mzanetti, it did! | 12:19 |
seb128 | mzanetti, thanks to ogra ;-) | 12:19 |
mzanetti | \o/ | 12:19 |
mzanetti | thanks you two | 12:19 |
greyback | mzanetti: https://code.launchpad.net/~gerboland/unity-api/adjust-application-api/+merge/183641 | 12:21 |
greyback | lemme know what you think | 12:21 |
mzanetti | greyback: I'd need a review when you have some time. albert already did a first round and I've fixed his comments: https://code.launchpad.net/~mzanetti/unity8/launcher-defaults-from-dconf/+merge/183420 | 12:21 |
greyback | mzanetti: ok, looking now | 12:21 |
mzanetti | yep, will check out yours | 12:21 |
=== hikiko|lunch is now known as hikiko | ||
larsu | mzanetti: seb128 advises to call QIcon::setThemeName("ubuntu-mobile") in unity8 until bug #1098578 is fixed | 12:25 |
ubot5 | bug 1098578 in Ubuntu UI Toolkit "On Ubuntu Desktop, the icon theme used by Qt is always 'gnome' (instead of the user set one)" [High,Confirmed] https://launchpad.net/bugs/1098578 | 12:25 |
larsu | mzanetti: who'd be the right guy to ask where to put that? | 12:25 |
seb128 | larsu, mzanetti: system-settings does "QIcon::setThemeName("ubuntu-mobile");" | 12:25 |
seb128 | that works fine on touch and desktop (when the theme is installed, we depends on it) | 12:25 |
mzanetti | larsu: I am until saviq is back... yeah. | 12:26 |
mzanetti | larsu: probably just dump it in the main.cpp where we set up all the other import paths | 12:26 |
larsu | mzanetti: sounds reasonable. I'll put up a MR right away | 12:27 |
mzanetti | larsu: just found this code in unity8: | 12:27 |
mzanetti | void resolveIconTheme() { | 12:27 |
mzanetti | const char *ubuntuIconTheme = getenv("UBUNTU_ICON_THEME"); | 12:27 |
mzanetti | if (ubuntuIconTheme != NULL) { | 12:27 |
mzanetti | QIcon::setThemeName(ubuntuIconTheme); | 12:27 |
mzanetti | } | 12:27 |
mzanetti | } | 12:27 |
larsu | ugh, that's ... ugly | 12:27 |
mzanetti | larsu: I guess I know why this is | 12:28 |
mzanetti | larsu: there was a time when we were transitioning between ubuntu-mobile-icons and ubuntu-icons-touch (or something like that) | 12:28 |
seb128 | that makes testing with a different theme easy though | 12:28 |
mzanetti | probably it was added to suite that | 12:29 |
seb128 | you might want to do that and have an else case setting the default theme | 12:29 |
mzanetti | +1 ^^ | 12:29 |
larsu | okay, fair enough | 12:29 |
greyback | mzanetti: you don't like the C++11 for iterator? "for" -> "Q_FOREACH" :) | 12:30 |
davmor2 | guys in saucy if I click on the power button the system shutsdown even if the system is in sleep mode I'm guessing at that being a bad thing, to reproduce let your machine lock hit the power button watch it shutdown with no prompt | 12:31 |
mzanetti | greyback: Q_FOREACH knows better how to deal with QHash and QMap | 12:31 |
mzanetti | greyback: other than that it's probably the same. | 12:31 |
greyback | mzanetti: really? huh | 12:31 |
mzanetti | greyback: yeah.. for example Q_FOREACH(const QVariant &value, m_myHash) directly iterates over the values | 12:32 |
mzanetti | greyback: whereas c++11 would need for(const QVariant &value, m_myHash.values()) which needs to create a copy of the values list first | 12:32 |
greyback | mzanetti: ah of course yes | 12:32 |
mzanetti | greyback: not really sure if that applies to all my usages... but then its only because I'm used to Q_FOREACH | 12:33 |
mzanetti | greyback: I do hate the overuse of "auto" tho | 12:33 |
mzanetti | which that merge gets rid of too | 12:34 |
greyback | mzanetti: I'm fond of it, if the object type is obvious | 12:34 |
larsu | I just found out why unity8 takes 10 seconds to load on my desktop and I started crying: `cat /usr/lib/x86_64-linux-gnu/hud/dbus-activation-hack.sh? | 12:34 |
mzanetti | greyback: auto variant = m_accounts->getUserProperty(m_user, "launcher-items"); | 12:35 |
mzanetti | greyback: what is it? | 12:35 |
greyback | mzanetti: yeah, not obvious there at all. But there are times, like in an iterator, where it doesn't cause any harm | 12:36 |
mzanetti | yeah... but once you have a file which _only_ uses auto you get upset and remove them all :D | 12:36 |
greyback | http://blog.qt.digia.com/blog/2013/09/02/new-scene-graph-renderer/ new renderer looks very interesting, I sense the fact there's much GPU retention will help shell and some app performance greatly | 12:41 |
larsu | mzanetti, seb128: https://code.launchpad.net/~larsu/unity8/fallback-icon-theme/+merge/183649 | 12:43 |
seb128 | larsu, why do you Build-Depends on it? is that used in tests? | 12:44 |
seb128 | larsu, or did you mean to Depends: rather? | 12:44 |
larsu | seb128: this is why I pinged you. I don't think before I push! | 12:45 |
seb128 | lol | 12:45 |
seb128 | larsu, you can also make it a Recommends: since that's not strictly required | 12:45 |
larsu | seb128: okay :) | 12:45 |
seb128 | larsu, danke | 12:46 |
larsu | seb128: done | 12:46 |
seb128 | larsu, done, set as approved (but comment only, I'm not an unity8 maintainer and I don't feel like approving their MRs) | 12:47 |
larsu | ya that makes sense | 12:48 |
larsu | merci! | 12:48 |
dandrader | dednick|lunch, FYI: taking over the review of https://code.launchpad.net/~nick-dedekind/unity8/indicator.multi-icon/+merge/181862 from Albert as he's on holidays | 13:00 |
=== alan_g|lunch is now known as alan_g | ||
greyback | apt-get update: "Fetched 17.8 MB in 4993d 10h 9min 6s (0 B/s)" lol | 13:23 |
mhall119 | hey guys, any chance of https://bugs.launchpad.net/unity8/+bug/1214428 getting fixed soonish? It's causing problems with click app installation | 13:30 |
ubot5 | Launchpad bug 1214428 in Unity 8 "Apps with capital letters in their .desktop file name won't launch" [Undecided,New] | 13:30 |
mzanetti | dednick|lunch: standup | 13:30 |
mzanetti | kgunn: fyi: standup time in case you wanna join | 13:32 |
greyback | sorry about the background noise | 13:36 |
=== dednick|lunch is now known as dednick | ||
dednick | dandrader: cool. thanks | 13:39 |
larsu | mzanetti: getenv is deprecated?! That's news to me... | 13:49 |
mzanetti | larsu: oh... my mistake | 13:50 |
mzanetti | larsu: it was only deprecated in VCS2005 | 13:50 |
mzanetti | which we probably don't really care about | 13:50 |
larsu | mzanetti: ah, right, just read that in the docs as well :) | 13:50 |
larsu | mzanetti: I don't care, if you prefer qgetenv I can fix it | 13:50 |
larsu | I had already moved unity-mobile-icons to Recommends | 13:51 |
mhall119 | mzanetti: any feedback from katie on the wording for the tour's final step/ | 13:52 |
mhall119 | ? | 13:52 |
mzanetti | mhall119: yes: | 13:53 |
mzanetti | mhall119: go for 3 lines, but no link | 13:53 |
mzanetti | mhall119: unity_tour_threelines.png but removing the "Skip intro" link completely | 13:54 |
mhall119 | ok, can I modify EdgeDemoOverlay.qml to add an alias to the link's visibility? | 13:54 |
kgunn | dednick: curious...did saviq talk to you about possibly picking up snap decisions while macslow is out ? | 13:58 |
dednick | kgunn: nope | 13:58 |
kgunn | dednick: mzanetti & i were just chattting about first time boot/sim card pin entry which would need snap decisions | 13:58 |
kgunn | dednick: awesome... :) | 13:59 |
dednick | kgunn: is that a "awesome, thanks for picking it up"? :) | 13:59 |
dednick | as in "now you know" | 14:00 |
kgunn | dednick: awesome as in i dropped the ball | 14:00 |
kgunn | dednick: so...now that you do know...mind taking a look this afternoon just to get a feel for it, if you think you might be able to effect it ? | 14:01 |
kgunn | guess you'd have to branch off macslows... | 14:01 |
dednick | kgunn: sure | 14:01 |
mzanetti | dednick: ping me when you want to get started. I can explain what the plan is | 14:02 |
dednick | this is in relation to the ext snap dialogs right? | 14:02 |
kgunn | dednick: i trust your judgement...if you think its best to just wait till monday for macslow to return, lemme know | 14:02 |
kgunn | dednick: thank you | 14:04 |
kgunn | dednick: and to be totally clear....yeah, ext snap dialogs | 14:05 |
kgunn | dednick_: also if you can make progress on ext snap dialog...just as a but of freedom/guidance for you it's better to have something landed than perfection | 14:23 |
kgunn | fwiw | 14:23 |
=== dandrader is now known as dandrader|afk | ||
nic-doffay | Cimi, got time for a quick review? | 14:43 |
nic-doffay | dednick_, fancy a one line review? | 14:51 |
=== dandrader|afk is now known as dandrader | ||
dednick_ | nic-doffay: can do | 14:52 |
sil2100 | pete-woods: hi! | 14:53 |
pete-woods | sil2100: hi | 14:53 |
nic-doffay | dednick_, https://code.launchpad.net/~nicolas-doffay/unity8/infographic-background-change/+merge/183675 | 14:53 |
nic-doffay | Should be easy to test too. | 14:53 |
sil2100 | pete-woods: did you get an e-mail from Robert about lp:ubuntu-voice? | 14:53 |
sil2100 | pete-woods: I mean, unity-voice | 14:53 |
sil2100 | pete-woods: since he outlined some problems with the symbols file and autopilot tests | 14:54 |
pete-woods | sil2100: I did a could of days ago, yes - I thought I had fixed everything he asked about | 14:54 |
dednick_ | that's 19 lines... | 14:55 |
dednick_ | larsu: you ever get that action working? | 14:56 |
larsu | dednick_: no, was distracted (and will be for the next hour as well, meeting) | 14:57 |
dednick_ | larsu: ok | 14:57 |
larsu | dednick_: maybe I can try to do it on the side if the meeting is boring :P | 14:57 |
dednick_ | :) | 14:58 |
sil2100 | pete-woods: oh, let me re-check then - thanks! | 14:58 |
pete-woods | :) | 14:59 |
dednick_ | nic-doffay: did design give you that asset? | 15:01 |
nic-doffay | dednick_, yeah | 15:07 |
dednick_ | nic-doffay: approved | 15:08 |
nic-doffay | dednick_, ta | 15:12 |
larsu | mzanetti: hm the merge failed because of some tests that seem unrelated. Can you please have a look what the problem there might be? | 15:20 |
mzanetti | larsu: yeah... I need to walk through the whole list of merges today anyways | 15:32 |
larsu | mzanetti: cool, thanks | 15:33 |
greyback | mzanetti: if you have minute: https://code.launchpad.net/~gerboland/unity-api/adjust-application-api/+merge/183641 | 15:40 |
=== om26er is now known as om26e | ||
=== om26e is now known as om26er | ||
greyback | mzanetti: thank you | 15:49 |
dednick | pete-woods: ping | 15:51 |
pete-woods | dednick: hi | 15:54 |
mhall119 | mzanetti: updated branch and screenshot on https://code.launchpad.net/~mhall119/unity8/finish-tour/+merge/183037 | 15:55 |
dednick | pete-woods: hey. were you working on pinlock with MacSlow? | 15:56 |
pete-woods | dednick: not pinlock - it's the wifi auto dialogue API I'm waiting on from MacSlow | 15:56 |
pete-woods | *auto -> auth | 15:57 |
dednick | pete-woods: i c | 15:57 |
dednick | pete-woods: thanks, maybe tedg knows | 15:57 |
dednick | tedg: ^ | 15:57 |
mzanetti | mhall119: it's approved already | 15:58 |
tedg | dednick, I think that'll be me when it gets done, but I don't know that MacSlow knows that yet. | 15:58 |
mzanetti | mhall119: err... I think I forgot to actually send the approval :D | 15:58 |
mhall119 | mzanetti: not according ot LP | 15:58 |
mzanetti | will do now | 15:58 |
mhall119 | :P | 15:58 |
pete-woods | dednick: np | 15:59 |
dednick | tedg: ok. cool | 15:59 |
tedg | greyback, Do I just need the Mir PPA to get your version of Unity on my phone, or is there something else I should do? | 16:01 |
greyback | tedg: just that ppa, update, dist-upgrade and reboot | 16:02 |
* tedg jumps | 16:02 | |
mhall119 | mzanetti: any idea why Jenkins is failing now, but wasn't previously, and on something that seems unrelated? | 16:05 |
mhall119 | https://code.launchpad.net/~mhall119/unity8/finish-tour/+merge/183037 | 16:05 |
mhall119 | sorry: https://jenkins.qa.ubuntu.com/job/unity-phablet-qmluitests-saucy/1461/? | 16:05 |
mzanetti | mhall119: yeah. the SDK released an update of the ToolBar and we weren't prepared for it | 16:06 |
mhall119 | ok, so it's not my branch that broke it | 16:08 |
larsu | dednick: this issue is like magic. Run unity8 it works (none of the weird errors). Apply your patch and run it, weird errors happen. Revert it, errors *still* happen. I want to cry. | 16:10 |
dednick | larsu: you havent got some weird import path variables set up have you? | 16:11 |
larsu | no.... | 16:11 |
dednick | and you're building every time you make a source change to unity? you're not editing build files again are you?! | 16:12 |
dednick | :) i have no idea | 16:12 |
larsu | nope, rebuilding every time | 16:12 |
larsu | I think its just my setup | 16:12 |
larsu | also: I need to go now | 16:12 |
dednick | larsu: i'll get someone else to test tomorrow. if it works, we will approve | 16:13 |
larsu | dednick: okay cool thanks. Have a good evening | 16:13 |
dednick | you too | 16:13 |
mzanetti | greyback: unity-api merged btw. | 16:15 |
greyback | mzanetti: yay! | 16:15 |
mzanetti | greyback: is there already some branch I could start working on top of? | 16:16 |
kgunn | greyback: you were saying you needed review on unity-mir before landing...? do you have a victim ? | 16:16 |
mzanetti | 29 branches to be reviewed... Saviq will kill us when he's back :D | 16:17 |
greyback | kgunn: Almost: I need reviewer for unity8 stuff. I got a preliminary review from Saviq, who had no major problems | 16:17 |
greyback | kgunn: and the bits I'm doing, and bits dandrader is doing. We'll review each-other's stuff there probably | 16:17 |
greyback | kgunn: so I think I'll steal dandrader for a while longer | 16:18 |
nic-doffay | mzanetti, you got a moment? | 16:22 |
mzanetti | nic-doffay: yeah | 16:22 |
nic-doffay | mzanetti, I need to pass the size of the whole qmlscene of unity to the PageHeader, any idea what the easiest way to do that is? | 16:23 |
nic-doffay | I can't just use narrowMode for this... | 16:23 |
nic-doffay | The size of the filters tab needs to be worked out based on the total height. | 16:23 |
mzanetti | nic-doffay: what for? | 16:24 |
mzanetti | nic-doffay: wait. need to check your code | 16:24 |
mzanetti | nic-doffay: you mean the height of the filterSelector Rectangle? | 16:25 |
nic-doffay | Yeah. | 16:25 |
nic-doffay | I think that needs to be based off the total height. | 16:26 |
nic-doffay | Or at least the amount of space for the MouseArea below to get rid of it. | 16:26 |
mzanetti | nic-doffay: for that use the InverseMouseArea | 16:27 |
nic-doffay | mzanetti, ^5 for that. | 16:32 |
mzanetti | nic-doffay: 5? | 16:32 |
nic-doffay | mzanetti, high five ;P | 16:32 |
mzanetti | ah :) | 16:32 |
=== dandrader is now known as dandrader|lunch | ||
nic-doffay | mzanetti, thanks for that screen | 16:36 |
nic-doffay | I noticed the filter button is already there? | 16:36 |
nic-doffay | in the search bar in addition to the one I added? | 16:36 |
mzanetti | nic-doffay: I was just playing around while doing the review | 16:36 |
mzanetti | so it's only hacked in in a local branch here | 16:37 |
nic-doffay | mzanetti, ah ok. | 16:37 |
nic-doffay | I was worried because I didn't see anything in the code anywhere. | 16:37 |
mzanetti | nic-doffay: so... have asked martin | 16:37 |
nic-doffay | mzanetti, are you setting your theme manually? | 16:37 |
nic-doffay | Still need to get to the bottom of why mine isn't displaying. | 16:37 |
mzanetti | the overlay should be relative to the screen height, keeping 8 grid units distance from the bottom | 16:37 |
nic-doffay | mzanetti, right so we'll need to pass the total height. | 16:37 |
mzanetti | nic-doffay: hmm... I have set up KDE to use the ubuntu-mobile-icons theme | 16:38 |
mzanetti | nic-doffay: so I probably get it as default. but larsu fixed it for you. once his branch has landed, you can merge with trunk and you'll se the icons | 16:38 |
nic-doffay | mzanetti, ah brilliant. | 16:39 |
nic-doffay | larsu, cheers | 16:39 |
nic-doffay | mzanetti, where would be the best place to pass the height into the PageHeader from? | 16:39 |
mzanetti | nic-doffay: for now you can do this to get it: | 16:39 |
mzanetti | export UBUNTU_ICON_THEME="ubuntu-mobile" | 16:39 |
mzanetti | before doing the ./run and the icons should show up | 16:40 |
mzanetti | nic-doffay: let me check the code for the height thingie | 16:40 |
nic-doffay | mzanetti, I'm assuming the individual selectors will be a fraction of the filter rectangle's height, but I'm double checking. | 16:40 |
mzanetti | nic-doffay: ok. have found the height thing | 16:54 |
mzanetti | nic-doffay: you have a moment so I can give you some instructions how to fix stuff? | 16:54 |
nic-doffay | mzanetti, yeah | 16:58 |
mzanetti | nic-doffay: ok. so... from what I can see the narrowMode change is on | 16:59 |
mzanetti | nic-doffay: ok. so... from what I can see the narrowMode change is only because you added the Icon, right? | 16:59 |
mzanetti | nic-doffay: so I'd suggest you revert line 37 + 38 of the diff to where it was before. | 17:00 |
mzanetti | nic-doffay: then remove your added icon | 17:00 |
nic-doffay | mzanetti, I've added my icon into the search bar instead. | 17:00 |
mzanetti | nic-doffay: and add the icon back inside the TextField as secondaryItem. | 17:01 |
mzanetti | yeah exactly | 17:01 |
nic-doffay | mzanetti, yeah done that bit. | 17:01 |
mzanetti | its still not 100% matching the design, but at least it looks good enough for now. | 17:01 |
nic-doffay | mzanetti, yeah agreed. | 17:01 |
mzanetti | and especially, it looks like the other search/filter labels | 17:01 |
nic-doffay | There still needs a black bar and some other additions. | 17:01 |
mzanetti | exactly. but we won't do that in here. for that we'll create a ModifiedTextField.qml and use that one here... in another merge request tho | 17:02 |
mzanetti | so. next thing is the height of that overlay. I asked martin. it should be screenHeight - units.gu(8) | 17:02 |
mzanetti | for that, add a property filtersHeight (or similar) to the PageHeader. | 17:03 |
mzanetti | and in the Dash*.qml, where the pageHeader is used, set it to the appropriate height | 17:03 |
mzanetti | nic-doffay: anything else you need more information? | 17:05 |
nic-doffay | mzanetti, where's that height stored originally thought? | 17:06 |
nic-doffay | I'd like to pass it into DashContent preferably and then on to the pageHeader. | 17:06 |
mzanetti | nic-doffay: for example: the PageHeader is used in Dash/DashContent.qml | 17:08 |
mzanetti | nic-doffay: I think that ScopeView is fullscreen. so just pass scopeView.height - units.gu(8) to the PageHeader in that file | 17:08 |
nic-doffay | mzanetti, perfect cheers | 17:18 |
dednick | dandrader|lunch: fixed those issues in indicator.multi_icon | 17:19 |
mhall119 | mzanetti: will I need to merge from trunk and push a new revision of my branch, or should jenkins just start being happy with mine once trunk is fixed? | 17:23 |
mzanetti | mhall119: I don't think you will need to merge. | 17:24 |
mhall119 | ok, thanks | 17:24 |
* mhall119 is unreasonably happy about contributing a fix to Unity :) | 17:24 | |
=== racarr is now known as racarr|lunch | ||
mterry | Is anyone else working on upgrading unity8 from icon://gicon/name to icon://theme/name? I can do that if not | 18:18 |
tvoss_ | mzanetti, ping | 19:06 |
mzanetti | tvoss_: hi | 19:06 |
=== racarr|lunch is now known as racarr | ||
=== dandrader|lunch is now known as dandrader | ||
=== dandrader is now known as dandrader|afk | ||
=== dandrader|afk is now known as dandrader | ||
=== dandrader is now known as dandrader|afk | ||
=== dandrader|afk is now known as dandrader | ||
=== sergiuse1s is now known as sergiusens |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!