=== chihchun_afk is now known as chihchun === faenil is now known as faenil_ === faenil_ is now known as faenil === athairus is now known as afkthairus === _salem is now known as salem_ === salem_ is now known as _salem === _salem is now known as salem_ [11:47] t1mp: Ready for review https://code.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/listsAndNameSpaces/+merge/298314 [11:48] (with, as I mentioned in Mumble, the caveat that it's not fully tested because I don't want to re-do everything here, which'll be done in the new api project anyway) [13:00] kalikiana_: cool, I'll check it [13:00] kalikiana_: it only checks the stuff we export to QML, right? So C++ API changes are still undetected? [13:01] I'm thinking what will happen for example when(if) we rename UCAction to UbuntuToolkit::Action [13:18] ahayzen: pong, sorry I lost power and with it my irssi backlog, but I see that you pinged me about a wallpaper contest [13:18] I think that's a fantastic idea [13:24] t1mp: I don't get that. Those are 3 things you mix up here. 1) UCAction, C++ class name 2) UbuntuToolkit, C++ namespace 3) Action, QML typename [13:25] What are you referring to exactly? [14:47] mhall119, \o/ woo [14:48] mhall119, yeah basically we were thinking of a wallpaper contest to get a set of images included into the default ubuntu image, so you don't just have that single image. ... I wondered if they could be 'convergent' images that work when cropped to portrait and when output to a landscape display [14:57] ahayzen: if you want to organize it, including defining rules like that, I will check on (A) prizes and (B) whether there is space on the default Ubuntu images for extra images [15:00] awesome :-) like it'd be nice even if it just the /ubuntu channel images doesn't have to be the OEM ones [15:00] ahoneybun, ^^ [15:02] ahayzen: let me check on space, IIRC we were pretty tight on how much more could go into the image given the partition sizes on the phones [15:02] ah yeah [15:02] would/could that not go into the user's partition though? [15:02] like ~/Pictures [15:03] i guess they are somewhere in a system folder though at the moment [15:03] I'll help with rules, and judging [15:04] awesome :-) [15:04] mhall119, how do we organise this sortof thing? start a wiki page somewhere detailing the rules? [15:04] (assuming there is enough space on the image etc) [15:04] ahayzen: usually we use developer.ubuntu.com for phone related contests [15:05] ah yeah [15:05] but i guess we could use similar rules to how the desktop one has been run [15:06] yeah, using flickr to collect the entries [15:06] :-) [15:06] yep flickr is great that at [15:29] ahoneybun: ahoneybun: you may also want to talk to nhaines for ideas/rules etc., since he runs the desktop wallpaper contest [15:35] kalikiana_: do I mix it up? [15:35] kalikiana_: the components.api used to have UCAction, and now it has Action. So in the MR it goes from C++ to QML name? [15:39] svij: good point ahayzen [15:39] yeah good idea [15:40] t1mp: "Used to" after the namespace MR broke it you mean? This is exactly what fixed. [15:40] kalikiana_: my question was: Does it only check the QML API? [15:40] because there used to be C++ class-names in there, and now only QML. [15:41] hmm.. that is strange though. UbuntuToolkit is the C++ namespace and Action is the QML component name [15:42] t1mp: Let's take an example. "Ubuntu.Components.Action 1.3 1.0 0.1 UbuntuToolkit::UCAction: QtObject" The C++ class name is at the end of the list of exported QML types and versions. We do that because the C++ class leaks into Autopilot introspection - we don't actually track C++ API. [15:46] kalikiana_: the MR has in many places stuff like: + property UbuntuToolkit.Action action [15:47] so it has CppNameSpace.QmlComponentName [15:47] Yes [15:47] if we track QML API, why do we have the CPP namespace? [15:47] it should be Ubuntu.Components.Action [15:47] UbuntuToolkit does not exist in QML [15:48] I think eventually we should track cpp api too. So we would have both UbuntuToolkit::UCAction and Ubuntu.Components.Action [15:48] but maybe in separate files [15:57] t1mp: Pushed a fix https://code.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/listsAndNameSpaces/+merge/298314 [15:59] kalikiana_: nice, thanks. [15:59] kalikiana_: do we not need the namespace on the right-hand side of the : for QML API? [15:59] kalikiana_: for example now we have Ubuntu.Components.Pickers.Dialer 1.3: StyledItem in the MR [15:59] should that be Ubuntu.Components.Pickers.Dialer 1.3: Ubuntu.Components.StyledItem [15:59] or is that useless? [16:02] t1mp: That'd be an enhancement... in theory it could be useful if the parent is from another module, so I might say, iff it came from another module the parent should be shown [16:02] (Rule of thumb is usually, include what's necessary, avoid redundancy) [16:02] Ubuntu.Layouts.Layouts 1.0 0.1 ULLayouts: Item [16:03] kalikiana_: Item comes from the QtQuick module [16:03] Right. So in theory it's ambiguous and should probably say QtQuick.Item [16:09] Ubuntu.Components.ListItems.Empty 1.3: AbstractButton [16:09] kalikiana_: ^ also that one. AbstractButton is in Ubuntu.Components, not in Ubuntu.Components.ListItems [16:09] Yes [16:09] perhaps it is the easiest to list the full import [16:09] or namespace [16:10] how do we call it in QML? [16:10] Nothing is easy here. :-) [16:10] Type system is everything but [16:10] t1mp: QML has no namespaces [16:10] ok, maybe not the easiest. But always listing the full namespace is the least ambiguous [16:11] ok, maybe not the easiest. But always listing the full import string is the least ambiguous [16:11] like that then? [16:11] how do you call the prefix for the Component name? [16:12] that's the module name. you import either all classes by their name or using "as" with a custom prefix (which looks like a namespace in C++ does) [16:13] but essentially the types are dumped freeform into the typesystem [16:15] for example "import Ubuntu.Components" is the module, if there's a conflict with another module, your problem [16:15] ah, ok module name then :) [16:15] I'd never call it a namespace because it really doesn't work like one [16:17] t1mp: Mind filing a bug for including the name of the module of parent classes? [16:17] sure [16:18] Thanks [16:19] huh, Ubuntu.Layouts.Layouts 1.0 0.1 ULLayouts: Item is correct? [16:19] I didn't notice ULLayouts before [16:19] yeah it is ULLayouts. [16:20] mhall119, i wonder if there isn't enough space, if we could create a click or zip file that the user could download a 'pack' of photos which is then imported into the settings app ... as a fallback plan [16:20] kalikiana_: perhaps we should show the namespace for the C++ parents? [16:21] kalikiana_: https://bugs.launchpad.net/ubuntu/+source/ubuntu-ui-toolkit/+bug/1596601 [16:21] Launchpad bug 1596601 in ubuntu-ui-toolkit (Ubuntu) "Show the module of the parent component in components.api" [Undecided,New] [16:23] kalikiana_: another one, [16:23] - readonly property UbuntuToolkit.QQuickMimeData data [16:23] 246 + readonly property MimeData data [16:24] kalikiana_: MimeData is not in a module/ [16:24] ? [16:24] also the newData() function after that [16:26] t1mp: MimeData comes from Ubuntu.Components - although it's perfectly possible it's not there because it was hard to do [16:27] (I probably wouldn't have seen a need to include if it's the same module, but I don't think it was a very specific decision to do or avoid) [16:39] kalikiana_: there are more issues, I commented on https://code.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/listsAndNameSpaces/+merge/298314 [16:41] t1mp: The syntax hasn't changed at all. If it *looks* confusing that's not a bug or regression in the context of this branch. [16:42] t1mp: Re "" list parsing. That's exactly what I'm fixing. [16:43] t1mp: the description of AlarmModel.get used to be this before the namespaces came into the picture, thus no regression "function UCAlarm* get(int index)" [16:45] You found a bug, though: "signal dragDirectionChanged(UCBottomEdge direction)" is not fixing the regression. [16:46] why is the module name missing here? -> it was never there ;-) [16:46] "function QQuickMimeData* newData()" is also not a regression [17:07] t1mp: So in summary, you found 2 genuine bugs (regressions caused by namespaces they weren't fixed), I recorded the reasoning in a comment and will investigate. [17:56] kalikiana_: okay, thanks. === kalikiana_ is now known as kalikiana === afkthairus is now known as athairus === salem_ is now known as _salem