[03:04] <Kb52> Ok found out after booting into windows that my USB Bluetooth Dongle is hardware defective. Have a TekRam TM-304 but cannot find whatever ubuntu wants to make it work, so no dice.
[03:08] <Kb52> The one that might work, because Ubuntu sees it ok, is hardware defective, it appears to work but does not. The Tekram which works fine in Windoze is not able to work in Linux. I guess nobody made a Linux driver for Tekram usb bluetooth dongle. So I am just dead in the water on figuring out how to do it. In lsusb it sees it as a Roper Class 1 Bluetooth Adapter.
[03:10] <Kb52> Hmm guess everyone here went away.
[06:26] <liuxg> does anyone get the Positioning working on ubuntu phone?
[08:22] <mivoligo> dpm: hi, not sure if you answered my question some time ago as I went offline. It was about getting translations from Launchpad and adding them to an app.
[08:23] <dpm> hi mivoligo, I think it was a bit of a broad question. Could you ellaborate?
[08:23] <dpm> in general this doc should help in setting up a project for translation auto-commits: https://wiki.ubuntu.com/Translations/LpProjectConfiguration
[08:24] <dpm> but if you tell me exactly which part is not clear, then we can look at it in detail
[08:24] <mivoligo> dpm: sure, I have a simple app which has been already translated to few languages. And now I want to get those translations and add them to the app
[08:24] <dpm> mivoligo, link to the lp project?
[08:25] <mivoligo> dpm: https://translations.launchpad.net/pop-that-wrap
[08:26] <dpm> mivoligo, ok. Then I'd recommend to follow all the steps from that wiki page. That will get your project set up for daily translation exports committed to a branch
[08:27] <dpm> generally the exports happen early in the morning European time, so we've missed the window for today. This means that if you set up the project for translations now, the first translations will be committed to your branch tomorrow
[08:27] <dpm> however, you can always manually download translations and manually commit them if you can't wait to the next export
[08:28] <dpm> there is a link for downloading them here: https://translations.launchpad.net/pop-that-wrap/trunk/+translations
[08:29] <mivoligo> dpm: ok, and how do I add them to the app? Should I just put them to "po" folder?
[08:34] <dpm> mivoligo, the po/ folder will initially contain all the .po files from LP (one per language). These are source files, which need to be built into binary .mo files, which is what you'll need to ship with your app
[08:36] <dpm> mivoligo, what kind of app is that? Is it QML? Does it use a build system (e.g. cmake)?
[08:36] <mivoligo> dpm: QML
[08:40] <mivoligo> dpm: I see I can manually choose to download MO formathttps://translations.launchpad.net/pop-that-wrap/trunk/+export
[08:44] <dpm> ah, yeah
[08:44] <dpm> mivoligo, if you don't have a build system and you're doing this manually, then that works
[08:45] <dpm> you'll then need to also manually craft the path where the .mo files will be in the final .click
[08:46] <mivoligo> dpm: so, should I just put .mo files into "po" folder and build a click package? Or is it more complicated?
[08:46] <dpm> no, the po/ folder is only for the source .po files, the .mo files go into a different path, just a sec
[08:46] <mivoligo> dpm: ah, ok
[08:49] <davidcalle> dpm, regaridng .mo files. I can build them from a default qml project from template, with make build-translations (which creates a local share/locale/...)
[08:49] <davidcalle> But I'm not able to start qmlviewer in a different locale. The suggested LANG=... qmlscene Main.qml doesn't seem to work
[08:51] <dpm> mivoligo, here's what the path should look like in the final click: http://pastebin.ubuntu.com/10831865/
[08:51] <dpm> davidcalle, yeah, that's what I mentioned in the document:
[08:52] <dpm> the translations are loaded from the default locations:
[08:52] <dpm> - either /usr/share/locale on the desktop
[08:52] <dpm> - or $CLICKFOLDER/share/locale on the phone from a click package
[08:53] <dpm> in the example, the .mo files are not installed under /usr/share/locale, so gettext cannot find them
[08:54] <dpm> loading translations for test purposes is always a bit more complicated than it should be
[08:54] <davidcalle> dpm, so in a nutshell, you need a device
[08:54] <dpm> either that, or you can use the bindtextdomain call to tell gettext an explicit location to load translations from
[08:54] <dpm> I'm not sure if we expose it in the QML API, though
[08:55]  * dpm looks at API docs
[08:55] <mivoligo> dpm: do I have to create that folder structure manually?
[08:55] <davidcalle> dpm, not in the doc
[08:55] <dpm> mivoligo, generally a build system takes care of that
[08:56] <dpm> if you don't have a build system, you'll need to do this manually, yes. Or generally, with a script that does it for you
[08:57] <mivoligo> dpm: by build system you mean "Build and validate click package" option in "Publish" menu in SDK?
[08:57] <dpm> mivoligo, no, I mean whether you are using cmake or qmake, or autotools... to build your app and click package
[08:58] <mivoligo> dpm: I use just the mentioned button :D
[08:58] <dpm> :)
[08:59] <mivoligo> cmake, qmake and their families are black magic to me
[09:00] <dpm> zbenjamin, quick question: for a QML app, if I've got a po/ folder containing translations (.po files), will the Publish option to create a click package detect that there are translations and will build them?
[09:01] <dpm> davidcalle, I think that's maybe a more important topic than launching in a different language: i.e. how do I package my translations ^
[09:02] <dpm> Ideally Publish should take care of it. I'm not sure exactly what's launched when clicking on "Build and validate click package"
[09:02] <dpm> bzoltan_, zbenjamin, do you know? ^^
[09:04] <mivoligo> dpm: let me experiment, I will put some .po files into po folder and run Build and validate thing
[09:05] <zbenjamin> dpm: hmm it should
[09:06] <dpm> mivoligo, it seems another option might be what davidcalle was mentioning: use the make build-translations option, which will create the folder structure for you, and then when you hit publish it will just package those already built .mo files. davidcalle, to be clear, is that the Build translations option from the menu (which didn't work for me)?
[09:06]  * dpm tries
[09:08] <zbenjamin> dpm: depends a bit on how old the project is. If its really old then it probably lacks the Makefile that does the right steps
[09:08] <dpm> zbenjamin, no, I'm talking about a project created a few days ago
[09:08] <dpm> testing it now
[09:08] <zbenjamin> that should do
[09:08] <dpm> I see the rule in the Makefile, yes
[09:10] <dpm> zbenjamin, what's the workflow nowadays for building clicks from a QML app? I see the "Build and validate" button under the Publish tab grayed out
[09:10] <zbenjamin> dpm: you need a ubuntu kit for that
[09:11] <dpm> ok
[09:11]  * dpm adds kit to the QML project
[09:16] <davidcalle> dpm, zbenjamin, if I try the menu option to build translations, it does nothing for me (po files not found)
[09:16] <dpm> same here
[09:16] <davidcalle> But, make build-translations in the project work
[09:16] <zbenjamin> dpm: davidcalle: hitting just "build" in qtc should do the trick
[09:17] <dpm> zbenjamin, it does not, as davidcalle says, you have to manually run make build-translations and only then Publish will pick up the built translations
[09:17] <dpm> it seems Publish does not run the build-translations target in the makefile
[09:17] <zbenjamin> dpm: what project?
[09:18] <davidcalle> dpm, any new one with the QML template
[09:18] <davidcalle> zbenjamin, ^
[09:18] <zbenjamin> davidcalle: dpm: if you are targeting the 15.04 framework use qmake based projects ...
[09:18] <dpm> zbenjamin, exactly. But for this particular one, we're looking at lp:howmanyapples
[09:18] <zbenjamin> davidcalle: dpm: soon we will deprecate the qmlproject type
[09:19] <zbenjamin> but lets see why its not working
[09:19] <dpm> zbenjamin, hm, how soon is soon? We're currently writing a tutorial for this
[09:19] <zbenjamin> dpm: 15.10 will most likely not show the qmlproject type in the "new project" wizard anymore
[09:20]  * davidcalle afk for a few min
[09:21] <zbenjamin> davidcalle: dpm: never add the .user file to your repository its valid just for your machine
[09:21] <dpm> zbenjamin, yeah, already mentioned that :)
[09:22] <davidcalle> :)
[09:22] <mivoligo> dpm: so how do I "manually run make build-translations"?
[09:23] <zbenjamin> dpm: davidcalle: ok that is a bug i'll fix that right now
[09:23] <dpm> mivoligo, tl;dr - in the meantime, you can drop to the command line, cd to your project's folder, ensure you've got .po files in the po/ folder, then run: "make build-translations" and then do the build as usual with the "Build and validate click package" button
[09:23] <zbenjamin> its seems to be just a one line fix
[09:23] <dpm> zbenjamin, that'd be awesome
[09:24] <mivoligo> dpm: thanks :)
[09:30] <dpm> zbenjamin, while you're at it: this rule in the Makefile would need a tweak as well: http://pastebin.ubuntu.com/10831970/
[09:30] <dpm> template.pot is too generic
[09:30] <dpm> it should be either the project name or the app id
[09:30] <dpm> probably the project name to make it more readable
[09:30] <dpm> davidcalle, ^
[09:31] <mivoligo> dpm: hmm.. I installed the package but the menu is still in English
[09:32] <dpm> mivoligo, does your .click contain the .mo files in the right locations?
[09:32] <mivoligo> dpm: I think so
[09:32] <dpm> mivoligo, I would need a bit more of info to be able to help :)
[09:33] <dpm> I'm not sure what you mean by "the menu"
[09:33] <dpm> or whether you're running this on a device, the emulator or the desktop
[09:33] <dpm> and whether the .click package does indeed contain the .mo files
[09:33] <mivoligo> dpm: link to mo file: /Pop-that-wrap/share/locale/pop-that-wrap-cs/LC_MESSAGES/mivoligo.pop-that-wrap.mo
[09:34] <mivoligo> dpm: I'm running on the device
[09:34] <dpm> mivoligo, ah, I see what it is
[09:34] <dpm> the part of the path that says 'pop-that-wrap-cs'
[09:34] <dpm> should only be 'cs'
[09:34] <dpm> I'm guessing
[09:34] <dpm> you downloaded translations from LP
[09:35] <mivoligo> dpm: yes
[09:35] <dpm> but LP has an old bug
[09:35] <mivoligo> :D
[09:35] <dpm> that means it adds $PROJECTNAME- to the language code in the name of the downloaded .po files
[09:35] <dpm> it does not affect the automatic exports, but it's still present when you manually download
[09:36]  * mivoligo is renaming
[09:36] <dpm> mivoligo, so what you'll need to do is to rename all .po files in the po folder to remove the pop-that-wrap- prefix
[09:40] <mivoligo> dpm: still no luck :(
[09:43] <zbenjamin> dpm: the finished .mo file is using the appid as name. The template.po is really just the template for new translations
[09:44] <dpm> zbenjamin, I know. What I'm saying is that template.pot is too generic. It should be projectname.pot
[09:44] <zbenjamin> err template.pot
[09:45] <zbenjamin> hmm using po/$(APP_ID).pot: $(QMLJS_FILES)  as target does not work for some reason
[09:45] <zbenjamin> make: *** No rule to make target '1,2', needed by 'po/zeller-benjamin.untitled4.pot'.  Schluss.
[09:46] <dpm> zbenjamin, davidcalle, to track the bugs that affect the i18n tutorial: https://bugs.launchpad.net/ubuntu/+source/qtcreator-plugin-ubuntu/+bugs?field.tag=i18n-qml-tutorial
[09:48] <dpm> zbenjamin, you can perhaps have a look at what Build > Ubuntu > Translations > Update Translations Template does
[09:53] <mivoligo> dpm: in the Makefile it says "#APP_ID needs to match the "name" field of the click manifest APP_ID=mivoligo.pop-that-wrap" but in the manifest.json I have name: "pop-that-wrap.mivoligo" Might that be the problem?
[09:56] <mivoligo> nik90: I bet you have some experience in QML project translation ;)
[09:57] <dpm> mivoligo, I think that might be it. Which one of the two is the correct app id?
[09:57] <mivoligo> dpm: how can I know that?
[09:58] <nik90> mivoligo: usually I test by setting the system language to dutch and testing apps out.
[09:59] <mivoligo> dpm: I guess pop-that-wrap.mivoligo is correct
[09:59] <dpm> hm, it seems https://wiki.ubuntu.com/AppStore/Interfaces/ApplicationId hasn't been updated to reflect the recent changes in app id
[09:59] <mivoligo> nik90: I have problem with adding translations to my app
[10:00] <mivoligo> dpm: so is this a bug in Makefile too?
[10:01] <nik90> mivoligo: just read the backlog
[10:03]  * davidcalle hugs dpm
[10:03] <davidcalle> dpm, I'll address the rest of your comments in the meantime
[10:03] <dpm> mivoligo, probably the comment in the Makefile is incorrect, and it shows the app id constructed the wrong way round?
[10:03] <dpm> thanks davidcalle :)
[10:04] <mivoligo> dpm: OW YEAH,  THAT WAS IT :D :D
[10:04] <dpm> nice :)
[10:04]  * mivoligo hugs dpm too
[10:05]  * dpm hugs mivoligo back :)
[10:05] <zbenjamin> dpm: seems this is the problem "--keyword=tr:1,2"   in the xgettext call. Do you know what it is used for?
[10:05] <dpm> mivoligo, would you mind filing a bug at https://bugs.launchpad.net/ubuntu/+source/qtcreator-plugin-ubuntu/+filebug nd tagging it with i18n-qml-tutorial ?
[10:05] <mivoligo> dpm: about Makefile?
[10:07] <dpm> zbenjamin, it's used to tell xgettext how to identify strings that are marked for translations and then put them in the final .pot file. In this case, it's been told to look for strings wrapped in tr() (from i18n.tr()). As per the arguments, I'll need to refresh my memory from the internet, just a sec :))
[10:07] <dpm> mivoligo, yeah, about the incorrect comment
[10:08] <mivoligo> dpm: the comment is actually correct, the APP_ID is somehow changed
[10:09] <dpm> mivoligo, oh, I see. So how did you actually fix your issue?
[10:09] <mivoligo> dpm: I've changed APP_ID in Makefile
[10:10] <dpm> zbenjamin, the xgettext --keyword invocation is correct. As far as I can tell, it's the same as we use for all core apps. The actual format is at https://www.gnu.org/software/gettext/manual/html_node/xgettext-Invocation.html (look for "--keyword[=keywordspec]")
[10:11] <mivoligo> dpm: so it's something wrong with how the Makefile is generated
[10:11] <zbenjamin> dpm: yes its correct, but the : is interpreted by the makefile as a target , hence the "target 1,2 not found"
[10:11] <zbenjamin> dpm: and it seems its not possible to escape : in a makefile
[10:11] <dpm> mivoligo, then I guess that's a bug to address in the makefile
[10:12] <dpm> zbenjamin, hm, xgettext has been around for ages, one would think that this is an issue that would have a solution
[10:12] <zbenjamin> maybe if i put them in a var
[10:13]  * dpm asks the internet
[10:15] <dpm> zbenjamin, folks seem to put this in an $(XGETTEXT_FLAGS) variable, yes: http://git.kaarsemaker.net/git/commit/adc3b2b27670f123bd7dd742ba0fa0e0a60e1b19/
[10:18] <zbenjamin> dpm: moving the stuff around revealed the real error. My editor converted the tab to spaces in the makefile.
[10:18] <zbenjamin> which is a nono :D
[10:19] <dpm> zbenjamin, aha, that's actually something I checked too, but the Makefile in the project I've got open does have a tab indeed
[10:28] <zbenjamin> dpm: https://code.launchpad.net/~zeller-benjamin/qtcreator-plugin-ubuntu/fixqmlclickbuild/+merge/256458
[10:30] <dpm> zbenjamin, jumping on a call in a minute, will look at it asap. davidcalle, perhaps you can do it in the meantime? ^
[10:33] <zbenjamin> dpm: btw forget about the menu item.. I wonder if we need it at all anyway. Hitting build should do both steps
[10:43] <davidcalle> zbenjamin, should I wait for a new commit (for the menu button) before reviewing the merge?
[10:45] <nerochiaro> oSoMoN: can you look at https://code.launchpad.net/~uriboni/webbrowser-app/bookmarks-in-suggestions/+merge/256459 when you are doing your next round of review ? i added lots of comments on the MR itself so hopefully we won't need a lot of back and forth
[10:46] <davidcalle> zbenjamin, one thing I noticed is that using the Translation Template menu item creates two templates : template.pot and appname.pot
[10:46] <oSoMoN> nerochiaro, will do just after lunch, thanks!
[10:47] <dpm> zbenjamin, davidcalle, about the menu option: there are two of them - 1) Update template and 2) Build translations. I'd be happy to drop 2), but 1) is actually useful, so that you can just update the template without building the whole thing
[10:47] <davidcalle> zbenjamin, disregard my last comment, I must have been doing something wrong :)
[10:47] <davidcalle> dpm +1
[10:48] <davidcalle> zbenjamin, ok, so, there is a template.pot appearing when I run the app from the SDK
[10:49] <dpm> davidcalle, I saw that too
[10:52] <dpm> davidcalle, zbenjamin, I'm guessing that will go away once the fix lands, as in that case when running the app the projectname.pot template will be created, but it will just overwrite the existing one
[10:53] <dpm> but nitpicking on it, I guess the template shouldn't be created in the first place when just running the app
[10:53] <davidcalle> dpm, I don't disagree, but why should the template be created at this stage?
[10:54] <dpm> davidcalle, that's what I'm saying, it shouldn't
[10:54] <davidcalle> Yeah, I've seen your reply when pressing return :)
[10:54] <dpm> :)
[10:55] <mivoligo> well, it was actually useful for me to upload the template to Launchpad
[10:56] <zbenjamin> dpm: thats not going to change though. Buildsteps are always executed at that stage. Also this whole qmlproject thing is just a hack since .qmlprojects where never meant to be used for what we do
[10:57] <dpm> mivoligo, yeah, but generally you'd do that either manually (via the menu option) or at build time, not when running the app
[10:57] <zbenjamin> dpm: the project format to be used is .pro
[10:57] <dpm> ?
[10:57] <dpm> zbenjamin, what do you mean?
[10:57] <zbenjamin> dpm: i was answering to [16.04 12:53:43] <dpm> but nitpicking on it, I guess the template shouldn't be created in the first place when just running the app
[10:58] <mivoligo> dpm: but in order to run the app you have to build it first, right?
[10:58] <dpm> yeah, I think that's the point
[10:58] <zbenjamin> dpm: i think even in cmake projects the translation template is built before the app is executed
[10:59] <zbenjamin> mivoligo: no, you can disable it in the buildsteps if you want. But then they also won't be packaged when creating your click package
[10:59] <dpm> zbenjamin, IIRC it's generally built manually by a rule that's not run as part of the build
[10:59] <dpm> at least that's how we do it for core apps
[11:00] <dpm> zbenjamin, davidcalle, but I think to keep things simple, I'd leave as it is
[11:00] <zbenjamin> +1
[11:00] <dpm> it'll just be a bit annoying for app devs, as the template file will be shown as changed in bzr every time they run the app
[11:00] <mivoligo> zbenjamin: OK, I'm just clicking buttons :D have no idea how it all works
[11:01] <zbenjamin> dpm: remove it from the default target
[11:01] <mcphail> If the .pot file changes, does it invalidate all the .po files?
[11:01] <zbenjamin> dpm: in the makefile
[11:01]  * mcphail has been stuggling with gettext for another project
[11:02] <zbenjamin> mcphail: no, you need to manually copy the pot to the po files
[11:03] <dpm> mcphail, it depends on how radical the changes are, I guess. If you are using Launchpad and automatic imports/exports, whenever you update the template, then Launchpad will take care of merging the .po files with the new template for you
[11:03] <mcphail> zbenjamin: and does the generation of a new .pot trigger a hook to email the translators?
[11:03] <dpm> mcphail, if you are not using Launchpad, when you update the template you should use msgmerge to manually do this merge
[11:03] <mcphail> dpm: ok
[11:03] <dpm> mcphail, a hook where?
[11:04] <mcphail> dpm: I was wondering whether hooks are set up in bzr to email translators if the .pot file changes. I think some projects do this with git hooks
[11:05] <mcphail> dpm: and that would be a problem if the pot file changes each run
[11:05] <dpm> mcphail, but it's always up to the maintainer of the project to set up the hooks
[11:05] <dpm> I personally don't appreciate automated e-mails that much :)
[11:05] <mcphail> :)
[11:06] <mcphail> dpm: just wondering. Bzr seems to have a bit of automated behaviour I don't understand yet
[11:06] <mcphail> So what triggers the call to translators when something needs updated?
[11:09] <popey> dbarth: if someone wants to file bugs against the webapps shipping on the phone, where do they do that?
[11:14] <dpm> mcphail, generally a human writing an e-mail and hitting the Send button :)
[11:15] <dpm> zbenjamin, bzoltan_, thanks a lot for addressing this superquickly, really appreciated! So that we have an idea of when we can publish the tutorial, do you know more or less when the fix will be released?
[11:17] <mcphail> dpm: aah. I thought it was automated in launchpad. Sorry for the digression :)
[11:17] <dpm> np :)
[11:24] <nerochiaro> oSoMoN: is that ok for you if I start taking care of https://trello.com/c/92qH2ZT7 or https://trello.com/c/oOlKUeeh ?
[11:29] <bzoltan_> dpm:  must be today :) Right?
[11:30] <dpm> bzoltan_, not necessarily, it's not a critical fix and we're still reviewing the tutorial, so we could wait. I was just asking to have an idea
[11:31] <dpm> but if you're offering today, I won't say no ;)
[11:31] <bzoltan_> dpm:  I mean, today is the _FREEZ_
[11:31] <bzoltan_> dpm:  so we do push it today
[11:31] <dpm> ah, of course
[11:31] <dpm> good timing :)
[11:31] <oSoMoN> nerochiaro, Arthur is already taking care of both tasks (should be landing soon),
[11:32] <nerochiaro> oSoMoN: great. any other stuff in the pipeline I can pick up then ? I have not any tasks claimed for this sprint apparently
[11:35] <oSoMoN> nerochiaro, nothing comes to mind right now, I’ll think about what you could tackle, in the meantime see with Bill if he has more stuff for you in other apps
[11:35] <nerochiaro> oSoMoN: ok
[12:00] <dbarth> popey: on the webapps-core project
[12:01] <dbarth> specifying the app name in brackets (since we bundle the main webapps in that single project for now)
[12:04] <popey> ok! thanks
[13:11] <kalikiana> zsombi_: https://code.launchpad.net/~zsombi/ubuntu-ui-toolkit/05-setpalette/+merge/253500
[13:13] <zsombi_> kalikiana: uhm... the two text component suse differet theme in the test
[13:14] <kalikiana> zsombi_: yes. and InnerText is the one that has custom theming
[13:14] <zsombi_> kalikiana: one uses the standard theme, the other one modifies th epalette
[13:14] <kalikiana> zsombi_: field from what I can see isn't applied correctly
[13:15] <zsombi_> lemme check
[13:15] <zsombi_> were you using a binding or a simple value?
[13:16] <kalikiana> zsombi_: it does seem odd since there's no specific focus handling in your changes… so it might be uncovering a lurking bug from elsewhere
[13:16] <kalikiana> zsombi_: just field: UbuntuColors.blue
[13:16] <kalikiana> nothing more
[13:16] <zsombi_> kalikiana: ok, a binding then
[13:17] <kalikiana> ah, sorry, I didn't realize what you were asking
[13:17] <kalikiana> lemme try plain colors
[13:17] <zsombi_> np :)
[13:17] <kalikiana> zsombi_: identical behavior
[13:18] <kalikiana> zsombi_: I'm checking TextAreaStyle to see if it might be doing something funny
[13:20] <zsombi_> kalikiana: weird... it gets stuck to the selected.field indeed...
[13:20] <zsombi_> kalikiana: that seems to be something with the style then...
[13:22] <kalikiana> zsombi_: styledItem.focus is apprently always true
[13:24] <kalikiana> hmmm hold on
[13:26] <zsombi_> kalikiana: it shouldn't be, as the other TextField behaves nicely
[13:27] <zsombi_> kalikiana: aaah, highlighted may mess up?
[13:27] <zsombi_> kalikiana: we should have removed support for that as gventuri asked us to get rid of it
[13:28] <kalikiana> zsombi_: I found it. "focus" should be "activeFocus"
[13:28] <kalikiana> we use it wrongly twice
[13:28] <kalikiana> changing two lines I made it work
[13:28] <kalikiana> zsombi_: so it's not a bug in your code
[13:28] <zsombi_> kalikiana: but... why the heck does work with the other text field then?
[13:30] <kalikiana> zsombi_: hmmm
[13:30]  * dholbach hugs davidcalle
[13:30] <dholbach> davidcalle, maybe you can join #snappy too? :)
[13:32] <davidcalle> dholbach, :)
[13:33] <mcphail> The app design guidelines suggest certain clours for certain components (e.g. UbuntuColor.red for negative actions). If the design guidelines change, all of these components will have to be updated in every app. Do you think there is a role for something like "UbuntuColors.negative"?
[13:34] <zsombi_> kalikiana: actually it doesn't work!
[13:34] <zsombi_> kalikiana: the first text field changes the color, but the second doesn't!
[13:34] <zsombi_> kalikiana: if I use focus only
[13:34] <DanChapman> mcphail: +1 to that
[13:35] <zsombi_> kalikiana: if I use activeFocus, all works fine
[13:38] <kalikiana> zsombi_: funny thing, if I have two custom themed TextField's they can switch colors correctly between each other
[13:39] <kalikiana> zsombi_: yes, found it. FocusScope at fault
[13:39] <zsombi_> kalikiana: weird... I pushed an update where text style uses activeFocus, and removed the styledItem.highlighted use
[13:40] <zsombi_> kalikiana: yes, I suspected that as well
[13:40] <kalikiana> zsombi_: I can place any TextField in a FocusScope to get it to show the issue
[13:40] <kalikiana> so this makes sense
[13:40] <zsombi_> yep
[13:41] <zsombi_> kalikiana: I'll add te missing test as well
[13:41] <kalikiana> zsombi_: dude, don't remove the highlighted
[13:42] <kalikiana> fix it :-)
[13:42] <zsombi_> kalikiana: why? we do not have use cases anymore, the highlighted was removed on request, gventuri did ask us +1 year ago
[13:42] <kalikiana> zsombi_: we still have that API
[13:43] <zsombi_> kalikiana: yes, and we will have it, but it will mean sthing else
[13:43] <zsombi_> kalikiana: otoh, the state it was representing doesn't exist anymore
[13:43] <zsombi_> kind of like deprecated behavior
[13:43] <kalikiana> zsombi_: which is what? right now I only see that you break it
[13:44] <kalikiana> and it's not deprecated
[13:45] <zsombi_> kalikiana: we don't have the 3rd state of teh text inputs, the property wasn't marked as deprecated as we thought that it can be used to show that there is a focus highlight frame around it
[13:45] <zsombi_> kalikiana: oren had this vision of having disabled, enabled non-focused, enabled-non focused highhlighted and enabled-focused states, which was wrong
[13:46] <zsombi_> kalikiana: ok, I'll put it back :)
[13:49] <kalikiana> zsombi_: at least I'd like to see a discussion on what we want it to mean if we change it… for TextArea it's typical to not dim when it loses focus, and changing it only for TextField won't be very obvious
[13:51] <zsombi_> kalikiana: well... I don't have the logs on that discussion, it was loooooooong ago... but you're right, a text input should stay with focus color when OSK is away. On desktop we don't have that state
[13:52] <zsombi_> but this change I made is also valid for TextArea, they have the same style
[13:52] <Diogo> hi guys
[13:52] <Diogo> can some one help me out plzz??
[13:52] <kalikiana> zsombi_: ah, true, it would break both. even worse :-D
[13:53] <zsombi_> yes :)
[13:53] <zsombi_> sorry
[13:54] <zsombi_> you've good eyes ;P
[13:55] <zsombi_> kalikiana: ok, so if I put the highlighted back, I don't get the normal.field coloring back :/
[13:56] <kalikiana> zsombi_: that's why I said above, the "focus" is wrong in more than one place. modules/Ubuntu/Components/TextField.qml and modules/Ubuntu/Components/TextArea.qml use it for highlighted
[13:57] <zsombi_> kalikiana: ah, you meant that.. ok
[13:57] <kalikiana> I should've been clearer
[13:58] <zsombi_> kalikiana: should we try to deal with it in a separate MR or in this one?
[13:59] <kalikiana> zsombi_: it's technically a separate issue… no strong feelings
[14:01] <zsombi_> kalikiana: ok, then I'll revert the style changes
[14:01] <zsombi_> kalikiana: as activeFocus doesn't help there
[14:02] <kalikiana> zsombi_: well, it does help… but with a separate branch we can also have a unit test without delaying your palette work
[14:03] <zsombi_> kalikiana: yep, pushed the changes
[14:04] <kalikiana> zsombi_: was about to file a bug… this might possibly even explain https://bugs.launchpad.net/ubuntu/+source/ubuntu-ui-toolkit/+bug/1378231
[14:07] <kalikiana> aaaand it does
[14:07] <kalikiana> yay
[14:07] <zsombi_> kalikiana: oh, yes, so teh bug suddenly becomes High importance wise
[14:10] <zsombi_> kalikiana: ok, so if no more requests on 05-setpalette, then push teh button :)
[14:10] <kalikiana> zsombi_: https://code.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/textComponentsActiveFocus/+merge/256495
[14:11] <zsombi_> kalikiana: aaaah, that makes sense :D
[14:13] <zsombi_> kalikiana: grrrr....
[14:14] <zsombi_> kalikiana: let's have https://code.launchpad.net/~zsombi/ubuntu-ui-toolkit/04-apply-styleset/+merge/252467 as prereq...
[14:14] <zsombi_> kalikiana: otherwise we will have to have a pile of iterations to fix the Theme to theme changes
[14:15] <kalikiana> zsombi_: hmm what do you mean?
[14:15] <zsombi_> kalikiana: using Theme will give deprecation warning, you shoudl use theme instead
[14:16] <zsombi_> kalikiana: 04 just got happroved by t1mp, so it'll land soon
[14:16] <zsombi_> kalikiana: but you can decide whether you wait till that gets merged and then resync with staging, or make it as prereq so I can happrove it and will get into staging once the prereq lands
[14:18] <kalikiana> zsombi_: good point. I'll wait, I trust it shouldn't take long
[14:18] <zsombi_> kalikiana: from 04 onwards, using Theme (with capitals) will give warning, so you should switch to use theme
[14:18] <zsombi_> kalikiana: max 1h
[14:24] <t1mp> zsombi_: that is if jenkins likes it
[14:24] <zsombi_> t1mp: it must like it!!!!!
[14:25] <zsombi_> t1mp: about the 07 to make it as blog post, I think it is too huge for a blog, but we can have it also as Page article
[14:28] <oSoMoN> nerochiaro, I pushed one comment to your MR
[14:29] <nerochiaro> oSoMoN: wow, was expecting more :)
[14:30] <oSoMoN> nerochiaro, don’t worry, I’m sure I’ll find things to tweak when doing an actual code review :)
[14:30] <nerochiaro> oSoMoN: oh, i thought that is what you were doing :D
[14:32] <nerochiaro> oSoMoN: hmm, that bug is strange. it works if i scroll with the scroll bar but not if i drag. i will look into it
[14:32] <oSoMoN> nerochiaro, I did a functional review first, then will do a code review (when everything works as expected)
[14:34] <kalikiana> zsombi_: make it a series of blog posts ;-)
[14:54] <nik90> DanChapman: Hey, just saw the backlog in ubuntu-touch-meeting. I did make a start on the first proposal, however I was not happy with the performance of the component on the phone. The issue is that since I am calculating x and y coordinates at every frame of the animation, it is leading to a laggy experience.
[14:55] <nik90> DanChapman: the issue is actually persistent on the current radial bottom edge, but it is within reasonable limits due to other variables not being changes such as opacity etc which is not the case with the design mockups provided to us.
[14:55] <nik90> DanChapman: I was thinking of opening it up to the community to come and help. May be they might other ideas on the implementation part.
[15:00] <t1mp> kalikiana: :)
[15:00] <t1mp> zsombi_, kalikiana: https://code.launchpad.net/~zsombi/ubuntu-ui-toolkit/04-apply-styleset/+merge/252467 merged :D
[15:01] <t1mp> kalikiana: sync your branches :)
[15:05] <kalikiana> zsombi_: here you go https://code.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/textComponentsActiveFocus/+merge/256495
[15:24] <nerochiaro> oSoMoN: fixes pushed on the MR (the one you asked plus another bug I found)
[15:27] <oSoMoN> nerochiaro, thanks, I’m doing a proper code review now, will post my comments in a moment
[15:28] <kalikiana> t1mp: is PageHeadConfiguration involved in anything fancy, like a loader? it's one of the things that seem to be affected by a race condition I'm tracking down (in the api check tool)
[15:36] <faenil> how are the modularScale values computed?
[15:36] <faenil> https://bazaar.launchpad.net/~ubuntu-branches/ubuntu/vivid/ubuntu-ui-toolkit/vivid/view/head:/modules/Ubuntu/Components/plugin/ucfontutils.cpp
[15:36] <faenil> 0.677 for example
[15:45] <faenil> it should be quite font-specific afaik
[15:46] <t1mp> kalikiana: in PageHeadStyle the overflowPanels depend on the Actions that are defined in PageHeadConfiguration
[15:46] <t1mp> kalikiana: and the overflow panels are only created when you click on the overflow button
[15:49] <oSoMoN> nerochiaro, more comments sent your way
[15:52] <nerochiaro> oSoMoN: fixing them
[16:06] <DS-McGuire> Would anyone know why my Ubuntu SDK doesn't run QML files and I have to use the terminal to do so?
[16:08] <faenil> DS-McGuire: describe "doesn't run"
[16:08] <DS-McGuire> faenil, Just don't do anything
[16:08] <DanChapman> nik90, ok thanks for the update, and for giving it a go. I agree I think it would be good to open it up to the community and see what comes from it
[16:08] <faenil> you play the green arrow and there's no output in "Compile Output" ?
[16:08] <DS-McGuire> I have to use: qmlscene ~/Ubuntu-Help-QML/main.qml
[16:09] <faenil> also, are you talking about running on desktop, or on device?
[16:09] <faenil> if on device, which chroot target?
[16:10] <DS-McGuire> desktop: here is an output from Compile Output: Error while building/deploying project Linux_Command_Cookbook (kit: Desktop)
[16:10] <DS-McGuire> When executing step 'Update translations template'
[16:10] <faenil> if you scroll up there should be something more detailed
[16:10] <faenil> feel free to copy the whole content of the output on some pastebin
[16:10] <faenil> and I'll have a look
[16:11] <DS-McGuire> faenil, Thanks, I will do so now
[16:12] <DS-McGuire> faenil, http://paste.ubuntu.com/10833823/ Here you go
[16:13] <faenil> DS-McGuire: do you have the "po" folder in your project?
[16:13] <faenil> if not, did you delete it?
[16:14] <DS-McGuire> faenil, It is a brand new project with nothing changed/deleted. Even if I start a new project it doens't work.
[16:14] <faenil> which template are you trying?
[16:16] <DS-McGuire> QML App with Simple UI (qmlproject)
[16:18] <DS-McGuire> faenil, Forgot to ping^
[16:26] <faenil> DS-McGuire: mmm maybe I know what's wrong
[16:26] <faenil> can you try getting rid of the spaces in "Ubuntu Touch Projects" ?
[16:26] <DS-McGuire> Really?
[16:26] <DS-McGuire> I will try now
[16:28] <DS-McGuire> faenil, You're amazing! hahah!
[16:28] <faenil> DS-McGuire: \o/
[16:28] <DS-McGuire> That has been holding me back for months!!!
[16:29] <DS-McGuire> Shall I report it as a bug?
[16:29] <faenil> probably, yes please ;)
[16:29] <faenil> DS-McGuire: sorry I'm in the community since a few weeks only :p
[16:29] <DS-McGuire> faenil, Awesome! Man, I am so freaking happy haha!
[16:29] <faenil> glad to hear that, get back to coding now! haha :D
[16:29] <DS-McGuire> faenil, No need to apoligize at a;;
[16:29] <DS-McGuire> all*
[16:29] <DS-McGuire> apologize*
[16:30]  * DS-McGuire goes back to coding
[16:50] <oSoMoN> nerochiaro, did you forget to push a revision by any chance? I’m not seeing the change for making SuggestionsFilterModel::filterAcceptsRow more tolerant
[16:50] <nerochiaro> oSoMoN: yes, i am on actually finishing adding tests for it right now, then will push it. sorry
[16:51] <oSoMoN> nerochiaro, ok, no worries! let me know when I can review again
[16:51] <oSoMoN> nerochiaro, btw, the change to use a var for the list of models is totally worth it
[16:52] <nerochiaro> oSoMoN: i find the assignment more clunky but it is worth using reduce()
[16:52] <oSoMoN> yes, and afaic the assignment is not that clunky, it looks very QMLish
[16:58] <nerochiaro> oSoMoN: would be better if you could create the two models inside the array constructor
[16:58] <nerochiaro> oSoMoN: but then QML will see it as a QML list and complain about the property not being a list property
[16:59] <oSoMoN> nerochiaro, sure, but it’s no big deal, it doesn’t impair code readability
[17:03] <nerochiaro> oSoMoN: yeah, worth it. pushed last change
[17:06] <oSoMoN> nerochiaro, thanks
[17:39] <oSoMoN> nerochiaro, one more (minor) comment on the MR
[17:39] <nerochiaro> oSoMoN: looking
[17:40] <nerochiaro> oSoMoN: fixing
[17:45] <nerochiaro> oSoMoN: fix pushed
[17:47] <oSoMoN> nerochiaro, thanks. At this point I know I’m picking nits, but it would be a tiny teeny bit more performant if you stored the expected count, rather than the expected set and then calling .count() on it at each iteration
[17:47] <nerochiaro> oSoMoN: that's fine. gimme a sec
[17:49] <nerochiaro> oSoMoN: done
[17:49] <oSoMoN> cheers
[18:11] <oSoMoN> nerochiaro, I added one suggestion to the MR
[18:12] <nerochiaro> oSoMoN: ok. btw, why is jenkins failing now ? I ran all tests locally and it was ok
[18:13] <oSoMoN> nerochiaro, I think that’s an intermittent failure where the unity8 greeter fails to unlock, you can safely ignore it
[18:15] <nerochiaro> oSoMoN: pushed fix. went for ordered in the name to make it clear, since usually these lists aren't
[18:16] <oSoMoN> nerochiaro, ok, fine by me
[18:17] <oSoMoN> nerochiaro, one last thing, and I think we’re good to go: test coverage (see my comment)
[18:18] <oSoMoN> nerochiaro, and before a new CI run kicks in (and fails), there’s one flake8 error in the autopilot tests
[18:18] <nerochiaro> damn, keep forgetting that
[18:19] <nerochiaro> oSoMoN: and i will check code coverage now
[18:19] <oSoMoN> nerochiaro, "make test" before every commit is your friend :)
[18:45] <nerochiaro> oSoMoN:  coverage should be 100% now on the model I added
[18:46] <nerochiaro> oSoMoN: and flake8 fixed
[18:47] <nerochiaro> oSoMoN: is there a way to clean coverage info without having to hunt files manually ? when i added coverage to qt-halide i had to add a target that did that, and made make coverage a dependency of it, otherwise when you change something the html reports will not update as far as I know
[18:55] <oSoMoN> nerochiaro, not that I know of, I usually just clean everything (very inefficient, I know)
[18:56] <nerochiaro> oSoMoN: personally i would rather have it regenerated from scratch every time you run make coverage. i submit a small patch tomorrow
[18:57] <oSoMoN> cool, thanks