=== chriadam_ is now known as chriadam [01:27] aquarius: Any pending changes from you on the icon branch? [01:27] If not, I'll address those points I had and push it [01:28] (I have icon downloader code waiting to go it) [01:28] rschroll, no, get stuck in. I have some half-finished code which actually fetches all the icons and does the "decide which icon" stuff inside manifestParser.js, but it can certainly wait [01:29] I haven't committed it because I honestly can't decide whether I should decide on an icon with full knowledge and then give you a URL, or give you a list of URLs in order and then you fetch the first one. [01:30] sounds good [01:30] I have a C++ plugin to do the icon fetching, waiting in the wings [01:30] Right now, it gets a list of urls and goes through them until it finds a good one. [01:31] yeah. The issue is this: if I fetch all the icons to decide, then we have to fetch the "best" icon twice (once in my code, once from C++). Which seems wasteful. But if I *don't* fetch them all, then I don't have enough information to make a decision about which to use. [01:32] We can worry about that later, I think [01:33] certainly [01:33] the current code (that is, all the code which is currently pushed) picks an icon [01:33] we currently fail if the site points at an icon which doesn't currently exist, but that's OK for now [01:34] I altered that a bit in my branch. Now it returns a sorted list. We can go through the list until we find one that downloads. [01:34] ah, ok [01:34] At least I hope we can. I need a broken site to test it on [01:35] try, I think, brucelawson.com [01:35] Do you want to review the branch before I merge it into master, or should I hit the big red button as soon as I'm ready? [01:36] (I need to make the URL textbox properly handle "www.example.com" and treat it as "http://www.example.com", etc) [01:36] nah, if you're happy with it, run with it [01:37] brucelawon: invalid hostname. It's broken, but not in the right way.... [01:39] bah. I can't remember which URL I used -- it indicated an icon on a different domain, which didn't exist [01:41] might be worth setting up half a dozen test websites somewhere [01:41] I can just insert some bad urls in the javascript code and check that the compiled code handles them properly... [01:41] ha! that's probably easier. [01:41] have a test suite would be better still, but that's lots of work [01:41] don't tell the QA people. [01:43] my lips are sealed [06:51] good morning === faenil_ is now known as faenil === justCarakas1 is now known as justCarakas === faenil is now known as faenil_ === faenil_ is now known as faenil [09:01] 30 hours to go for UbuContest! [09:02] And we need much more nominations, so go crazy! :) [09:05] http://ubucon.de/2015/contest/nominate-individual === chriadam is now known as chriadam|away [10:12] t1mp: why would you need to get the z-value always max? [10:13] daker: ping [10:14] zsombi: to put the header on top of other stuff [10:15] zsombi: but in the end I just went with the default z-value [10:15] t1mp: you mean overlay the Page content, or whatever it is the content? [10:15] zsombi: yes [10:15] t1mp: why don't you leave that to the user? [10:15] zsombi: but I can more easiliy do that later in the page with header [10:15] zsombi: right, that's what I'm doing now in the Header [10:15] zsombi: see https://code.launchpad.net/~tpeeters/ubuntu-ui-toolkit/cppHeader/+merge/271347 [10:15] t1mp: the Page can then set the z-order if wants, then any other case would need to be custom, right? [10:15] zsombi: it is ready [10:16] zsombi: right [10:16] t1mp: ahh, finally :D [10:16] t1mp: especially that if you use it in a ListView.header, then you don't have to care about that [10:16] cool [10:17] zsombi: this is only the 'empty' header [10:17] quite a lot of code for such a small API :s [10:17] but it works :) [10:17] t1mp: why do you rename ListItems.Header? [10:17] zsombi: see the Description of the change [10:17] zsombi: I couldn't get qdoc to work properly if there is a Header.qml *anywhere* :s [10:18] zsombi: all the \qmlproperty Header::flickable and such go to the ListItem.Header [10:18] t1mp: fuuuuck!!! wdoc... [10:18] qdoc [10:18] t1mp: but you still export it as Header in qmldir right? [10:19] zsombi: yes [10:19] ok [10:19] zsombi: but the ListItems.Header docs are gone :( [10:19] t1mp: that's because of \deprecated [10:19] qdoc appears to care about filenames first [10:19] zsombi: right. Without that I would still have the problem. [10:20] zsombi: so it seems different "namespaces" for our module is a bad idea for qdoc [10:20] t1mp: crap [10:21] zsombi: I chose to do it like this because the ListItems module will be deprecated anyway [10:21] t1mp: right [10:21] zsombi: there is just one more thing I started to have doubts about, [10:22] zsombi: how to deal with styling [10:22] t1mp: .... [10:22] zsombi: I could make the header a StyledItem and give it a background (+divider) [10:22] t1mp: told ya... [10:22] but that can also come in a subcomponent for the the contents of the Header [10:22] hmm [10:22] t1mp: ... it can... but why a subcomponent? [10:22] zsombi: will making it a StyledItem mess up the implicitWidth? [10:23] t1mp: why not to derive from Header? [10:23] zsombi: with subcomponent I meant to derive from Header [10:23] t1mp: StyledItem takes the implicitWidth if set in the style only [10:23] zsombi: ah, good :) [10:24] t1mp: if you set one in teh component, that will override the style one [10:24] t1mp: check loadStyleItem() or postStyleLoaded() [10:24] zsombi: I was actually thinking of possible next-gen theming, and whether I should deal with the styling by, for example, adding a background property to the Header [10:25] t1mp: even if we go with next gen theming, the base type will be something we can control focus and others, so you shoudl deal with it [10:26] right. So I'll make it a StyledItem [10:31] zsombi: good, that didn't break any of the unit tests :) [10:31] t1mp: push it [10:32] zsombi: I will add a style for the background, and move the implicitHeight to the style as well [10:32] t1mp: ping me when you are ready... [10:32] zsombi: okay [10:35] zsombi: pong [10:35] daker: hei, I think me and faenil needs your help :) [10:36] daker: we have an issue finding out the correct layout for RTL ListItem progression [10:36] zsombi sure [10:36] zsombi: show me something [10:36] daker: liek where would you expect to have the progression, on the lright side or on teh left, pointing to the left direction? [10:37] daker: uhh, that'd be hard... do you have the UITK gallery in hand? [10:37] zsombi: screenshot that part [10:38] daker: in the UITK gallery, in the menu you will see items with chevron [10:38] zsombi: i don't have the latest version, using 14.04 [10:38] daker: https://www.dropbox.com/s/yw52ay91pqd2gg8/Screenshot%202015-09-17%2013.38.42.png?dl=0 [10:39] daker: see the chevrons on the first column [10:39] daker: how woudl you expect to see this layout in RTL? [10:40] zsombi: ok so in theory, the first column should be at right [10:40] daker: ok, how about the header layout? and the list items? [10:42] daker: do you expect to have the header actions in teh same place as in LTR? [10:42] daker: also the chevron... [10:42] zsombi: wait a sec [10:42] daker: however if you said the left column shoudl be on the right, then I think intuitive woudl be to have teh chevrons pointing to the left, isn't it? [10:42] daker: ah, ok, sorry :) [10:49] zsombi: i would expect something like this https://imgur.com/OW6iV2D [10:49] zsombi: I need to create a UCHeaderPrivate in order to set the styleName from cpp? [10:49] faenil: ^ see daker's screenshot? [10:50] t1mp: why? no, you shouldn't [10:51] t1mp: if you don't want to use Q_D() then you can simply go for UCStyledItemBasePrivate::get(this)->setStyleName() [10:51] t1mp: also, for you, daker's image^^ [10:51] zsombi: yep [10:51] t1mp: for teh header layout [10:51] popey: what is happening with reminders? [10:51] daker: thanks, buddy, a lot!!! [10:52] daker: thanks, may I ping you in the future for more difficult examples? :) [10:52] zsombi: ah, okay :) [10:52] zsombi: yw :) [10:52] faenil: sure anytime [10:52] great, cheers [10:52] t1mp: also, if you want to use Q_D() then you need to use Q_DECLARE_PRIVATE_D() macro [10:53] t1mp: and use QScopedPointer d_ptr [10:53] zsombi: do I need to set themeVersion? [10:53] t1mp: uhm... yeah.. you should [10:53] faenil: as i said in tablet mode this column at the right, should be at the left [10:54] t1mp: perhaps we should have a proper solution for that as well... right now this is embedded in the ListItem... [10:54] faenil: in other word everthing should be read from Right to Left [10:54] daker: sorry I'm confused now :) you siad the column in tablet should be on the right, is that correct? [10:54] said* [10:55] t1mp: I was thinking to have a solution for that as well, so we don't necessarily have to derive 13, 14, etc classes [10:55] faenil: what did confuse you? [10:55] faenil: the layout I had had two columns [10:55] "as i said in tablet mode this column at the right, should be at the left " [10:55] faenil: the one on teh left should be on teh right and th other on teh right shoudl be on teh left. [10:55] yes, he said the first column should have been on the right, now he said on the left :) [10:56] zsombi: yeah, that's what I thought [10:56] davmor2: kalikiana was looking at the sdk issue, but do you have logs for the apparmor issue? [10:56] ehm.. whatever :) [10:56] just trun the tablet around [10:56] ogra_: deal [10:56] faenil: ah sorry, you are correct [10:56] popey: there was a pastebin that I threw up let me dig it out [10:57] daker: ok I see what you mean now, I misinterpreted your last message, neverming ;) [10:57] nevemind* [10:57] davmor2: ta [10:57] but we got it :) [10:57] faenil: if you have a layout with colomuns, the order need to be reversed in RTL [10:57] faenil: 1|2 in LTR => 2|1 in RTL [10:58] zsombi: alternatively I could set the style in PageHeader.qml that will inherit from the cpp Header [10:58] daker: sure, sure :) [10:58] t1mp: do you want to make the PageHeader in cpp or QML? [10:59] zsombi: qml seems the best solution. It will be a Header with two ActionBars and some title. [10:59] or a Header with a SlotsLayout, depending on how flexible that will be ;) [10:59] t1mp: right, then do it there [11:00] popey: line 15 on I think http://paste.ubuntu.com/12419623/ [11:00] zsombi: in that case, https://code.launchpad.net/~tpeeters/ubuntu-ui-toolkit/cppHeader/+merge/271347 is ready! [11:02] zsombi: our generated docs don't show inheritance [11:02] davmor2: I'm getting issues on rc-proposed with reminders too now. [11:02] t1mp: that's bad, qdoc crap [11:02] davmor2: "add new account" doesn't work [11:03] t1mp: I wonder how Qt does... what tags do they use [11:03] zsombi: right. They have inheritance in the docs [11:03] faenil: https://2.bp.blogspot.com/-txMrX6kMxtY/UR6AW28Jx9I/AAAAAAAAB9E/DhVAtPhkPSI/s1600/settings-rtl.png [11:03] faenil: https://4.bp.blogspot.com/-EzUf6UQhhus/UR6ATD23lZI/AAAAAAAAB88/_APE31dL4zw/s1600/setings-ltr.png [11:03] popey: that's the one, but if you create the account in system settings, and then add that app it's called com.reminders_reminders rather than notes right? [11:03] zsombi: the docs for Header don't show that it is a StyledItem. But I guess that's a separate issue. [11:04] daker: yup :) [11:04] davmor2: yes, only user facing parts were renamed [11:04] t1mp: use \inherits, dear [11:04] popey: but in ota6 it's called notes [11:04] "it"? [11:04] t1mp: see QQuickFlickable docs [11:04] right [11:05] faenil: if an icon is pointing to a direction, like the chevrons they should point to the correct direction [11:05] daker: yeah, that's what I expected [11:05] davmor2: the user facing parts were called notes, but not the plugin. this is expected [11:06] zsombi: we should do that also for AbstractButton to show that it is an ActionItem [11:07] t1mp: sure [11:07] popey: http://people.canonical.com/~davmor2/notes.png [11:07] popey: that is ota6 [11:07] right, what's wrong? [11:08] [M#mZRequestAccess failed: QDBusError("com.ubuntu.OnlineAccountsUi.InvalidApplication", "Invalid client application") [11:08] this is an issue [11:08] popey: in the installed version it is called com.reminders_reminders not notes so I wondered if that was triggering some of the apparmor issues [11:08] no [11:08] the user facing bits are called notes [11:08] underneath it's reminders [11:08] that should be fine, been like that for ages [11:09] ^^ the QDBusError is more of a concern! [11:09] popey: right okay [11:09] why is it suddenly unable to launch online accounts [11:09] popey: no idea [11:10] yeah, i'll file a bug [11:10] ah, you did https://bugs.launchpad.net/reminders-app/+bug/1496084 [11:10] Ubuntu bug 1496084 in Ubuntu Notes app "0.5.490: Can't connect to reminders account from inside the app" [Undecided,Incomplete] [11:11] popey: oh can you change it to confirmed [11:11] done [11:12] mzanetti: can you do a wipe of you device at all? It seems that myself and popey who have done a wipe can see both issue but you can't, not sure if it is your main phone though or something :) [11:13] i haven't wiped [11:13] just using my retail device which has had no debs or any nonsense [11:19] popey: I thought your rc-proposed device was wiped [11:24] no === JMulholland_ is now known as JMulholland [15:28] popey, I'd be ready in the hangout already (a bit too early, I know but I've just finished a task) [15:28] okay [15:58] mzanetti: [M#Š8 'ubuntu-sdk-15.10-dev1' is not a supported framework [15:58] we _are_ using that framework, surely that's the issue then? [15:59] (that was from building mardy's branch) [15:59] ah, trunk uses 15.04, so the mardy branch will certainly fail. [16:02] popey, it should be there by now [16:03] mzanetti: not on ota-6 [16:03] right... [16:03] popey, probably this branch was the reason why mardy started the conversation in the first place [16:03] yeah [16:03] * popey is updating chroots for another build of r490 [16:03] * popey gets coffee [16:27] mzanetti: it works! (cc davmor2 ) [16:27] popey, which one? [16:28] ah, with the updated build chroots [16:28] the one built with an updated chroot! [16:28] still weird [16:28] yes [16:29] * davmor2 just blames the store and JamesTait in particular it's easier [16:30] * popey flips it back to "Ready for QA" [16:31] popey: does that fix both issues then? Or is the other one an sdk issue? [16:32] the black box is sdl [16:32] *sdk [16:32] popey: no the reminder setter [16:32] popey: black box was weather app ;) [16:33] oh, yes. [16:33] sorry, confused [16:33] or ride on time [16:33] yes, it fixes both [16:33] on my ota-6 retail krillin [16:33] \o/ [16:33] ya [16:33] * popey updates build scripts to update chroots before builds [16:35] popey: man not even a batted eyelid at throwing ride on time in there, I'm disappointed popey ;) [16:35] I am immune to that song [16:36] it washes over me [16:38] davmor2: to be clear, date thing still busted on rc-proposed, which kalikiana was looking at, but works fine on ota-6 [16:38] wonder if this is a vivid overlay issue. [16:38] popey: of all the things that you said.....that you said....are running through my head.......running through my head....../me now pictures popey tapping his foot and cussing me for finding tatu [16:38] davmor2: lemme do more testing, I'll undo my citrain switch [16:39] hahaha [16:39] :) [16:40] popey: what's your opinion of titanium? It's a song that grows on you :) [16:45] meh [16:52] mzanetti: 163 packages to install in each chroot, I suspect that will fix it ;) === JamesTai1 is now known as JamesTait [17:29] mzanetti: built with fully updated chroot including the ppa, and the date problem is still there, so I am sure it's an SDK issue, because it only happens on rc-proposed, not retail bq [17:34] popey, yes. I talked to zsombor about it. seems they are aware of it [17:34] ok [17:34] sweet [17:34] some binding loop decided to go mad [17:35] so davmor2 so, what do you want me to do? the date picker is busted, but it seems it's not us? [17:40] Qt Creator question: Where are the build and run settings stored? They've become corrupted for my project [17:40] I've deleted my .user file and the whole build directory, but the settings persist [17:49] I've removed the relevant kit from that project and re-added it. Still broken [17:50] rschroll: wild guess but are they stored under ~/config/QtProject/* [17:50] *.config [17:51] That was my guess to, but grepping for the project name in that directory gave nothing useful [17:52] I guess I could nuke the whole directory and see if that helps [17:52] zbenjamin: ^^ [17:55] rschroll hmmm yeah. Well i see there are some kit related directories .config/QtProject/qtcreator/ubuntu-sdk/ubuntu-sdk-15.04-armhf/ no idea whether they are related or what they are for :-) [17:56] DanChapman: No luck with nuking that directory. Settings must be elsewhere [17:57] meh :-( [17:57] I dont' have a ubuntu-sdk subfolder. Hrm.... [17:57] I'm on 14.04, if that makes a difference [17:58] Well, I do, but it's under ~/.config itself. [17:59] Nothing project related that I can see, though [17:59] popey: let me have a play with it and I'll get back to you [18:00] thanks davmor2 [18:01] davmor2: [M#uR53cad72310673c52cefedb551bde3055 com.ubuntu.reminders_0.5.490_multi.click [18:01] thats the one that should work :) [18:01] popey: thanks [18:10] Hi, anybody having issue with ubuntu-emulator (segmentation fault) ? [18:18] popey: is the link the same I assume so just a different check sum right? [18:18] yes [18:18] cool ta [18:18] re-download and check [18:18] np === ahayzen_ is now known as ahayzen [20:02] mzanetti, popey: yes, indeed the 15.10-dev1 framework is the reason why that branch is not working at all :-/ [20:03] :( mardy [20:04] popey: well, my understanding from that e-mail thread is that no one really objects to the creation of a new framework, so things will eventually be solved [20:49] rschroll: per project settings are only stored in the .user file [20:50] rschroll: the other settings are stored in ~/.config/QtProject/qtcreator [21:24] !ping [21:24] pong! [21:45] zbenjamin, DanChapman: The problem was the Qt Creator couldn't parse my manifest file because of CMake variables I had in there. (Why isn't it looking at the manifest file after configuration?) [21:46] Thanks for the helpful attempts, anyway!