=== chriadam|away is now known as chriadam [08:00] good morning === chriadam is now known as chriadam|away [09:19] Good morning all; happy Single Tasking Day! :-D === teknico_ is now known as teknico [09:37] dpm: o/ [09:37] hey mzanetti, morning! [09:38] dpm: can you give this a shot when you have some time? lp:~mzanetti/reminders-app/tablet [09:38] dpm: It's not complete, but I'd like a quick discussion about it [09:38] once you've seen it [09:38] mzanetti, sure, I don't have a tablet, so I'll be testing it on the desktop. popey has borrowed a N10 and can probably test it on a real device [09:38] desktop is fine [09:39] I didn't test it on a real tablet yet either [09:39] ok, give me 5 mins and I'll give it a go [09:39] * dpm is looking forward to seeing tablet mode [09:40] don't get too excited yet [09:41] zsombi: hey good morning [09:41] zsombi: I got a small question about Ubuntu Layouts [09:41] nik90: good afternoon :) [09:41] nik90: gpo ahead [09:41] zsombi:timezones yes [09:41] ;) [09:42] zsombi: can I define a Flickable inside a conditional layout instead of a row or a column? [09:42] nik90: everything should be possible [09:42] nik90: we don't have samples for that, but Fliockable is yet another Item, right? [09:42] Flickable I mean [09:42] zsombi: I am getting an issue where the page header doesn't detect that the convergent layout is using the flickable [09:42] zsombi: as a result the page header doesn't hide when scrolling up [09:43] zsombi: that's why I thought perhaps I was doing something wrong [09:43] nik90: yep, the page header is an issue... it does not detect runtime changes [09:43] same issue here :) [09:43] zsombi: but yes Flickable is an item [09:43] zsombi: so If I set the flickable property of the page manually, then hopefully that should fix the issue? [09:43] nik90: it's a problem of page header unfortunately [09:44] nik90: yep, that shoudl do the job [09:44] mzanetti: nice to know :) .. if one of us has a solution to it, then let the other person know. [09:44] zsombi: thnx.. [09:44] I just started yesterday evening with the tablet mode for reminders [09:44] the content itself works quite well [09:44] but header and toolbar get messed up badly when unsing 2 pagestacks next to each other [09:45] mzanetti: nik90: you guys should have started the whole app with that in mind ;) :P [09:45] mzanetti: I started in the weekend but had to stop because of a critical crash when switching layouts.. [09:45] mzanetti, ack [09:45] mik90: I'm investigating the crash there [09:45] mzanetti: Florian figured out that using a Dialer widget caused it [09:45] zsombi: well... its just about not pushing everything to the same pagstack [09:45] mzanetti: ;) [09:45] zsombi: thnx [09:46] but yeah... need to figure how to deal with panel and header [09:46] nik90: it's a setup, where the item laid out contains a loader. Dialer is one of those, but Picker is such as well... [09:46] seems the last push()/pop() to any pagestack updates them [09:47] dpm: so yeah... that's what I wanted to discuss. How do other apps deal with this? [09:47] mzanetti: yes, because push/pop updates the flickable property so header can adjust its logic accordingly [09:48] zsombi: yeah... makes sense if you have only one pagestack. makes me wonder if I'm actually supposed/allowed to have 2 of them next to each other [09:48] mzanetti: not really [09:48] mzanetty that messes up the header for good [09:48] mzanetti: ^ [09:48] yeah... so what do you suggest? [09:49] mzanetti: you'd need a two'column pagestack kinda setup? [09:49] mzanetti, so far we've only got weather and clock that we are also looking to get tablet modes for MWC for, so this is also relatively new to everyone, but as zsombi is here as the master of layouts, he can give us some advice :) [09:49] zsombi: yeah [09:49] zsombi: let me show you a few screenies.... gimme a minute [09:50] mzanetti: hehe :) I have a branch which turns PageStack into a two-column layout automatically using Layouts, but as ckpringle went away, that stuff got forgotten onto /dev/null [09:50] standard view phone: http://i.imgur.com/thyV6T4.png [09:50] after clicking on a note on the phone: http://i.imgur.com/1FTesPj.png [09:51] standard view tablet: http://i.imgur.com/TsHrtMF.png [09:51] clicking on a note on the tablet: http://i.imgur.com/nUlHE5j.png [09:51] now... as you can see there's an edit button for the note, which would push the EditPage on top of the note view page [09:51] mzanetti: yes, this is what I had for my PageStack [09:52] but at the same time, the tabs in the header allow you to switch to notebooks [09:52] which in turn would push pages to the left hand side [09:52] so yeah. not entirely sure how to deal with it right now [09:52] zsombi: are there any chances your 2 column pagestack will make it to trunk at some point? [09:53] mzanetti: it depends on how UX will pick it up... they're pretty busy with the whole UI reorg atm [09:53] heh... true. why am I doing this before the UI reorg actually? [09:53] mzanetti: but hey, the standard view phone is a Tabs, right? [09:53] zsombi: yeah, it is [09:54] mzanetti: same as the standard tablet, right? [09:54] mzanetti: and Notes is a PageStack [09:54] zsombi: yeah... I just make the pagestack non-anchors-fill just like the uitk gallery [09:54] Notes tab I mean [09:54] and put a second pagestack next to the other [09:55] and do some conditional pushes. the rest is identical between phone and tablet [09:55] mzanetti: ok, is the PageStack depth 2? [09:55] zsombi: if you go to Notebooks, you'll get a list of notebooks, you enter one and it pushes a page with the notes in there. [09:55] yeah, that's when the depth == 2 [09:56] mzaneti: brilliant! [09:56] or also, if you press the search button in the toolbar it pushes a searchPage to the left pagestack [09:56] mzanetti: then, in Phone layout you use PageStack, on tablet layout you use two columns, no need for pagestack [09:57] its not /that/ easy I'm afraid [09:57] mzanetti: the search will end up in the header, so forget about pushing a page [09:57] mzanetti: why? [09:57] cause both stacks need panels for example [09:57] mzanetti: what panels? [09:57] or well, toolactions, wherever they might be [09:57] left one needs: Add note for example [09:57] mzanetti: the tools will land on the header [09:58] right one needs: edit note, delete note, etc [09:58] mzanetti: you can update those tools based on what you have active [09:59] mzanetti: and yes, the problem in here is that the whole header-toolbar-search is changing, and everything will be upside down... [09:59] zsombi: ah, so I'd need to drop the toolbars from all the pages and only have one toolbar in the MainView, updating contents as I need it? [10:01] mzanetti: ..... damn it... it would be really good to have that two-column PageStack mode solved by now... [10:01] oh yeah, it would :D [10:01] why are you chatting around here? get to work :P [10:01] mzanetti: :P I'm trying to help you out... [10:02] I know. just kidding [10:02] mzanetti: here's the MR: https://code.launchpad.net/~zsombi/ubuntu-ui-toolkit/two-pane-pagestack/+merge/172288 [10:03] zsombi: oh man... this is *exactly* what I'm doing... if its on the left side, pushing, if its on the right side replacing [10:03] mzanetti: so this setup was supporting depth >2 too [10:03] mzanetti: ;) [10:04] mzanetti: this was doing the way that if push() was used on the right pane, the right pane was moved to the left pane, and the right one was updated with the new pushed page [10:05] mzanetti: so it all depended whether the pushOrReplace() or push() was used [10:05] ah... ok... then its a bit different... more like a column-view in a file manager [10:05] mzanetti: it all depends how you use it [10:06] mzanetti: "When pages are pushed from the current page (right pane) the current page will be moved into the left pane and the new one will be placed into the right one. This results in a push operation. When a page is pushed from a left pane page, the current page (from the right pane) will be replaced with the page that is just pushed." [10:06] mzanettI: even I don't remember exactly the functionality :D [10:07] mzanetti: so PushOrReplace() did all 4 U [10:07] zsombi: I have a small issue with labels not being positioned correctly despite using anchors.horizontalcenter to place them properly. As result I get, phone->http://imgur.com/03FmgaI , tablet->http://imgur.com/diTyYFa [10:07] right... although not exactly how I'd need it according to this description [10:08] I also have the use case that I have depth 1 on the left side and depth 2 on the right side. and then need to push to the left side without changing the right one [10:08] zsombi: Although If I replace those label ItemLayouts with Rectangles, I see that they are placed correctly [10:08] zsombi: https://code.launchpad.net/~nik90/ubuntu-clock-app/converge-clock-tab [10:08] mzanetti: if the active page was on the left side, the pushOrReplace() did show up the page on the right pane [10:09] mzanetti: then every consecutive pushOrReplace on the left page was doing a replace :) [10:09] mzanetti: the right-to-left (and vice versa) comes into force when depth > 2 [10:09] nik90: the anchors.verticalCenter will anchor the Label, not the text within [10:10] ok. would need to try it... not sure if its really exactly what I'd need... [10:10] but anyways, still leaves the issue with the header and the toolbar I guess [10:10] nik90: you should use Label.verticalAlignment [10:11] mzanetti: this stuff was working with those pretty well [10:11] zsombi: will give that a shot nnow [10:11] mzanetti: and I think we should get this branch up alive again [10:11] ideally I'd have 2 headers [10:11] mzanetti: why would you? [10:11] one for the tabs, and one for the content title for the right side [10:12] mzanetti: ah, yeah, that was also discussed, but that is pretty ugly... and misleading [10:12] zsombi: that did the trick...I didn't have to do this before [10:12] nik90: beware of this with Labets ;) [10:12] Labels [10:14] zsombi: :) [10:20] dpm: what do you think is better http://imgur.com/AYpB1Ir or http://imgur.com/diTyYFa === zsombi1 is now known as zsombi [10:32] nik90, both look good, but I personally find http://imgur.com/diTyYFa better. I find the shadow on the other one a bit too deep [10:32] popey, can you have a go at lp:~mzanetti/reminders-app/tablet on the N10? [10:33] dpm: yeah I reverted it to the one that you liked before you said it...It needs more work to make it look better.. [10:38] nik90, cool. Did you work around that crash? I've not been following the conversation... [10:40] dpm: no the crash is still there, however I believe Kaleo and zsombi are on it. [10:40] dpm: I am meanwhile trying to implement the phone and tablet views for all tabs. [10:40] awesome [10:40] dpm: when the crash is resolved and pushed to phone, then I will propose a MR [10:42] kalikiana: I saw a boat load of fixes land in the u1db trunk...nice that you got someone to review the MP finally.. [10:42] kalikiana: is there a planned landing slot for them? [10:43] nik90: no idea yet, still waiting to find out how the fancy new process works, and then to see how to apply it to u1db-qt [10:43] I'll know more later today [10:44] kalikiana: nice..I already got most of the fixes from the PPA on my 13.10 :) [10:46] dpm: ok [10:53] (12.40.04) nik90: dpm: no the crash is still there, however I believe Kaleo and zsombi are on it. [10:54] nik90: dpm: yes, we are on it [10:54] oSoMoN: do you have a minute to talk about gallery-app and thumbnails ? I'm trying to figure out the best way to approach the problem I'm having and I can use some advice [10:56] oSoMoN: essentially when you edit the picture the gallery doesn't save the changes on the image on disk, but keeps them in an internal database and applied them to the image on the fly. So if I use the SDK thumbnailer, it will use the images on disk which are different from what the gallery is actually displaying [10:57] oSoMoN: and the thumbnails won't match, which is not a huge deal when browsing the pictures, but it's a problem when scrolling through them in the photo viewer [10:57] nerochiaro, and the SDK thumbnailer can’t be fed with live data, it understands only files on disk? [10:58] oSoMoN: yes, you pass it a path and it generates the thumbnail for it (or returns it if it's already generated and fresh) [10:58] oSoMoN: it saves them on disk based on the md5 of the original image path [11:01] dpm: rotation is broken on nexus 10. [11:02] oSoMoN: i wasn't expecting it to do something like this, i thought it would save changes to the image immediately (losslessly when possible). i think this whole undo/redo stack is problematic as it is implemented now also for integrating the gallery with content hub, as it will have to "flatten" the edit stack to disk when exporting images via the hub anyway [11:02] nerochiaro, how about saving modifications to the original images to a separate file on disk, and use that one to generate the thumbnail [11:02] ? [11:03] oSoMoN: that might work, but it's one extra disk write of the full image that we're doing just to generate a thumbnail [11:04] nerochiaro, well, not just to generate the thumbnail, the image can then be used to be displayed in the app, instead of having to apply all the modifications on the fly [11:06] dpm: mzanetti http://imgur.com/t5xxEpB,528Nscu,2GjsSPb,fseUpjH ⍨ [11:06] popey: oh [11:08] * popey updates tablet [11:14] popey: hmm... works here :/ [11:14] fine here now I updated [11:14] oSoMoN: so basically you think that the undo/redo stack should work by keeping around: (1) a copy of the original image (2) the edited version of the file (3) the stack of operations that were applied on it. Then at each operation added to the stack, the file is modified. And when undoing from the stack if the last operation can be rolled back losslessly over the edited image (e.g. rotation for jpegs) then it's don [11:14] e that way, otherwise the original is taken, and all remaining operations are applied to it again in order minus the one being undone [11:15] oSoMoN: does it make sense ? [11:16] nerochiaro, doesn’t sound like a good idea, worded like this :) [11:17] oSoMoN: well, it's the best that can be done if there are operations which are lossy, like crop or enhance [11:17] nerochiaro, I guess that really depends on what operations we want to optimize [11:17] or rather, what use-cases we want to optimize [11:17] oSoMoN: well at the moment we only have 3: rotation, crop and enhance [11:18] oSoMoN: but one would expect to be a lot more [11:19] nerochiaro, what I meant is: I think it makes sense to ensure that viewing images in their current state is the fastest operation [11:19] argh [11:19] scratch "operation" [11:19] fastest use case [11:19] and it’s probably acceptable to have some overhead when applying new operations, or undoing them [11:20] nerochiaro, what are the other options we have, re- generating thumbnails? [11:20] oSoMoN: i was specifically asked to use the SDK thumbnailer for gallery, that's why I ended up in this situation. before we were using an internal thumbnailer for gallery [11:21] oSoMoN: i think what we can do if we want the easy way out is to use the SDK thumbnailer for browsing and keep the internal for the photo viewer [11:21] oSoMoN: the thumbs for browsing will still not be 100% accurate though [11:21] oSoMoN: so if that's important too we can't really switch [11:21] nerochiaro, and is modifying the SDK thumbnailer to accept live data an option? [11:22] oSoMoN: not that I know of [11:22] oSoMoN: we can add that ability I guess, but it will be completely custom for us, not part of the freedesktop standard in any way [11:24] doesn’t sound like the best idea [11:24] dpm: mzanetti http://imgur.com/tpUFHdc,w3KLulY,ATWS6NG,VZstxmj,GH7B1cM,1TYThPp,UQfIhG6,nrlcgJ1,1yJ2BGx better [11:24] nice [11:25] oSoMoN: other editors seem to keep undo/redo history in memory only while you're editing a picture and they ask you to save when you browse away from it [11:25] popey: so yeah. the question what to do with the toolbar still persists [11:26] oSoMoN: if we were doing that too, it would solve the issue I think [11:26] oSoMoN: as we will regen the thumbs only when we finish with editing [11:27] oSoMoN: but it requires (a) a decision to actually do it and (b) design input on how to ask for save or discard [11:28] nerochiaro, I’d tend to think that if we’re not asking the user to save/discard, it’s because it was designed this way [11:28] nik90: ok, I think I found the problem causing Layouts crash [11:28] oSoMoN: i wouldn't be surprised if a persistent undo/redo stack is something we didn't ask for and yorba decided to put n [11:28] in [11:29] nerochiaro, I don’t know, but in any case you should check with design first [11:30] oSoMoN: right. so while I do that and put this work on hold, is there any other priority that i should take care of, or should i go back and hack on gallery and camera bugs from the high priority list ? [11:49] nik90: ping [11:54] mzanetti, https://code.launchpad.net/~mzanetti/reminders-app/fix-attaching-images/+merge/206816/comments/485207 [11:55] rpadovani: nice, thanks [12:03] ybon: pong [12:03] zsombi: oh nice :) [12:03] zsombi: is it a simple fix? [12:03] nik90: 3 lines of code :) [12:04] zsombi: lol [12:04] nik90: as autopilot tests are working for you, can you just give me the version of the libs you have: http://paste.ubuntu.com/6954036/ [12:04] just in case I've a bad one [12:04] zsombi: in the SDK or the clock app? [12:04] I'm still unable to run them, and that's annoying :( [12:04] nik90: in Layouts [12:04] nik90: SDK [12:08] omg, my irc client didnt send my last few messages..lets try agian [12:08] zsombi: ok...convergence is something I need for the MWC. [12:08] zsombi: can we expect a fix to land on the phone (not just the ppa) by friday? [12:08] ybon: http://paste.ubuntu.com/6954086/ [12:08] nik90: I got that, I'm already preparing the MR [12:09] nik90: more than sure [12:09] zsombi: thnx! you rock! [12:09] nik90: thanks :) [12:09] so my autopilot-desktop seems out of date [12:10] I've two versions in my cache, let's take the good one :) [12:10] ybon: :) [12:11] ah: [12:11] The following packages have unmet dependencies: [12:11] autopilot-desktop : Depends: gir1.2-glib-2.0 (>= 1.39.0-0ubuntu1) but 1.38.0-0ubuntu1 is to be installed [12:11] ybon: which version of ubuntu are you running? [12:11] 13.10? [12:11] sorry I may have asked this before [12:11] yep [12:13] nik90: do you where you get the gir1 lib from? [12:13] ybon: btw, the entire QA is on a sprint...they are generally online around 19:00 UTC [12:13] ybon: let me check [12:14] ybon: I have 1.38.0-0ubuntu1 installed [12:14] oh, like mine so [12:14] humm [12:14] ybon: what are the autopilot ppa you have? [12:14] do you have more than one? [12:15] humm [12:16] does the 0 sized ones can be of any conflict here: http://paste.ubuntu.com/6954120/ ? [12:17] no idea [12:17] when you removed the other 2 ppas, do you purge them? [12:18] humm, I've just --remove I think [12:18] let me remove those [12:18] I think you should do a ppa-purge on the ppa to remove all other dependencies it upgraded completely [12:18] right [12:25] can reach to get ppa-purge to get anything (could not find package...), not sure they are really out or I'm missing the syntax for the name [12:25] did you do sudo ppa-purge ppa:name-of-ppa ? [12:26] yup [12:26] and without ppa: and with -p option also [12:33] I think they are purged, at least they do not appears in y-ppa-manager [12:44] Has someone experience about having an own update mechanism and leftover files? [12:45] I'll tried to add some of the self-updated files to /var/lib/dpkg/info/.list and they will removed correctly. Is that the way to go? To Update the listfile? [12:50] dholbach, shouldn't the QtWebKit test be a warning, not something that prevents approval (application 41)? I'll look into whether module module Ubuntu.Components.Extras.Browser UbuntuWebView will work in my situation. If the experimental features work, I should be able to make the transition [12:52] wellsb, great - let me know how it goes === dholbach_ is now known as dholbach === gatox is now known as gatox_brb [13:13] dholbach, I'm sure this is probably well documented somewhere, but I've missed a lot of the recent developments. How can I execute the click test you run before I upload to the store so I can verify there are no errors or warnings and not waste your time? [13:16] mzanetti: Do you know why a label takes so much width? http://imgur.com/OyOIsaB [13:16] mzanetti: I set the label width: contentWidth [13:18] mzanetti: nevermind I fixed it [13:21] :) [13:42] nik90: sorry, me again ;) Can you show me the autopilot-desktop details you have? http://paste.ubuntu.com/6954457/ [13:42] I don't understand at this point how you have autopilot-desktop 1.4 BUT gir1 1.38 [13:43] nik90: the MP that fixes the layouts crash: https://code.launchpad.net/~zsombi/ubuntu-ui-toolkit/layouts-fix1280359/+merge/206930 [13:43] ybon: http://paste.ubuntu.com/6954464/ [13:43] or in other worlds why my version of autopitlo-desktop want gir1 > 1.39 [13:43] thanks :) [13:43] nik90, zsombi got your fix :) https://code.launchpad.net/~zsombi/ubuntu-ui-toolkit/layouts-fix1280359/+merge/206930 [13:44] Kaleo: yup he just told me :) [13:44] Kaleo, zsombi: Will test it immediately..I finally figured out to build SDK source branches and test stuff out :) [13:44] nik90, nice ) [13:44] ybon: np [13:44] ah, we don't have the same exact version [13:44] 1.4+14.04.20131125bzr397saucy0 vs 1.4+14.04.20131125bzr411saucy0 [13:44] zsombi: it is funny how the solution is few lines but the test cases are huge :) [13:45] ybon: I told you I am running the autopilot experimentat PPA(trunk) while you are using the 1.4 PPA [13:45] aaaahh, damn, missed this point, sorry [13:45] ybon: mine gets updated every few days :) [13:45] let's upgrade once again, so :) [13:46] ybon: ensure you first purge your 1.4 ppa properly [13:46] nik90: yes :) and there were more than 3 lines this time, as there was an other issue with items removed during re-layouting [13:46] yep :) [13:46] randomcpp: hey ... its been a long time..how is it going? [13:48] rhââ, still not the same: 1.4+14.04.20131125-1experimental409~ubuntu13.10.1 [13:48] I'm studying for exams :S how are you? [13:48] nik90: this one ppa:autopilot/experimental ? [13:49] maybe in march I will be able to work again on my projects, I hope [13:49] zsombi: it works! No more crashes yay :) [13:49] randomcpp: good luck with your exams.. I am doing good [13:49] nik90: überüawesomness :D [13:49] randomcpp: looking forward to saucy bacon new features [13:49] ybon: yeah [13:50] ehh, I mean über-awesomness [13:50] ybon: https://launchpad.net/~autopilot/+archive/experimental [13:50] I have some planned, I will develop them as soon as possible when I'll have time :) [13:50] I promise [13:50] randomcpp: I also got time to publish my own app as well :) [13:51] nik90: are we sure autopilot-desktop comes from this ppa? [13:51] Kaleo: :) (15.49.20) nik90: zsombi: it works! No more crashes yay :) [13:51] nik90, if you have suggestions just post them on saucybacon issue tracker on github [13:51] nik90, nice :) [13:51] randomcpp: definitely like old times :) === gatox_brb is now known as gatox [13:51] yeah :) [13:51] :) [13:52] ybon: it seems not [13:52] ybon: I am not sure if autopilot and autopilot-desktop are the same thing [13:52] me neither :) [13:52] but my tests are failing and it's the only package I have different from yours [13:53] so I need to test updating it to see if I can finally run autopilot tests, and push MRs :) [13:54] ybon: just continuing proposing MRs. Jenkins will test your branches until then [13:54] ah, right [13:54] ybon: the only drawback of this situation is you are unable to run your tests, but for now don't let that stop your main work which is the calendar code [13:54] well, let's try upgrading anyway, just in case :) [13:54] ybon: upgrading to 14.04? [13:54] no no :) [13:55] just trying to upgrade autopilot-desktop [13:55] ybon: ah ok [13:55] I'm more comfortable if I can run my tests :) [13:55] so now that i've a possible explanation, let's try to solve, and if it's still not work, I will follow your advice :) [13:58] ybon: ok..you can then later ask in the autopilot channel around 19:00 UTC to get help if it still doesn't work. [13:59] I will be in my plane for Berlin at this time ;) [13:59] but I will in the incoming days :) [13:59] and thanks for your help :) [14:00] np [14:01] ap-desktop weems here https://launchpad.net/ubuntu/+source/autopilot in the description, but then not in the ppa list https://launchpad.net/~autopilot/+archive/1.4 [14:02] (not event https://launchpad.net/~autopilot/+archive/ppa ) [14:07] dpm or rpadovani: https://code.launchpad.net/~mzanetti/reminders-app/fix-compile-warnings/+merge/206938 [14:08] looking at it now... [14:10] network and compilation race :D [14:12] mzanetti, confirm, no warning, good work :-) [14:14] mzanetti, why do you change int i in unsigned int i? It's a question to learn, not a doubt on your code ;-) [14:14] rpadovani: because size() from stdlib returns a unsigned int [14:14] rpadovani: and the compiler was complaining that I compared int and unsigned int [14:15] mzanetti: am I right in saying you said you'd file a bug about the multi-pagestack header issue? [14:15] rpadovani: which might cause an issue if we have more than 2 and a half million notes [14:15] mzanetti, ahhh, ok, thanks :-) [14:18] mzanetti, reviewed and commented [14:22] dpm: :D [14:22] :) [14:22] dpm: I'm happy to explain stuff to you if you are interested [14:22] just ask [14:24] mzanetti, that's fine. I am interested indeed (it's been quite a while since I programmed in C and a bit in C++, and sometimes I miss it a bit, perhaps in a masochistic way :), but I might not have time today. [14:25] dpm: btw. the bit at line 9 is still exactly the same code as before... [14:25] dpm: I just put a #ifdef in between to suppress a unused parameter warning [14:26] dpm: because the parameter "policy" is only used when HAVE_SCHED_GET_PRIORITY_MAX || HAVE_SCHED_GET_PRIORITY_MIN [14:26] so without those its: static int toPthreadPriority(POLICY, PRIORITY priority) { [14:26] and with those its: static int toPthreadPriority(POLICY policy, PRIORITY priority) { [14:26] mzanetti, gotcha, thanks. Yeah, I wasn't sure when or why those defines would be defined [14:27] seb128, kenvandine: hi guys, I have a fix for the ItemSelector regression, here you go for testing: https://code.launchpad.net/~zsombi/ubuntu-ui-toolkit/itemselector-fix1275861/+merge/206945 [14:27] zsombi, great, I'm going to test in a bit, thanks! [14:27] Kaleo: ^ also, could you pick it up for review? [14:29] * zsombi brb [14:29] zsombi, awesome! [14:30] zsombi, nice!! [14:30] seb128, kenvandine, sorry for missing it [14:31] seb128, kenvandine: yep, sorry guys, it slipped our eyes during the sprint :( [14:33] Kaleo, zsombi: np, thanks for fixing it! === gatox is now known as gatox_lunch === gatox_lunch is now known as gatox [16:45] nik90: ping [16:45] balloons: pong [16:46] nik90: so still looking into things today. Any changes from your end? I'm assuming something changed that broke clock at this point [16:46] balloons: I did not dig into it today...I am busy today with convergence stuff.. [16:47] nik90: no worries, just checking in and letting you know what was up [16:47] balloons: I have a feeling EDS could be a potential culprit since I noticed lots of console output when you delete alarms [16:47] balloons: however I am not sure why it causes trunk to fail since there the alarm tests are skipped [16:47] nik90: I see alot of EDS noise too.. is there a rev pre-EDS to try? [16:48] balloons: well that is like rev 200 or something..pre-alarms period [16:48] I don't believe EDS is the reason the app fails to launch however [16:48] balloons: +1 [16:48] but there seems to be an issue with it too :-) [16:48] nik90: ok, I'll go run a rev from there just to see :)) [16:48] balloons: at this point, my guess would be u1db since that was the main change in the backend, but you tested the rev before the u1db transition and it still failed for you [16:49] balloons: let me know how your investigation goes :) [16:49] nik90: yep.. again, trunk failing, and pre-u1db failing, etc.. It can't be clock :=-) [16:50] balloons: I guess then as far as I know the packages that changed in the past week (which affects clock) are EDS and ui-toolkit [16:51] balloons: can you try an older ui-toolkit release? [16:51] balloons: let's leave EDS to the end [16:51] nik90: I cannot, but yes I agree we need to rollback dependencies until we find out what's up [16:52] balloons: wait you mean in the archive itself? or just locally in your computer? [16:52] nik90: on the jenkins box specifically, as we can't reproduce locally [16:52] balloons: just a heads up, I need the latest ui-toolkit for my upcoming critical patches :P [16:52] nik90: yes, textfield emulator :-) [16:52] balloons: well that and the performance and convergence patches for MWC [16:53] nik90: I think part of the trouble for rolling back jenkins is it's used by more than clock [16:53] balloons: isn't jenkins just a normal desktop running 14.04? [16:53] balloons: it is really weird that we cannot reproduce the failures on our laptops [17:05] hmm really weird I don't see much of popey these days...I need a test bot popey :) [17:36] WOW, how many awesome partenership! [17:36] http://insights.ubuntu.com/news/growing-app-ecosystem-for-ubuntu-phones/ [17:44] neat [17:47] dpm: popey: can you please test lp:~mzanetti/reminders-app/tablet and tell me what you think? [17:47] popey: I've been playing with another way of working around our current issues. [17:47] imo quite successful, but with downsides. [17:48] it's not polished yet, it'll break if you try to save a note in editpage and the editpage needs to be adapted to make the cheating less visible. [17:48] but if you're ok with the direction in general I can fix that [17:50]