=== chriadam|away is now known as chriadam | ||
=== salem_ is now known as _salem | ||
=== chriadam is now known as chriadam|away | ||
=== hypera1r is now known as hyperair | ||
dholbach | good morning | 07:10 |
---|---|---|
=== chriadam is now known as chriadam|away | ||
justCarakas | good morning dholbach | 08:31 |
mivoligo | hi all, I'm looking for a nice and simple example of using ContentHub. I tried the one on https://developer.ubuntu.com/api/apps/qml/sdk-15.04/Ubuntu.Content.index/ but it only worked with Pictures and Videos ContentType. I want to be able to import Music | 08:38 |
faenil | DanChapman: actually, let's move it here :) | 08:53 |
faenil | I'll be visiting a flat during Dekko meeting today, so I was wondering if I could help you with anything ;) | 08:53 |
faenil | like, if you have questions, need help, or anything like that :) | 08:53 |
faenil | since I won't be able to attend the meeting | 08:53 |
rhuddie | Elleo, hello, would you be able to give this MP for ubuntu-keyboard a review please? https://code.launchpad.net/~canonical-platform-qa/ubuntu-keyboard/fix-1467449-autopilot-key-names/+merge/262576 | 08:59 |
dholbach | hey justCarakas | 09:04 |
DanChapman | faenil: sorry delivery arrived. anyway.... ok that's cool. I don't have any specific questions right now. If any come up during the meeting I'll try and grab you at some other point :-) | 09:20 |
DanChapman | faenil, actually quick question. What's the plans with the bottom edge component? | 09:22 |
DanChapman | like will it just be an empty container for developers to fill, or will it have ways to define multiple views depending on which tab is pulled up in a multicolumnview. | 09:25 |
snizzo | mhall119: have I to play the game to get an invite for the mx4 also if I'm a (more less) longtime app dev? | 09:26 |
faenil | DanChapman: afaik there will be a bottom edge hint provided in the SDK....just the hint (i.e. the button at the bottom), all thought with convergency in mind | 09:26 |
DanChapman | faenil: oh just the hint | 09:26 |
faenil | yes...there is a discussion ongoing about whether we should provide a default template for the panel as well | 09:26 |
faenil | it depends on our resources, we've got a lot of stuff to work on, and the bottom edge is something that should be up to the application by design | 09:27 |
faenil | but at the same time, it'd be good to have some default bottom edge panel template | 09:27 |
faenil | which covers the "normal" usecases | 09:28 |
faenil | DanChapman: ^ | 09:28 |
mcphail | faenil: can I +1 that idea. The developer docs emphasise the role of the bottom edge but it isn't apparent how you actually implement one | 09:28 |
faenil | mcphail: because at the moment there isn't even a bottom edge hint (button, whatever) provided in the sdk, it's all up to the app | 09:29 |
faenil | there will be a hint, I'm just not sure whether we'll be providing a panel as well...we'd like to | 09:29 |
mcphail | faenil: and just about every app does it "badly" (to a greater or lesser extent) so a default panel would be good | 09:30 |
faenil | mcphail: most of the apps at the moment are using the same component, PageWithBottomEdge, which is not an sdk one | 09:31 |
DanChapman | faenil: ok great thanks. Probably easier right now that it is something custom made as dekko is still some way from moving to 1.3+. But i do agree it would be nice to have something default | 09:31 |
mcphail | faenil: yes | 09:31 |
faenil | and every app copies the qml into its project, something which we all agree is really suboptimal :D | 09:31 |
mcphail | faenil: absolutely, although when the UCS gets properly integrated it will be ledd of a problem | 09:32 |
mcphail | *less | 09:32 |
faenil | mcphail: it would be good to have a panel as well, we're just so packed with tasks at the moment about the convergence story that we're finding it difficult to find a slot for that as well | 09:32 |
DanChapman | hmm do we have SplitView as part of the sdk | 09:32 |
DanChapman | ahh no that QQUick Controls | 09:32 |
faenil | yeah that's QQC | 09:32 |
mcphail | Unfortunately, PageWithBottomEdge makes the pixels on either side of the hint "dead" to touch and obscures the bottom part of the UI | 09:34 |
snizzo | popey: same question: is there an easy way for developers to get just an invite for a mx4? | 09:35 |
faenil | mcphail: yes and we need a component which behaves "decently" in a convergent world anyway | 09:46 |
faenil | s/decently/reasonably | 09:46 |
mcphail | faenil: will converged desktop have bottom edge? | 09:46 |
faenil | mcphail: at the moment, it seems it will | 09:46 |
mcphail | "interesting" decision... | 09:47 |
faenil | it's not that easy | 09:47 |
faenil | everyone agrees that it's not ideal, alternatives are being researched | 09:48 |
Elleo | rhuddie: can do next week, away on holiday at the moment | 09:48 |
mcphail | faenil: I'm quite happy with important UI elements at the bottom, but it goes against the "top left" mantra we've been hearing for years | 09:49 |
rhuddie | Elleo, oh, no problem! enjoy. | 09:51 |
ogra_ | mcphail, http://www.markshuttleworth.com/archives/1328 mark about the bottom edge ... "...the place you really express the single most important aspects of your application, because it’s the fastest, grooviest gesture in the book, and it’s all yours on Ubuntu." | 09:52 |
ogra_ | mcphail, there is really no way to have something like a "translation layer" if every developer is encouraged to implement his own custom thing | 09:53 |
mcphail | ogra_: I quite agree it should be an SDK component. Problem is, the developer docs push the bottom edge very heavily and there isn't an easy way to implement it beyond the component from UCS. So it needs to go in to the SDK or UCS needs to become tightly integrated | 09:55 |
mcphail | ogra_: if it was the latter, the component in UCS would attract bug fixes | 09:56 |
mcphail | Best bottom edge I've used is dekko | 09:57 |
DanChapman | \o/ | 09:57 |
mcphail | ha! | 09:57 |
ogra_ | mcphail, how could it be an SDK component if we give the developer "absolute freedom" ? | 10:07 |
ogra_ | UCS wont help either if i develop my complete own idea and dont push it to UCS | 10:08 |
meles | How can I purge an app on the phone? | 10:09 |
mcphail | ogra_: tbh, I don't see how having an SDK component for a bottom edge is any less restricting than having an SDK component for a Page. The latter still gives freedom | 10:11 |
mcphail | and even sabdfl's blog post imposes (voluntary) restrictions on the absolute freedom by pointing at the design guidelines | 10:13 |
ogra_ | mcphail, well, i personally have never used the SDK at all for my QML apps ... i just throw together my QML code as it fits and push it to the store ... while the toolbar is an SDK component ... if i implement a "triangilar menu" in the bottom edge there wont be any generic way to translate that to a desktop app | 10:13 |
mcphail | ogra_: so having the component in the SDK isn't going to hinder you at all. Evryone is a winner! | 10:14 |
mcphail | ogra_: whereas it will make life easier for numpties like me | 10:15 |
ogra_ | no, the desktop user loses, since they will have to use my "triangular menu" that is optimized for touch | 10:15 |
ogra_ | and has no way to do a generic translation to a usable desktop thing | 10:15 |
mcphail | ogra_: but that is a design decision you have made... | 10:15 |
ogra_ | because i followed the "it's all yours" paradigm we promote | 10:15 |
mcphail | ogra_: If you choose to avoid the SDK, you won't get the benefits of the SDK. I can't see that's a problem | 10:16 |
ogra_ | if it isnt all mine, we need to say that ... if i can only use SDK components so it works in convergence, we need to promote that restriction ... but we don't currently | 10:16 |
mcphail | ogra_: difficult to promote an SDK component which doesn't exist :) | 10:17 |
mcphail | ogra_: the triangular menus in dekko work well on the desktop anyway | 10:18 |
ogra_ | i'm not saying that ... i'm saying we should tell devs that they have not the ultimate freedom ... but have to operate within the SDK boundaries | 10:18 |
ogra_ | mcphail, i meant a hypothetical triangular menu that doesnt exist and is possibly implemented in a 1000 line javascript in my app | 10:19 |
mcphail | ogra_: I'm all for pointing people towards the SDK. Sound's like you'll need to persuade sabdfl about that, though, as he seems to think that will stifle innovation | 10:20 |
ogra_ | well, i think we need to make clear in our docs that it isnt all that freedom we promote currently | 10:21 |
ogra_ | meles, press and hold the icon in the app scope and you get a "remove" option | 10:22 |
mcphail | ogra_: developers are going to ignore the guidance anyway. Look what happened with android, despite strict guidance on UI components | 10:22 |
ogra_ | well, i doubt our development can go as much out of bounds as androids ... but yeah, i understand what you mean | 10:24 |
meles | ogra_ does this remove the cached install files? I'm having an issue with the clock app, which I build myself. Now I would like to clean it up and install it fully from the software center. | 10:27 |
ogra_ | meles, ah, well, take a look under ~/.cache/ and ~/.cache/QML/Apps/ ... and also in ~/.config ... might well be that some data stays around there | 10:29 |
mcphail | and ~/.local/share | 10:30 |
meles | #ogra_ thanks it was indeed in ~/.cache/QML/Apps I only checked /var/cache. | 10:36 |
=== athairus is now known as athairuzzz | ||
nik90 | Is there a way to trigger a manual listview model refresh? | 11:02 |
nik90 | the model I am using doesn't seem to fire the changed signal and thereby the listview is showing old data | 11:03 |
DanChapman | nik90: is it a qml ListModel or a QAIM based model? | 11:24 |
=== MacSlow is now known as MacSlow|lunch | ||
dholbach | I'll have a look at your dev.u.c MP, so I get a bit more familiar with how the importers work | 11:57 |
dholbach | mhall119, ^ | 11:57 |
dholbach | mhall119, maybe you can have another look at mine later on? | 11:57 |
nik90 | DanChapman: strangely it was a QT LocalStorage database whose data was returned as a var which was used as the listitem's model. | 12:18 |
nik90 | DanChapman: I worked around it by converted the var into a ListModel and then feeding that to the ListView. | 12:18 |
nik90 | s/converted/converting | 12:18 |
=== _salem is now known as salem_ | ||
dholbach | mhall119, can you maybe give me a bit of background, so I can review the docs importer? | 12:27 |
dholbach | mhall119, from a fresh branch with imported docs, I can't run "manage.py import_sphinx" (api_docs.models.DoesNotExist) | 12:28 |
dholbach | mhall119, and ./update-apidocs.sh gives a lot of warnings/errors too | 12:28 |
=== Neo31` is now known as Neo31 | ||
balloons | mhall119, are you wanting me to review lp:~mhall119/developer-ubuntu-com/add-autopilot-scopes-docs? | 13:26 |
dpm | balloons, after seeing dholbach's reply about the clock unit tests and accessing the network... one thing I don't understand is why the tests don't fail on the core apps PPA. Even if Jenkins triggers the build... ultimately the PPA builders with no network access will do the build | 13:53 |
popey | nik90: Elleo any thing you can do about unsupported media types in podbird? | 14:09 |
popey | e.g. http://static.aboveandbeyond.nu/grouptherapy/podcast.xml contains m4a files. | 14:09 |
popey | which don't play | 14:09 |
balloons | dpm, it's unclear how the tests in trunk started failing. nik90 any insights for why trunk clock unit tests fail? In theory it simply means the last time they merged / built, they did pass. Since the tests didn't change, something else did | 14:10 |
Elleo | popey: don't think there's much we can do, just needs support adding in media-hub | 14:11 |
Elleo | popey: and iirc it's not actually that m4a is unsupported | 14:11 |
Elleo | popey: it's specifically m4a with jpeg streams (for cover art) | 14:11 |
Elleo | popey: there's a media-hub bug logged for that somewhere iirc | 14:11 |
Elleo | popey: an ocean o | 14:11 |
Elleo | oops | 14:11 |
Elleo | popey: https://bugs.launchpad.net/media-hub/+bug/1408681 | 14:12 |
ubot5 | Launchpad bug 1408681 in Media Hub "Certain m4a files with an embedded cover art jpeg stream are not playable" [High,Triaged] | 14:12 |
popey | haha, the guy asking me about this filed that bug :) | 14:12 |
Elleo | heh | 14:12 |
=== iphone is now known as Guest54540 | ||
nik90 | Elleo: hey did you already merge lp:~nik90/podbird/17-migrate-mainpage-listitems ? | 16:56 |
nik90 | Elleo: if you haven't then I would recommend not doing just yet. I noticed one issue with it that I fixed locally, but haven't pushed yet. Let me do a couple more tests before it gets merged in trunk. | 16:57 |
Elleo | nik90: okay, haven't done anything with that one yet | 16:57 |
nik90 | Elleo: I'm hitting this bug https://bugs.launchpad.net/ubuntu/+source/ubuntu-ui-toolkit/+bug/1465582. The fix for it is to show the episode description in a dialog and change the default click behavior to play the episode. | 16:59 |
ubot5 | Launchpad bug 1465582 in ubuntu-ui-toolkit (Ubuntu) "TrailingAction widths needs to be clamped" [High,Triaged] | 16:59 |
Elleo | nik90: okay, I'd like play to be the default anyway with the move to swiping to the side; otherwise the thing that people want to do most frequently seems a bit hidden | 17:00 |
nik90 | Elleo: true I discussed this extensively with Kevin..initially we thought of showing the play button always..but then people might think that's the only button and might not know about the trailling edge actions. | 17:01 |
nik90 | so yeah default click to play seems like a good idea. | 17:02 |
Elleo | okay, cool | 17:03 |
=== dpm_ is now known as dpm | ||
=== athairuzzz is now known as athairus | ||
nik90 | Elleo: alrite finished patching the branches. | 19:52 |
Elleo | nik90: cool | 20:20 |
nik90 | Elleo: btw I think I might have an idea to workaround the app freeze on startup. At the moment it refreshes the list *every* time the app is started which is a bit of a waste. May be we should refresh the episodes list once every 4 hours or something. | 20:22 |
nik90 | It doesn't fix the issue, but definitely improves the experience quite a bit I think | 20:22 |
Elleo | nik90: sure, sounds reasonable | 20:23 |
zbenjamin | nik90: hey | 20:49 |
zbenjamin | nik90: i hacked some example projects for ucs https://code.launchpad.net/~zeller-benjamin/+junk/ucstemplate | 20:49 |
zbenjamin | nik90: there is a qml-only case, a c++-only case and a mixed case. Only the mixed case does not completely work. Code completion somehow is broken | 20:50 |
nik90 | zbenjamin: wow awesome | 20:51 |
nik90 | zbenjamin: what do you mean by mixed case? | 20:51 |
zbenjamin | nik90: like a c++ component that also has qml files | 20:51 |
zbenjamin | nik90: sadly it requires some code in the plugin that is not yet released to completely work | 20:51 |
zbenjamin | nik90: executing won't work yet | 20:51 |
nik90 | ah ok | 20:51 |
zbenjamin | nik90: but you can already see that code completion works nicely | 20:52 |
nik90 | zbenjamin: I will try it out tomorrow morning. Looking at the code at the moment. | 20:53 |
zbenjamin | nik90: ok, the baseapp is just a plain app from one of our templates | 20:54 |
zbenjamin | nik90: the other 3 are the provided components | 20:54 |
nik90 | zbenjamin: so you added the qml component with the correct qmldir definitions and also edited the pro file? | 20:54 |
zbenjamin | nik90: yeah, the important thing is that the source folders are structured correctly | 20:55 |
zbenjamin | nik90: they need to look like a correct qml import path | 20:56 |
zbenjamin | otherwise code completion breaks | 20:56 |
zbenjamin | thats something the component authors need to be aware of | 20:56 |
nik90 | yes I noticed that | 20:58 |
nik90 | zbenjamin: We can check that when they submit the component | 20:58 |
nik90 | zbenjamin: do you know any good links to learn more about qmake? | 20:59 |
zbenjamin | not rly, only the qt documentation | 20:59 |
nik90 | I tried it out with some of my new projects but at some points couldn't figure out how to inform qmake to install certain directories at a certain place | 20:59 |
nik90 | ok | 20:59 |
nik90 | I will check it in more detail | 20:59 |
zbenjamin | nik90: thats a install target | 21:00 |
zbenjamin | nik90: http://doc.qt.io/qt-5/qmake-advanced-usage.html#installing-files | 21:00 |
zbenjamin | nik90: http://doc.qt.io/qt-5/qmake-manual.html | 21:00 |
nik90 | the qmake manual seems really well described..thnx a lot | 21:01 |
zbenjamin | nik90: yw | 21:02 |
Elleo | nik90: http://mikeasoft.com/~mike/podbird-sync.png synced from my phone to the web service :) | 21:17 |
Elleo | nik90: next step, getting the server to poll them all regularly and issue notifications... | 21:18 |
nik90 | Elleo: ooooh awesome! | 21:43 |
Elleo | :) | 21:43 |
nik90 | Elleo: does it use the ubuntu one account to create user accounts? | 21:44 |
Elleo | nik90: it just has its own accounts at the moment, will probably switch to something oauth based with future versions though | 21:44 |
nik90 | ok. this is cool stuff. | 21:45 |
nik90 | Elleo: btw will your server be able to handle polling and sending notifications to 2000+ users? | 21:48 |
nik90 | I guess the common podcasts wil be polled only once while the notifications will be sent individually to each user | 21:48 |
Elleo | yeah, its setup so it'll poll all podcasts once and send notifications to all users subscribed to them based on that user's last update time (so they don't get notifications for things they already know about) | 21:51 |
Elleo | we can see how the server copes, I need to shift onto bytemarks cloud system at some point anyway which will give a bit more flexibility if we need to up the resources dedicated to it in the future | 21:52 |
Elleo | but in the podbird settings it also lets you specify the server address, so if it gets unsustainably large people can run their own servers as a fallback | 21:52 |
Elleo | (or if people have privacy concerns or similar) | 21:53 |
nik90 | Elleo: yeah I was thinking of that since people with privacy concerns might not like sharing their subscriptions | 21:53 |
nik90 | Elleo: also pretty sure that if it goes to the point where we cannot sustain large userbase, we could open it up for donations. | 21:54 |
nik90 | It seems people *really* want to donate to podbird ;P | 21:54 |
Elleo | heh, yeah | 21:54 |
Elleo | we manage to run libre.fm mostly on donations | 21:54 |
Elleo | although there we get a lot of server stuff donated to us too | 21:54 |
nik90 | ah | 21:55 |
nik90 | hey you never know what will happen..people might surprise us | 21:55 |
Elleo | yeah, well I don't mind backing it financially myself up to a point as long as canonical keep paying me generously ;) | 21:56 |
nik90 | haha | 21:57 |
Elleo | but I expect my old VM will cope with the current number of users | 21:57 |
Elleo | especially since they probably won't all bother registering for accounts | 21:57 |
nik90 | Elleo: well if you want push notifications then you need a account, no? | 21:58 |
Elleo | yeah | 21:58 |
nik90 | but I guess not all users might not want push notifications | 21:58 |
Elleo | yeah, some people just won't care enough about that to want to go through the hassle of signing up to yet another service | 21:58 |
nik90 | true | 21:59 |
nik90 | btw feel free to share your progress on the g+ page to generate interest | 21:59 |
Elleo | plus we can probably fund it by demanding that popey gives us back-handers to feature the ubuntu uk podcast as the pick of the week for all time ;) | 21:59 |
Elleo | good idea | 22:00 |
nik90 | lol ... yes | 22:00 |
Elleo | set up a bidding war between them and bad voltage | 22:00 |
Elleo | then we'll be set for life | 22:00 |
nik90 | ;) | 22:00 |
* nik90 adds it to his plan of world domination | 22:01 | |
nik90 | s/his/podbird | 22:01 |
Elleo | heh | 22:01 |
=== salem_ is now known as _salem | ||
popey | \o/ | 22:06 |
nik90 | popey: so much fun to pull you into everything ;P | 22:10 |
* nik90 is curious as to when that started to happen in the community ;) | 22:10 | |
nik90 | popey, mhall119: Could you check why Podbird page shares on Ubuntu App Developer community are not visible. It only works if either me or Elleo share the post. May be its stuck in the review queue to prevent spam? | 22:14 |
Elleo | ah yeah, I forgot about that | 22:14 |
akiva_ | I have returned | 23:20 |
=== akiva_ is now known as akiva-thinkpad |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!