Mirvmorning, and yes QCommandLineParser usage should lead to >= 5.2.0 dependencies automatically04:06
=== karni is now known as Guest47462
=== peTERR is now known as Lunner
=== MacSlow is now known as MacSlow|errand
MerucodClassicIs there any effort to port Unity over to bsd?08:18
CimiI had connection issues, they fixed this morning08:38
tsdgeosSaviq: ping08:47
Saviqtsdgeos, pong08:47
tsdgeosSaviq: can you give  lp:~aacid/unity8/unity8_real_async_loader  a quick try and see what you think of the behaviour?08:47
Saviqtsdgeos, yup08:47
tsdgeosi think it looks faster/smoother and acceptable in look&feel08:48
tsdgeosand my interwebs crashed08:56
tsdgeosSaviq: not sure you got08:56
tsdgeos<tsdgeos> i think it looks faster/smoother and acceptable in look&feel08:56
tsdgeos<tsdgeos> but would like a more impartial opinion before i make it pass tests etc08:56
Saviqtsdgeos, yup, doing08:58
Guest47462Hi guys. Is it possible to point a result of one scope to another scope? i.e. trending tweets -> the trending scope result points to the 'twitter' scope with proper search term09:03
=== Guest47462 is now known as karni
Saviqmhr3, ↑09:03
karniThat would be an annotation, I suppose (?)09:04
Saviqkarni, we call those "canned queries"09:04
karniah cool09:04
Saviqkarni, I'm not sure what exactly their status is, but it's coming09:04
* karni wants to know moar :)09:04
mhr3karni, yes it is09:04
karniTell me about it please :) Or point me to the docs (good joke, right ;D?)09:05
mhr3you'd just return ActivationResponse with a query to run09:06
mhr3same thing scopes scope does really09:06
karnimhr3: do remote scopes support that as well?09:06
mhr3not atm09:07
mhr3there's a bug about it09:07
Saviqtsdgeos, it does feel a bit better... I feel that we should do even better than that...09:07
mhr3karni, they will09:07
karnimhr3: I see09:07
karnimhr3: until then, I suppose I can point to twitter from the preview instead.09:08
tsdgeosSaviq: there isn't much more things i can do, basically everything is async now except the text and the card.qml creation itself09:09
tsdgeosbut card.qml is just the text and the AbstractButton09:09
tsdgeosbut i'll keep digging :)09:10
mhr3karni, fyi https://bugs.launchpad.net/unity-scopes-api/+bug/130514709:11
ubot5Launchpad bug 1305147 in unity-scopes-shell "Canned queries need to be supported via "scope://" uri schema" [Undecided,New]09:11
Saviqtsdgeos, well, sure, but if you just put a Rectangle { color: "grey" opacity: 0.3 } or so behind the actual delegate (and load whole Card.qml async), it's totally smooth09:11
karnimhr3: subscribed, thanks!09:12
Saviqtsdgeos, so maybe we should think of something like that09:12
tsdgeosSaviq: yes, we can do that09:12
mhr3karni, oh wait, i think i misunderstood09:13
Saviqtsdgeos, but this is a step in the right direction for sure09:13
karnimhr3: no, I think this is exaclty it09:13
mhr3karni, you can't override what happens on tap, that always opens a preview09:13
karnimhr3: suppose "icecream" is trending topic right now on twitter. I can point from "Twitter Trending" scope to "Twitter" scope with 'icecream' search term09:13
mhr3unless it's an annotation09:13
Saviqtsdgeos, I'm only worried that once we put a workaround in place, we won't look at actually making the Card itself faster ;)09:13
tsdgeosSaviq: well, not sure, i mean if we go for the rectangle and card.qml totally async, i'm not convinced we need this09:13
mhr3karni, and we don't support annotations yet either :)09:14
karnimhr3: ah, that's fine. button in preview is fine09:14
Saviqtsdgeos, yeah, exactly ;)09:14
karnimhr3: baby steps :)09:14
tsdgeosSaviq: yeah, true, but i'm not sure we can make Card itself much faster, other than making the UShape faster09:15
tsdgeosSaviq: so basically you tell me what you want, try to make what i have faster without the async Card, go back to async Card + Rectangle, try to make Card itself faster09:16
Saviqtsdgeos, I still worry we'd be better off with purpose-built minimal Cards, falling back to the current behemoth Card.qml09:16
tsdgeosi'm open to anything :)09:16
mhr3+1 on specializing at least the apps cards09:17
Saviqtsdgeos, I think what we need is more analysis - like simplify Card.qml down to Shape + Label, and take measurements in the process to see what's the heaviest thing there09:18
Saviqwe could also try with The New Shape™... but...09:18
Saviqtsdgeos, basically, if we can't see much improvement between a minimal "Shape + Label" and your asynchronized Card.qml, we'll know that this is not the bottleneck09:20
Saviqwhich I'm not sure is the case still, there's still too much happening in Card.qml I'm afraid09:20
tsdgeosi'll try that09:21
mhr3tsdgeos, been seeing this quite often with latest images http://imgur.com/q6fbfuC something to be worried about?10:02
tsdgeosshould not happen10:02
mhr3is there a bug opened about it?10:04
mhr3it's not like this was the first time i saw it10:04
mhr3only seems more common10:05
tsdgeosi do not know of any bug open about this10:06
mhr3Saviq, aware of one ^?10:08
Cimiseb128, https://code.launchpad.net/~unity-team/ubuntu-system-settings/wizard.wifi/+merge/212675 :)))))))))10:25
seb128Cimi, get mterry to approve it and I can put a landing10:26
Cimiseb128, he approved10:26
Cimiseb128, but sine he participated in the branch, he asks for review10:27
seb128well, I've no slot for settings atm, still busy with the desktop LTS10:28
seb128but I can try to have a look next week10:28
Cimiseb128, someone else?10:29
Cimiif you're busy..10:29
seb128Cimi, we are short on people working on settings10:29
seb128you can try to ping kenvandine when he gets online10:29
Cimiok thx10:29
seb128or tedg10:30
seb128he wrote the wifi settings10:30
seb128so he might want to review the wizard side10:30
CimiSaviq, split common merging soon?10:37
tsdgeosSaviq: it is amazignly much faster :S10:38
Saviqtsdgeos, that's what I was afraid of10:41
tsdgeosSaviq: pushed to lp:~aacid/unity8/specialized_cards10:41
tsdgeosgive it a try10:41
tsdgeosit's obviously not production ready since it's a total hack10:42
tsdgeosbut the speed difference is quite woah10:42
didrocksSaviq: I've move the IRC bot to canonistack, it's stable enough for that. Feel free to tell me if you see any issue10:43
Saviqdidrocks, just joined, thanks!10:43
didrocksgreat ;)10:43
Saviqtsdgeos, yeah, I can still make it skip frames, but yeah, it's a lot better10:45
tsdgeosSaviq: you can make it skip frames? not here, or my eyes are not that trained :D10:46
Cimitsdgeos, we love hacks10:46
tsdgeosSaviq: and that's without the UbuntuShape being async10:46
Saviqtsdgeos, oh yeah, easy - open "Available" and scroll up as fast as you can10:46
Cimitsdgeos, with async might be faster?10:46
tsdgeoswell not faster10:46
Cimitsdgeos, I mean the loader10:46
tsdgeosbut less blocky10:46
Cimiyou removed async from loader10:47
tsdgeosi know10:47
tsdgeosi don't mean that one10:47
tsdgeosi mean another one that in this branch doesn't exist10:47
SaviqCimi, we're waiting for U to open, don't want to SRU10:47
tsdgeosadding async there makes you end up with basically and empty dash if you scroll faaaaaaaaast10:47
CimiI see10:48
Cimitsdgeos, what's the bottleneck?10:48
tsdgeosGL i'd say10:48
Cimitried not using ubuntushape but rectangle?10:48
tsdgeosbut can't tell10:48
Cimitsdgeos, just to see if it's sdk or not10:49
tsdgeosCimi: yes i have10:49
tsdgeosthat's the thing i did day 010:49
tsdgeosit is obviously faster10:51
Saviqtsdgeos, so yeah, I think this is the right direction, that's the only way that will give us fast delegates, downside is obviously code duplication, but it's a tradeoff we need to take I think10:53
tsdgeosSaviq: so the question is, how we do special case10:53
tsdgeosbasically do a if template is vertical and has art and title and nothing else -> load this?10:54
Cimitsdgeos, 100% fast or not?10:54
tsdgeosCimi: you try it10:54
Saviqtsdgeos, yeah10:54
Cimitsdgeos, Saviq because this ubuntushape thing can be problematic10:54
Cimiin 2014 we can't have slow performance because we use too many shaders10:54
Saviqtsdgeos, basically select the "smallest" delegate that can deal with what the template and components describe10:55
Cimiall other competitors are going flat (no shaders almost)10:55
Cimiwhile we use gpu a lot10:55
Saviqwe should cache the shaped images, is what we should...10:56
CimiSaviq, vram then :)10:56
Cimibut you still need to compute them once10:56
Saviqshaping them in GPU every time is kinda wasteful, when 90% of the time they will look the same every time10:56
SaviqCimi, sure, once10:56
SaviqCimi, you have to download them once, too, etc.10:57
CimiSaviq, I think those are questions we need to answer asap10:57
CimiSaviq, or we better use less this ubuntushape10:57
Cimiwhich is abused10:57
Cimieven when imho doesn't look nice10:57
Cimi(like everywhere)10:58
Cimigallery app for example10:58
CimiI'd rather have squares without padding10:58
SaviqCimi, and why are you telling me this?10:58
Cimifaster, focusing on the content10:58
SaviqCimi, don't you work from London, where... guess what! all the design happens!10:58
CimiSaviq, because you're my colleagues? :)10:59
CimiSaviq, well I first talk with friends/colleagues, if they agree I can jump and ask marcus10:59
SaviqCimi, we don't have to agree ;D10:59
Cimiwhy not!11:00
SaviqCimi, but worse than that, I don't think I can be objective any more11:00
SaviqI've gotten used to it, and probably a little tired, too11:00
=== alan_g is now known as alan_g|afk
CimiSaviq, it's exactly what I want!11:00
Cimiusers should not be tired!11:00
SaviqCimi, but we're not users11:01
CimiAqua on OS X11:01
Cimitheir theme11:01
Cimiit has 14 years11:01
Cimistill people love it11:01
Cimithey nailed it11:01
Cimigood designs can last for long11:01
Cimiif we're tired of something in less than a year...11:02
mhr3pff Aqua sucks :P11:02
mhr3it's dated11:02
CimiI like the ubuntu shape11:02
Cimiwe can probably use it with more balance11:02
Cimiwe gain in performance, battery usage, ram11:02
=== MacSlow is now known as MacSlow|lunch
Cimiand will feel less tiring11:03
Saviqtsdgeos, Cimi, bug #1311599 btw11:03
ubot5bug 1311599 in ubuntu-ui-toolkit (Ubuntu) "UbuntuShape should be available as an image provider" [Undecided,New] https://launchpad.net/bugs/131159911:03
tsdgeosthat'd help11:04
SaviqCimi, oh yeah, I'm tired of Aqua even though I don't use OS X at all ;)11:04
Cimimhr3, https://www.youtube.com/watch?v=vXQlYhLcQ2c11:05
CimiSaviq, ok but after 14 years for god sake!11:05
mhr3Cimi, ok, you got me :P11:05
CimiSaviq, it lasted enough...11:06
CimiSaviq, thanks for the bug11:07
CimiSaviq, +!11:07
Cimianother example of too many ubuntushape is the calender11:08
CimiI was talking with matthieu about this11:08
Cimi(he hates all those ubuntushapes everywhere)11:09
mhr3but he's working with them all day long11:10
mhr3Cimi, it's like you hating mir :P11:10
Cimimhr3, it's fundamentally different11:11
Cimimhr3, if mir will be as fast as wayland/sf it won't change anything for me, I chase performance11:12
Cimimhr3, but we are talking about a design here, different11:12
mhr3the point was that you don't have outsider perspective11:13
=== alan_g|afk is now known as alan_g
Cimimhr3, http://i1-news.softpedia-static.com/images/news2/Remember-Ubuntu-Touch-Apps-and-Unity-8-Are-Available-for-Ubuntu-14-04-LTS-438499-3.jpg11:14
Cimimhr3, http://r.phonedog.com/shared/images/2014/1/180699-newmotoxgallerycrop.jpg11:15
Cimimhr3, it's design choice, the latter focuses on the content, photo, the reason why I am running the app11:16
Cimithe photos are prioritised11:16
Cimiin our app, there's lots of padding and corners cropped too11:17
mhr3if you want to look at few last shots, yes11:17
Cimiwith all the issues of performance too11:17
mhr3but the app is much better as a photo journal11:17
Cimimhr3, you can have the same result with rectangles11:18
mhr3also, unfair comparison, bunch of nice pictures on an actual phone vs screenshot of screenshots of website11:18
Cimimhr3, http://cloud.addictivetips.com/wp-content/uploads/2013/10/Ubuntu-Touch-16.2-Apps-Gallery-Photos.png11:19
Cimimhr3, http://photos.pcpro.co.uk/blogs/wp-content/uploads/2013/01/Ubuntu-Phone-photo-gallery.jpg11:19
mhr3see, this second one isn't that bad11:20
mhr3although yes, quite a lot of empty space11:20
mhr3Cimi, but don't forget that the ushape is also making the platform as a whole distinct11:21
Cimimhr3, I agree11:21
Cimimhr3, but it's already everywhere11:21
mhr3*you* see it everywhere11:22
Cimimhr3, we don't need to blindly put it in every corner11:22
Cimiwe have it for buttons, widgets, etc etc11:22
Cimihere we're really abusing it in my opinion11:22
mhr3maybe, but then again, it makes the whole thing feel more concise11:22
Cimieven when design wise (UX) is counterproductive11:23
Cimimhr3, not necessarily concise, it just makes it more repetitive11:23
CimiI agree with consistency as a general rule11:24
Cimibut with a pinch of salt11:24
rvrI need someone's review/approval for this branch https://code.launchpad.net/~vrruiz/unity8/helpers-notify/+merge/21570011:29
rvrAnd I have a question: are you able to execute the autopilot tests for Unity8 in an up-to-date Trusty?11:29
Cimirvr, first, checklist! https://wiki.ubuntu.com/Process/Merges/Checklists/Unity811:30
rvrCimi: Already done11:30
Cimirvr, sorry we usually put it in the description, but it's fine in the comments11:30
rvrCimi: Oh, next time I will do it that way11:31
Cimirvr, you should be able to run tests11:33
Cimirvr, read the CODING file iirc11:33
Cimibasically make install then running with exporting PYTHONPATH11:33
Cimisomething like that11:33
rvrCimi: What I mean is that I was able to execute them, but no longer11:40
Cimirvr, which error?11:40
Cimitrying out now11:40
rvrNo error, it gets stuck11:40
rvrUnity8 window doesn't appear11:41
Cimitsdgeos, might be the error with upstart?11:41
tsdgeoshe may have the same problem i have with the indicators yes11:42
tsdgeosi still don't understand why noone else other than me has that11:42
tsdgeosrvr: status unity811:43
tsdgeoswhat does that return?11:43
rvrLet me check11:43
rvr$ status unity811:44
rvrunity8 start/pre-start, process 225111:44
tsdgeosrvr: yeah you are unlucky as i am11:46
rvrtsdgeos: lol11:46
tsdgeosrvr: my solution to that is11:46
tsdgeosreboot or relogout/in11:46
tsdgeosinitctl emit indicator-services-start11:46
tsdgeosthen stuff will work11:46
rvrI see "initctl emit indicator-services-start" in ps11:46
rvrubuntu    2279  0.0  0.0  19388  1184 ?        S    12:28   0:00 initctl emit indicator-services-start11:46
ubot5Ubuntu bug 3308 in Launchpad itself "duplicate for #2279 Bugs on Maintained Software times out" [Medium,Fix released] https://launchpad.net/bugs/330811:46
tsdgeosyou can try killing 15 that one11:46
tsdgeossee if it helps11:47
Saviqrvr, tsdgeos, there's a branch fixing this in LP11:47
Saviqrvr, if you add the --no-wait to the unity8.conf file it won't wait for indicators to start11:47
rvrtsdgeos: Yeah, that helps11:47
rvrSaviq: Trying11:48
rvrSeems work, thank you11:52
Cimirvr, you don't have emit indicator services here11:55
rvrCimi: Where?11:56
Cimirvr, in your upstart job11:56
Cimiof your branch11:56
Cimiit's in trunk with a merge11:58
rvrCimi: My branch is related to notification tests11:58
rvrSome refactoring we are doing11:58
CimiI was looking to see if the hang was due to a not updated branch11:58
rvrAutopilot wasn't able to start unity8 anymore after I upgraded the VM11:59
rvrBut now the problem is gone with tsdgeos and Saviq's suggestions11:59
=== alan_g is now known as alan_g|lunch
Saviqrvr, yup, that'd be the same issue indeed12:00
CimiSaviq, trying to test your branch12:00
CimiI added no wait here /usr/share/upstart/sessions/unity8.conf12:01
Cimiand locally12:01
Cimistill stuck12:01
rvrI donwloaded trunk and executed ./run, and got the same result: stuck12:01
SaviqCimi, you probably have one in ~/.config/upstart12:01
Saviqrvr, copy data/unity8.conf to ~/.config/upstart/12:02
Saviqrvr, only then will upstart pick it up12:02
Saviq(the fixed one from my branch, that is)12:02
rvrSaviq: Ack12:02
CimiI dont12:03
CimiI relogin12:04
rvrYeah, me neither12:04
rvrSaviq: Just modifying it in /usr/share/ works for me12:08
Saviqrvr, sure, the one in ~/.config (if present) just overrides the system-wide installed one12:08
Saviqrvr, so on one hand it's good to put it in ~, so that you don't mess system-wide ones... but then you might forget about the one in ~, and see weird things when it gets upgraded in the package...12:09
karnimhr3: Do you know who/when will be assigned to work on the scope:// uri? (I assume annotations are far down the line compared to supporting scope:// )12:28
mhr3karni, it's "Fix committed" already12:29
mhr3just didn't land12:29
mhr3cause T/U12:29
mhr3although shell part is still missing, i'll do that12:29
CimiSaviq, tsdgeos I have the issue of autopilot restarting my session12:31
Cimiwhich was the fix for that?12:32
karnimhr3: it's fix commited in ubuntu-unity-api, not ubuntu-unity-shell. that's a difference, no?12:32
karnimhr3: ah, there you wrote it12:32
karnimhr3: yes, that's what I was asking about hehe12:32
karnimhr3: any ETA when you can get to it? (for situational awareness, let's say :) )12:32
karnimhr3: T/U ?12:33
mhr3karni, it's tiny, shouldn't take long12:33
SaviqCimi, I don't know about that issue12:33
mhr3karni, trusty / u??? transition12:33
Cimitsdgeos, how did you fir the issue with autopilot restarting your session?12:34
karnimhr3: ah heheh12:34
mhr3karni, but yea, not being able to land stuff kinda sucks atm12:35
karnimhr3: gotcha. "shouldn't take long" sounds good enough :)12:35
mhr3karni, just bug me about it for the next few days ;)12:36
karnihehheeh, gotcha12:36
Cimiok ruynning autopilot restarts my desktop session12:41
SaviqCimi, rather it sounds like something is crashing12:42
SaviqCimi, there's no reason why autopilot runs would restart your session, unless the unity8 job kills your session upstart12:42
SaviqCimi, which means you should check whether you didn't b0rk it12:43
=== alan_g|lunch is now known as alan_g
=== MacSlow|lunch is now known as MacSlow
tsdgeosCimi: i didn't _/12:57
tsdgeosCimi: i just run them on the phone12:57
Cimiok now it didn't crash12:57
Saviqwe need a way to run them on desktop in xvfb, too, but that'd basically mean a complete new session under xvfb (with dbus et al, for indicators and such)13:02
pstolowskididrocks, hey! can you take a look if this https://code.launchpad.net/~stolowski/unity-scopes-api/install-client-tool/+merge/216875 looks good from packaging pov?13:13
didrocksmaybe sil2100 you want to have a look? ^13:16
MacSlowSaviq, updated the tripe-set of branches regarding the modal-snap-decisions adjusting all the version-bumps and build-dependencies... should be all set now.13:23
=== _salem is now known as salem_
SaviqMacSlow, kk13:25
sil2100What's up13:25
sil2100pstolowski: I'll take a look in a amoment, just finishing lunch13:26
SaviqMacSlow, sorry for being an a$$ about it, it's a rather tangled approach, but the only one we could think of that made sense13:26
MacSlowSaviq, no worries... that's inherent to the nature of these three branches spreading numerous parts13:29
pstolowskisil2100, sure, no rush, thanks13:32
karniSaviq: Would we want to make Card description make links within text clickable?14:11
Saviqkarni, no14:11
karniIf not, it would make sense for me to add another preview button.14:11
karniSaviq: What's the reson behind that decision? It's heavy?14:11
Saviqkarni, unless you mean in the preview?14:11
Saviqkarni, so not card summary, but preview description?14:12
karniSaviq: oh. yeah, that's what I meant :)14:12
* karni looks at the preview right now14:12
Saviqkarni, that might be possible, but not implemented, comment on the dash spec?14:12
karniSaviq: will do!14:12
CimiSaviq, https://bugs.launchpad.net/unity8/+bug/1089962 ?14:36
ubot5Launchpad bug 1089962 in touch-preview-images "[window management] add an 'opaque' flag for the shell so that other surfaces are not composited" [High,Confirmed]14:36
SaviqCimi, that will be out of scope with QtComp14:40
CimiSaviq, ok, worth commenting14:40
SaviqCimi, yup, did14:41
pete-woodshey Saviq, did you manage to get anywhere with the infographic stuff?14:42
Saviqpete-woods, was planning to get back on it now that stuff's slightly less crazy again14:43
pete-woodsSaviq: cool, was just checking really. I understand your plate is large and full!14:44
Saviqpete-woods, yeah, I get you14:44
Saviqpete-woods, do we have a blueprint for those btw?14:44
* Saviq wants moar blueprints!14:44
pete-woodsSaviq: seems there's this - https://blueprints.launchpad.net/ubuntu/+spec/client-1311-infographics14:45
pete-woodsbut it's not exactly useful information14:45
Saviqpete-woods, right, and it's old - should probably be marked completed, and we need a new u-targeted oen14:46
Saviqbut we'll need for u to be open for that...14:47
tsdgeosSaviq: what about? https://code.launchpad.net/~aacid/unity8/specialized_cards/+merge/21688515:27
tsdgeosit's best of both worlds15:27
tsdgeosasync loaded UShape + simplified card15:27
=== gatox is now known as gatox_lunch
=== dandrader is now known as dandrader|lunch
=== dandrader|lunch is now known as dandrader|afk
paulliuSaviq: If I want to make API documents inside the code, do we have any formal style. Like doxygen? Or maybe Qt has its own way of documentation?15:32
paulliuSaviq: For qt I think it is qdoc. Though I haven't read the documents about it.15:37
Saviqpaulliu, yes, we're using doxygen15:39
paulliuSaviq: ok. got it.15:39
Saviqpaulliu, `make doc` will build docs in builddir/doc15:39
Saviqpaulliu, you can then check source files to see how we style them15:39
paulliuSaviq: ok.15:39
Saviqtsdgeos, yeah, something like this, although I think we'll have more than just this one...15:42
tsdgeosSaviq: sure, but i guess we can try to get this reviewed and in and slowly add more as we need them?15:43
Saviqtsdgeos, sure, yes15:43
Saviqgotta finish early today o/15:43
=== jhodapp is now known as jhodapp|lunch
=== gatox_lunch is now known as gatox
=== dandrader|afk is now known as dandrader
=== jhodapp|lunch is now known as jhodapp
=== dandrader is now known as dandrader|afk
=== dandrader|afk is now known as dandrader
=== ChrisTownsend1 is now known as ChrisTownsend
=== bschaefer_ is now known as bschaefer
=== dandrader is now known as dandrader|afk
=== dandrader|afk is now known as dandrader
=== jhodapp is now known as jhodapp|afk
=== salem_ is now known as _salem
=== NoNameYet_xnox is now known as xnox

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!