| pavolzetor | sorry about that rant, but sad thing is, that was true | 00:44 |
|---|---|---|
| pavolzetor | and I didn't feel particular well after it | 00:44 |
| pavolzetor | how constructive? | 00:45 |
| pavolzetor | should I spend ages to trying fix bug in stack, I have minimal knowledge of C++ | 00:45 |
| pavolzetor | I am going to wait for answer, I was just honest, and I think you might not give enough man-power to fixing small bugs, which annoys a lot. | 00:49 |
| pavolzetor | I give you example, there is new video feature in USC, and it causes crashes in fullscreen, this should not happen in LTS | 00:49 |
| pavolzetor | I do not say my app is perfect, it is far from it, but I am not core app | 00:50 |
| pavolzetor | it is not core app | 00:50 |
| pavolzetor | and about being constructive, I have not idea, how can I be constructive? I prefer to not fix bugs in compiz, as I am not developer of compiz, and it should be much easier to fix it for compiz devs, as it is their project | 00:51 |
| pavolzetor | what I wanted to say, you cannot attract developers (I mean real one, not hobbyist like me), if you don't have working toolkits (as you probably know) | 00:54 |
| pavolzetor | and other think, please, do backport new clutter release, which fixes segfault (to be constructive) | 00:54 |
| pavolzetor | tomorrow will check answer, thanks | 00:56 |
| dholbach | good morning | 07:09 |
| aquarius | hola, dholbach | 08:00 |
| dholbach | hey aquarius | 08:00 |
| pavolzetor | dpm: still no answer, and I check irc logs too | 09:37 |
| dpm | hi pavolzetor, no answer for what? | 09:38 |
| pavolzetor | you said yesterday, "great, I hope he felt well after his rant" | 09:39 |
| pavolzetor | "didn't even leave time to finish my reply" | 09:39 |
| pavolzetor | "talk about being constructive! | 09:39 |
| pavolzetor | ! | 09:39 |
| pavolzetor | " | 09:39 |
| pavolzetor | so here we go | 09:39 |
| dpm | :) | 09:40 |
| * dpm rewrites reply | 09:41 | |
| pavolzetor | waiting... | 09:41 |
| pavolzetor | you could post it yesterdayt | 09:42 |
| pavolzetor | I check logs, I do not spend all time on computer, so I check logs sometimes | 09:42 |
| dpm | pavolzetor, sorry, I'm on the phone, let me come back to you when I've finished | 10:00 |
| pavolzetor | no problem | 10:01 |
| pavolzetor | I will be there for whole day, so you cannot say it is ranting ;) | 10:01 |
| dpm | :) | 10:03 |
| dpm | pavolzetor, in that regard, app development in Ubuntu is not different from any other area of contribution in Ubuntu. You get all of the tools for free, they put a great potential in your hands and you're even able to improve them and expand them. I won't deny that there are bugs, some minor, some critical, but I trust that you're not new to open source and are aware how it works. We cannot fix every bug under the sun, so we have to focus on the thi | 10:06 |
| dpm | ngs that can be done that make the most impact. If we were to wait until all bugs are fixed to promote app development, we would never move forward, and looking at the apps in the archive people do manage to write awesome apps with the tools available. | 10:06 |
| pavolzetor | I understand that | 10:07 |
| pavolzetor | it was always like that since I started with Feisty Fawn | 10:07 |
| pavolzetor | But | 10:08 |
| pavolzetor | there were infos about making unittesting and so on, TDD | 10:08 |
| dpm | yeah, but look how far we've come since Feisty! | 10:08 |
| pavolzetor | I probably exaggerated, but I just wanted to point out | 10:09 |
| pavolzetor | that people do not care that much about new feautres | 10:09 |
| pavolzetor | it is much better to have less features and have them polished | 10:10 |
| dpm | well, that's actually what we've done with Unity | 10:10 |
| pavolzetor | how can you explain, that software-center crashes in fullscreen video? (in LTS?) | 10:10 |
| pavolzetor | I reported about 8 bugs in unity/compiz, none of them fixed ;) | 10:10 |
| pavolzetor | mainly minor stuff, but annoys | 10:11 |
| pavolzetor | ing | 10:11 |
| pavolzetor | and you cannot expect ,that I will study compiz code base just to get my app working | 10:11 |
| dpm | I believe you're picking on a very focused example, I'd say most apps, if not all, work with the recommended tools and toolkits | 10:13 |
| dpm | as per the Software Center, I cannot comment without more details, and then again, I'm not an USC developer :) | 10:13 |
| pavolzetor | it was just example | 10:14 |
| pavolzetor | of core app being buggy | 10:14 |
| dpm | what's the bug number? | 10:14 |
| pavolzetor | Okay, shotwell, gtranslator have same problem as mine app, they use GTK, I use CLutter | 10:14 |
| pavolzetor | and Gtk is probably recommended toolkit ;) | 10:15 |
| pavolzetor | https://bugs.launchpad.net/ubuntu/+source/software-center/+bug/1006959 | 10:15 |
| ubot2` | Launchpad bug 1006959 in software-center "video playback freezes, crashes in fullscreen" [Undecided,New] | 10:15 |
| dpm | it is :-) perhaps we can focus on what's the issue with your app and someone can help | 10:15 |
| pavolzetor | it is issue of compiz (probably), as unity2d, gnome shell works nicely | 10:16 |
| pavolzetor | also, stack should be up-to-date | 10:16 |
| dpm | re: the Software Center, you're pointing out to a bug that was filed less than a day ago and that hasn't been confirmed | 10:16 |
| pavolzetor | so clutter is 1.10.6 | 10:16 |
| pavolzetor | but in ubuntu is 1.10.0 | 10:16 |
| dpm | I believe clutter is not supported in 12.04 | 10:17 |
| pavolzetor | and I reported some segfaults and they were fixed, stack should be up-to-date | 10:17 |
| pavolzetor | it works here | 10:17 |
| dpm | for what you're saying it does not really work :) | 10:17 |
| dpm | anyway, what's your app and what's the issue? As I say, perhaps someone on the channel can help | 10:17 |
| pavolzetor | re: USC, I wanted to say, that it should not be in LTS, if it is buggy/not well tested | 10:18 |
| pavolzetor | https://bugs.launchpad.net/unity/+bug/993204 | 10:18 |
| ubot2` | Launchpad bug 993204 in unity "bottom of windows are not drawn" [Undecided,Confirmed] | 10:18 |
| pavolzetor | this one | 10:18 |
| pavolzetor | it affects also gtranslator, shotwell | 10:18 |
| pavolzetor | and sorry for yesterday, but I hope something will change, I believe, bugless stack is very important | 10:19 |
| dpm | pavolzetor, there is a team of highly skilled developers working very hard on Software Center. Saying that it is not well tested without evidence, discarding their hard work and pointing to an unconfirmed bug sounds a bit rushed to me | 10:20 |
| dpm | if you want to bring that bug to their attention, they're hanging out on #software-center | 10:20 |
| pavolzetor | They can be hardworking | 10:20 |
| pavolzetor | but, nobody cares outside community, people do use APP | 10:21 |
| pavolzetor | I know, it is tough to make it works | 10:21 |
| pavolzetor | but basic users do not simply care, and it is true | 10:21 |
| pavolzetor | they just want to finish task | 10:21 |
| pavolzetor | it can be well tested, then tests are not good enough ;) | 10:23 |
| pavolzetor | okay, so any help with that compiz bug? | 10:23 |
| dpm | again, let me point out that the USC works for the vast majority of users and that you're basing your argument on an unconfirmed bug filed after release | 10:24 |
| pavolzetor | please, do me a favour :) | 10:24 |
| pavolzetor | go to software-center | 10:25 |
| pavolzetor | open some app with video inside | 10:25 |
| pavolzetor | and make it fullscreen and change volume :) | 10:25 |
| dpm | pavolzetor, why do you think it's compiz? From the bug description, I cannot really tell which package is to blame (not picking on it, just trying to genuinely help) | 10:25 |
| dpm | And the bug seems to be filed against unity, not compiz | 10:25 |
| pavolzetor | how can you say, it works for vast majority? how do you know? (errors.ubuntu.com does not show ration used without and with crash | 10:27 |
| pavolzetor | I am not sure if it is compiz, but it occurs only in unity 3d | 10:28 |
| pavolzetor | which is plugin into compiz, afaik | 10:28 |
| dpm | It may well be that it is compiz, but until it is tracked down, it's just speculation. It might be helpful to try and narrow it down to the component that might cause that behaviour. Do you know if you're using any technology that your app has in common with the other apps where you can reproduce the bug? It may well lie in there. You're mentioning Clutter, could there be anything else? | 10:31 |
| pavolzetor | It was doing with Gtk version too | 10:32 |
| pavolzetor | and the are on bottom, which is not drawn is clickable | 10:33 |
| pavolzetor | so you can manage window under it | 10:33 |
| pavolzetor | it has to do something with window manager | 10:35 |
| pavolzetor | I will send you video soon | 10:35 |
| pavolzetor | but whole purpose of this, I want to see increased quality, focus on details | 10:36 |
| pavolzetor | I can tell you, almost no app is beautiful | 10:37 |
| pavolzetor | http://ubuntuone.com/7UmNjnV2l2QMu4ZVpVzpiI | 10:38 |
| dpm | pavolzetor, oh wow, floaty is cool, is that the app you're developing? | 10:40 |
| pavolzetor | it is not cool, interface is not polished | 10:43 |
| pavolzetor | just early 0.2 | 10:43 |
| pavolzetor | and service is not async | 10:43 |
| pavolzetor | but thanks | 10:43 |
| JanC | it's difficult to say whether the bug is in compiz or in the applications or in the toolkit | 10:44 |
| pavolzetor | not very usable now | 10:44 |
| dpm | pavolzetor, I consider any attempt to write an app cool :) I understand it might not be polished yet, but looking at it you've gone a long way already in the process of app development | 10:45 |
| dpm | and I agree with JanC, the bug needs more investigation and more data before we can say where the issue is | 10:45 |
| pavolzetor | but it works in 2d and shell | 10:46 |
| pavolzetor | so it should not be in toolkit | 10:46 |
| dpm | why not? | 10:46 |
| pavolzetor | the are at bottom would not be clickable | 10:47 |
| JanC | it's like the developers found lots of bugs in ALSA drivers after PulseAudio started to use driver functions that almost nobody used before | 10:47 |
| pavolzetor | it is obvious that window manager is not handling it properly | 10:47 |
| pavolzetor | JanC: I think shotwell is pretty conmmon | 10:48 |
| dpm | it is not obvious to me without data. It might well be that some layer above or underneath does not pass correct data to the window manager | 10:49 |
| pavolzetor | dpm: about writing my app, it is sad, that you need to write app for reading RSS :/ (I used liferea and I have that paradigm) | 10:49 |
| JanC | what I mean: maybe some toolkits or applications (forget to) do something that only causes trouble in compiz | 10:49 |
| dpm | pavolzetor, what's happened to Liferea? | 10:49 |
| * dpm used to use it a long time ago? | 10:49 | |
| JanC | of course, it can also be a compiz bug ;) | 10:49 |
| dpm | :) | 10:49 |
| pavolzetor | dpm: if it would not pass correct data, other window managers would be affected too | 10:50 |
| pavolzetor | *I hated that paradigm | 10:50 |
| JanC | not if those wm don't use that data | 10:50 |
| pavolzetor | all RSS apps, same interface, I looked at al for linux/mac os x (there are few nice) and windows | 10:51 |
| JanC | findings bugs can be difficult sometimes | 10:51 |
| dpm | pavolzetor, without hard data we can only speculate. What I would recommend is to try to narrow it down to the technology that might be causing the issue. Perhaps start removing dependencies in your app until you can find what it's causing it | 10:51 |
| pavolzetor | JanC: it might be, but if I look back, in natty, there wasn't this bug | 10:52 |
| pavolzetor | and gtransaltor is pretty same | 10:52 |
| dpm | anyway, I have to go onto something else now. Let us know about the progress! | 10:52 |
| JanC | hm, what language/libs are those apps written in/for? | 10:52 |
| pavolzetor | dpm: nothing, actually they speed it up really nicely, but I do not like to clicking much, I want to use my screen proaperly | 10:52 |
| pavolzetor | dpm: thanks | 10:53 |
| pavolzetor | JanC | 10:53 |
| pavolzetor | it was is clutter/gtk | 10:53 |
| pavolzetor | and the issue was there in sole gtk too | 10:53 |
| pavolzetor | lang is python32 | 10:54 |
| pavolzetor | 3 | 10:54 |
| pavolzetor | I need to restart compiz, windows freezed | 10:56 |
| pavolzetor | that's awkward | 10:57 |
| pavolzetor | I tested it with gtk and clutter, seems same | 10:58 |
| pavolzetor | anyway, how can I get my window bar working, restarting unity/compiz doe not help | 10:58 |
| JanC | seems like gtranslator is C + Gtk (but no Clutter) | 10:58 |
| pavolzetor | ye | 10:58 |
| pavolzetor | s | 10:58 |
| pavolzetor | shotwell is Vala -> C | 10:58 |
| pavolzetor | + gtk | 10:58 |
| JanC | so the bug is probably not in the python bindings or in clutter | 10:58 |
| pavolzetor | too | 10:59 |
| pavolzetor | I need to restart X | 10:59 |
| pavolzetor | sorry about it, just somehow it freeze :) | 10:59 |
| pavolzetor | d | 10:59 |
| pavolzetor | frozen | 10:59 |
| pavolzetor | okay I am back | 11:01 |
| pavolzetor | when I close window, it saves size | 11:02 |
| pavolzetor | and size is as fullscreen window should have | 11:02 |
| pavolzetor | when yo uresize window, it works | 11:04 |
| pavolzetor | so I assume, compiz does not calculate it properly at first launch | 11:04 |
| JanC | hm, I don't think you should save the window size for maximized windows, but save something like maximized=yes, an then ask the wm to maximize it on next run, right? | 11:16 |
| JanC | I wonder if there is any difference between calling gtk_window_maximize () before or after showing the window | 11:17 |
| pavolzetor | going to chewck | 11:21 |
| pavolzetor | hmm, it works | 11:22 |
| JanC | what works? | 11:22 |
| pavolzetor | so it is definitely bug in compiz automaximize than | 11:22 |
| pavolzetor | when I use | 11:22 |
| pavolzetor | window.maixmize() | 11:22 |
| pavolzetor | there is function in compiz, which maximizes window automatically | 11:23 |
| pavolzetor | therefore, there must be bug | 11:23 |
| pavolzetor | because it is diplayed as maximized, but it is actaully not | 11:27 |
| pavolzetor | I will store if window is maximized, since it does work | 11:29 |
| pavolzetor | but there is still bug | 11:29 |
| pavolzetor | :/ | 11:29 |
| pavolzetor | thanks for help | 11:30 |
| pavolzetor | i didn't know I can workaround it this way | 11:30 |
| JanC | normally that does the same as when you double click the title bar | 11:32 |
| pavolzetor | yes, I tested it | 11:32 |
| pavolzetor | but the bug is | 11:32 |
| pavolzetor | now, if I do not use it | 11:32 |
| pavolzetor | window show as maximized | 11:33 |
| pavolzetor | but bottom is not drawn | 11:33 |
| JanC | How do you restore window size? | 11:33 |
| pavolzetor | set_Default_size | 11:33 |
| JanC | I think that's wrong in case of maximized windows anyway, as the screen size might be different on next run? | 11:35 |
| JanC | e.g. external monitor or projector connected to laptop | 11:35 |
| pavolzetor | yo uare right | 11:36 |
| pavolzetor | I am going to change it | 11:36 |
| pavolzetor | but it is still bug ;) | 11:36 |
| JanC | there might still be a bug in either gtk or compiz | 11:37 |
| pavolzetor | thanks anyway for help | 11:40 |
| pavolzetor | and sorry about rants, I just thought you do not really know, how developers feel about platform | 11:41 |
| pavolzetor | and I am not even developer | 11:41 |
| pavolzetor | so they probably expect more | 11:41 |
| dpm | pavolzetor, we do talk to developers all of the time, and we share both the pain in some cases, but also the successes :) Your feedback is very useful, so feel free to hang around here, on the mailing list or in any of the other channels for discussion! | 11:42 |
| pavolzetor | I hang a lot at clutter IRC | 11:44 |
| pavolzetor | and I can hang here | 11:44 |
| pavolzetor | as Gtk4 should work much better with clutter | 11:44 |
| JanC | I added what we found to the bug report | 11:44 |
| dpm | thanks JanC | 11:45 |
| pavolzetor | thanks :) | 11:45 |
| dpm | pavolzetor, when it's finished, have you considered submitting your app to the Software Center? | 11:45 |
| pavolzetor | I do not know | 11:45 |
| dpm | you should try :) | 11:46 |
| pavolzetor | I am waiting for breaking point | 11:46 |
| pavolzetor | because geeks will probably do not like my app | 11:46 |
| pavolzetor | so I want to make it feel nicely for newbies (I still feel, that RSS is more comfy than facebook for news) | 11:46 |
| JanC | maybe best to fix most issues before putting it in SC, or your users will rant... :p | 11:47 |
| pavolzetor | I just want to make it beautiful | 11:47 |
| pavolzetor | I know :D | 11:47 |
| pavolzetor | I have some tests | 11:47 |
| dpm | pavolzetor, well, that's exactly what we're trying to do both for Ubuntu and for the app developer process: reach out to non-tech users | 11:47 |
| pavolzetor | especially on service side, I do not know how to test Gtk/clutter properly though | 11:47 |
| dpm | in any case, when you're ready, the Software Center is just a few steps away: https://developer.ubuntu.com/publish/ :-) | 11:48 |
| pavolzetor | thanks, I will have a look | 11:48 |
| dpm | excellent | 11:48 |
| JanC | there are some tools to test GUI apps | 11:48 |
| pavolzetor | basically I use unittest | 11:48 |
| pavolzetor | and I want to make service really good | 11:48 |
| pavolzetor | so anybody who writes widget for RSS can plug and use it | 11:49 |
| dpm | nice | 11:49 |
| JanC | service is in python too? | 11:51 |
| pavolzetor | yes | 11:51 |
| pavolzetor | I have started in Vala | 11:51 |
| pavolzetor | but I haven;'t seen benefits | 11:51 |
| pavolzetor | harder to maintain, and I have got python in school | 11:52 |
| pavolzetor | it is still in python2 though, because of u1db | 11:52 |
| JanC | one important thing with a service/daemon is to make sure you don't leak memory | 11:52 |
| * dpm loves Python | 11:52 | |
| pavolzetor | me too :D | 11:52 |
| JanC | and don't fragment it too much | 11:52 |
| pavolzetor | I see | 11:53 |
| pavolzetor | I have no idea how to do it in python | 11:53 |
| pavolzetor | but it should not fragment | 11:53 |
| JanC | especially for people like me that run desktops with 1 month uptime ;) | 11:53 |
| pavolzetor | you can test it then :) | 11:54 |
| pavolzetor | I have uptime 2 days | 11:54 |
| pavolzetor | and memory usage is same | 11:54 |
| pavolzetor | for service | 11:54 |
| pavolzetor | also for client | 11:54 |
| JanC | that's good then | 11:54 |
| pavolzetor | (which uses about 70 megs, but 50 megs are just mesa stuff) | 11:54 |
| pavolzetor | which is actually not real, just allocated | 11:54 |
| pavolzetor | anyway, I just wonder about unity and compiz | 11:55 |
| pavolzetor | is actaully compiz developer by somebody outside ubuntu? | 11:55 |
| JanC | compiz still leaks memory, although not as bad as a year ago | 11:56 |
| JanC | the main compiz dev works for Canonical | 11:56 |
| pavolzetor | I see | 11:57 |
| pavolzetor | because I have tried clutter last year | 11:57 |
| pavolzetor | and I use it now, it is pretty quick | 11:57 |
| JanC | https://launchpad.net/~compiz-team/+members#active -> at least 6 of them work for Canonical | 12:08 |
| pavolzetor | I see | 12:17 |
| pavolzetor | why has compiz been chosen? | 12:17 |
| pavolzetor | I struggle with getting right state | 12:31 |
| pavolzetor | do you know how to get it reliably | 12:31 |
| pavolzetor | because snappy windows does not emit configure event | 12:32 |
| dpm | pavolzetor, I believe at the point when the decision was made to use compiz, clutter was first tested and was found to have performance issues. But I'm just an onlooker in that topic, so if you're interested someone on the #desktop team or on the Unity channels should be able to give you a proper explanation | 12:32 |
| pavolzetor | thanks, I asked | 12:34 |
| JanC | not sure if clutter had performance issues, but mutter did | 12:36 |
| pavolzetor | I see | 12:37 |
| JanC | mutter = metacity + clutter + javascript | 12:37 |
| pavolzetor | but was it year ago or now? | 12:37 |
| JanC | more than a year ago | 12:37 |
| JanC | the first version of unity was on mutter | 12:38 |
| pavolzetor | it has probably change | 12:38 |
| pavolzetor | d | 12:38 |
| JanC | right, but porting current unity to mutter and make it stable again would probably take 6-12 months with no new features etc. | 12:39 |
| pavolzetor | I see | 12:42 |
| pavolzetor | so it is easier to maintain compiz too | 12:43 |
| JanC | maybe | 12:43 |
| JanC | not my task to decide what's better long-term | 12:43 |
| pavolzetor | thanks for explanation | 12:44 |
| JanC | the future could also be a new compositor on Wayland | 12:47 |
| JanC | and if so it would make even less sense to switch to mutter now and then to Wayland a year later or so | 12:47 |
| pavolzetor | I see | 12:49 |
| pavolzetor | I am looking forward | 12:49 |
| pavolzetor | but they should decrease memory usage | 12:49 |
| JanC | lots of applications/frameworks should decrease memory usage... :-/ | 13:17 |
| pavolzetor | my app uses about 80 megs | 13:18 |
| pavolzetor | service 15 | 13:18 |
| pavolzetor | client about 65 | 13:18 |
| pavolzetor | but empty window with clutter stage uses about 50 megs | 13:18 |
| pavolzetor | so my app uses together 30 megs, which is not bad I think | 13:19 |
| pavolzetor | if I look at other apps, it is quite good | 13:20 |
| pavolzetor | hi, I have other quesiton | 15:37 |
| pavolzetor | if I enter video inside USC | 15:37 |
| pavolzetor | can it be youtube stufF? | 15:37 |
| pavolzetor | with youtube music? | 15:37 |
| pavolzetor | this is just example | 15:40 |
| pavolzetor | http://www.youtube.com/watch?v=UaVhlYzIU-s | 15:40 |
| pavolzetor | music is from youtube directly | 15:40 |
| pavolzetor | I just wonder for future release in USSC | 15:41 |
| pavolzetor | USC | 15:41 |
| zoopster | pavolzetor: if you are talking about submissions through the myapps portal into the USC - currently limited to vimeo, youtube is coming | 16:07 |
| pavolzetor | yes | 16:08 |
| pavolzetor | I see | 16:08 |
| pavolzetor | okay | 16:08 |
| pavolzetor | I am quite far from 1.0 | 16:08 |
| pavolzetor | but I wanted to check it | 16:08 |
| pavolzetor | thanks | 16:08 |
| zoopster | sure | 16:08 |
| JanC | pavolzetor: you can always look at Jamendo for music | 18:23 |
| pavolzetor | I see, but it is quite hard to find something lovely | 18:45 |
| pavolzetor | special | 18:45 |
| pavolzetor | what describes your product | 18:45 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!