quequotion | hello? | 06:05 |
---|---|---|
nhaines | quequotion: you have sucessfully joined #ubuntu-unity. | 06:05 |
quequotion | I've come to ask about lp bug 1025535 | 06:06 |
ubot5 | Launchpad bug 1025535 in Unity "Unity 6.0 on precise: Unity panel visible over full-screen applications" [Undecided,New] https://launchpad.net/bugs/1025535 | 06:06 |
duflu | Hi quequotion | 06:07 |
quequotion | hello! | 06:07 |
quequotion | I think the root of my problem is a rendering bug that affects the cube. | 06:09 |
quequotion | I'm not sure if it's a rendering bug in unity or compiz, but it was suggested that it's in unity. | 06:10 |
quequotion | If seeing were believing, I'd say that all windows are rendered on a side of the cube--fullscreen or otherwise--but unity is not. Unity is rendered separately, layered over the cube. | 06:12 |
nhaines | quequotion: I'll say one thing, you're not going to get anyone to make the panel autohide. :) | 06:14 |
quequotion | I am well aware :) | 06:14 |
nhaines | Is there a particular reason you're running Unity 6.0 on precise? | 06:15 |
duflu | quequotion: That's right. Unity (the Unity Shell) is a compiz plugin that is not affected by screen transformations. So it will stay in place even when the cube is active | 06:16 |
quequotion | Unity 5.16 was not available at the time, and it should have fixed this bug. My bug report is a duplicate of bug 734908 (sort of). | 06:16 |
duflu | I think it's mainly designed that way so you can still use it (start apps from the launcher etc) while in expo mode (Super+S) | 06:16 |
ubot5 | Launchpad bug 734908 in Compiz "Unity is visible on top of fullscreen apps" [High,In progress] https://launchpad.net/bugs/734908 | 06:16 |
quequotion | >>duflu that's good to know. Is this the reason why I have unity on top of everything (full-screen windows, screensavers, etc)? | 06:17 |
duflu | quequotion: Still don't know. But everyone else using the same Unity packages in Ubuntu 12.10 alpha seems to have no such problems | 06:18 |
duflu | And the same logic backported to 5.16 is looking good. Just a little work required in Nux and Compiz for that future precise update. | 06:19 |
duflu | But right now, more important things are happening... http://mars.jpl.nasa.gov/msl/multimedia/nasatv/ | 06:19 |
quequotion | the cube has been disabled by default, and I no one replied when I asked if they'd tested against the cube. | 06:20 |
quequotion | at least, not yet. | 06:20 |
quequotion | i wonder what the average IQ is on the stage... | 06:25 |
MCR1 | duflu, quequotion: Haha, I just wondered why the panel rendering on the cube is fixed here - then I remembered that I am running the Unity-2d-panel @ the moment :-D | 06:31 |
quequotion | 2d+3d=? | 06:32 |
MCR1 | quequotion: Unity-2d + Compiz, yes | 06:33 |
quequotion | Unity 2.5d? | 06:36 |
quequotion | It seems like unity 2d has been working well for a while | 06:37 |
quequotion | Someone mentioned it in a thread I started on the forum | 06:37 |
quequotion | I have to get back to work, but I'll be back later. Any help is appreciated, please leave a comment on the bug if you can. | 06:39 |
MCR1 | duflu: Got a minute ? Hi, btw :) | 06:50 |
duflu | MCR1: Yes? | 06:50 |
MCR1 | duflu: I need some advice how to cope with the remaining problems cppcheck detected: | 06:51 |
MCR1 | duflu: 1. We have a lot of scanf without field width limits can crash with huge input data. To fix this error message add a field width specifier warnings left. | 06:51 |
duflu | MCR1: Yes, field widths are good | 06:52 |
MCR1 | duflu: ok, I just read also that the use of scanf is generally considered dangerous and is not recommended in release versions... | 06:53 |
duflu | MCR1: That's probably overcautious. You can use scanf safely if you're careful | 06:54 |
MCR1 | duflu: ok, I'll try to fix those then with field width limits :) | 06:55 |
MCR1 | duflu: 2. We have a "suspicious calculation" in /plugins/animationaddon/src/polygon.cpp:459 and 2 "suspicious conditions" (assignment + comparison) in /plugins/colorfilter/src/parser.cpp:177 that can be clarified with parentheses | 06:57 |
MCR1 | duflu: Also a "Suspicious expression" in /plugins/place/src/place.cpp:439 - Boolean result is used in bitwise operation. The ! operator and the comparison operators have higher precedence than bitwise operators. It is recommended that the expression is clarified with parentheses. | 06:59 |
duflu | MCR1: They sound like things worth cleaning up. But FYI, animationaddon and colorfilter are not being ported to GLES (compiz 0.9.8.0) right now | 07:00 |
MCR1 | duflu: Is GLES the only future for Compiz ? | 07:01 |
duflu | MCR1: Yes it is the only requirement to reach 0.9.8.0 right now | 07:01 |
MCR1 | duflu: I mean, will Compiz on desktop machines run on OpenGL ES as well in the future ? | 07:02 |
duflu | MCR1: Sorry, that was confusing | 07:02 |
duflu | The branch is called "gles2", but it uses plain OpenGL on desktop. And uses OpenGLES if compiled for ARM etc | 07:02 |
MCR1 | duflu: Then we should IMHO port all the plug-ins to be compatible with GLES sooner or later... AFAIK it is easy to turn compiling of problematic ones off in that branch. | 07:06 |
MCR1 | duflu: I do not want to see a "crippled" Compiz with reduced functionality - it would be a bad compromise | 07:06 |
duflu | MCR1: Yes, we have ported 60-70 out of 80 already. But they won't all be done at release | 07:07 |
duflu | Community contributions welcome | 07:07 |
MCR1 | duflu: Sure, I try my best already ;) | 07:07 |
duflu | Only the plugins used by the default Unity/ubuntu profile are a priority right now | 07:07 |
MCR1 | Sure | 07:08 |
duflu | And Ubuntu/Unity only uses 23 of the 80 plugins | 07:08 |
* MCR1 already noticed that. | 07:08 | |
MCR1 | duflu: I just can't live without all the great animations ;) | 07:09 |
duflu | MCR1: I agreed with that when the project was first proposed. But there are just no resources or justification to port everything right now | 07:10 |
MCR1 | duflu: Ofc, I understand that. | 07:10 |
MCR1 | duflu: For me main priority is maximum stability, Compiz got to be rock-stable. 2nd priority is to not have to sacrifice looks or functionality in modern Compiz. | 07:12 |
MCR1 | duflu: Once all the suspicious stuff is cleaned up, I want to target the bugs nerving myself most. | 07:13 |
MCR1 | duflu: The strange thing is that some bugs are just occuring if Compiz+Unity are running together, Compiz standalone does not show many of them... | 07:14 |
duflu | MCR1: Makes sense. Wanting compiz to be better was what got me started | 07:15 |
MCR1 | duflu: I am a C++ newbie, but I coded before and am a fast learner, so I hope I'll be able to target harder stuff soonish | 07:15 |
MCR1 | duflu: Do you know the reason to have a shortcut::HARDCODED_OPTION ? I mean Compiz handles shortcuts perfectly (IMO), so shouldn't those be replaced by shortcut::COMPIZ_KEY_OPTION ? | 07:37 |
MCR1 | sil2100: Hi :) Tried e4rat already ? | 07:38 |
duflu | MCR1: Where is shortcut::HARDCODED_OPTION? Not in the compiz code... | 07:41 |
MCR1 | duflu: No - and probably that is the main problem. | 07:42 |
duflu | Sorry, what? | 07:42 |
sil2100 | MCR1: hello :) No, not yet - sadly had completely no time during the weekend :( | 07:43 |
MCR1 | duflu: As you know there are a lot of bugs reported regarding Unity/Compiz bindings problems - I found out that the main problem is that Compiz is not aware of Unity shortcuts | 07:43 |
MCR1 | sil2100: Highly recommended - it will probably cut your boot time in half | 07:44 |
duflu | MCR1: Yes, it's a Unity design problem. Unity intercepts key events from compiz, which was probably a bad idea. | 07:44 |
MCR1 | duflu: We should change that and Compiz should be in control of all the shortcuts | 07:45 |
duflu | MCR1: Agreed. Please look in lp:unity for all that | 07:45 |
MCR1 | duflu: It would eliminate multiple problems at the same time - yes I'll take care of it (once I finished the cppcheck stuff) | 07:46 |
=== deegee__ is now known as drussell | ||
Andy80 | hi all | 08:34 |
Andy80 | I've one proposal for the upcoming liv events. For example this evening we will have the "Ubuntu Accomplishments" live class. There is nothing about it on G+ Ubuntu profiles and there is no event created. Why don't we try to keep the G+ profile more updated and we create events for these lives? So people who has Ubuntu in their circles would automatically get the invitation and would not miss it. What do you think about? | 08:36 |
=== JanC_ is now known as JanC | ||
didrocks | sil2100: hey, how are you? | 10:22 |
sil2100 | didrocks: hello, good thanks, how are you? | 10:24 |
didrocks | sil2100: I'm fine, thanks | 10:27 |
didrocks | sil2100: is the release in progress? ;) | 10:27 |
sil2100 | didrocks: yes ;) Testing on the way | 10:28 |
didrocks | sweet \o/ | 10:29 |
tsdgeos | sil2100: speakin of release, these bugs never made it into a release yet? https://bugs.launchpad.net/~aacid/+assignedbugs | 10:46 |
sil2100 | tsdgeos: good point, there was no unity-2d release in like, forever | 10:47 |
sil2100 | Will try to release one this week maybe? | 10:47 |
tsdgeos | up to you (or someone else) | 10:47 |
tsdgeos | i haven't being doing unity-2d in a while | 10:47 |
tsdgeos | just seems a "waste" to have the fixes there lying around | 10:47 |
tsdgeos | but not really in users hands | 10:48 |
sil2100 | True, since most of the bugs are even prepared for SRU and everything | 10:48 |
sil2100 | didrocks, seb128: what do you think? Could I prepare an unity-2d SRU after unity quantal this week? | 10:48 |
seb128 | sil2100, works for me, it will probably not land before a few weeks since precise is frozen for LTS .1 at the moment, so you have time | 10:56 |
sil2100 | Thanks! | 10:57 |
didrocks | sil2100: looks good to me :) | 11:13 |
didrocks | (shower time after exercising now) | 11:13 |
sil2100 | It's so hot today that I don't want to even move outside ;) So I envy you didrocks | 11:20 |
didrocks | sil2100: contrary here for the past 3 days, more rainy/cloudy than anything, but the sun should come back tomorrow! | 11:23 |
=== _salem is now known as salem_ | ||
=== yofel_ is now known as yofel | ||
Andy80 | I try to repeat my question now that there should be more people: I've one proposal for the upcoming live events. For example this evening we will have the "Ubuntu Accomplishments" live class. There is nothing about it on G+ Ubuntu profiles and there is no event created. Why don't we try to keep the G+ profile more updated and we create events for these lives? So people who has Ubuntu in their circles would automatically get the invitation and | 12:53 |
Andy80 | would not miss it. What do you think about? | 12:53 |
sil2100 | I don't really use G+, but this sounds like a rather good idea | 12:56 |
quequotion | helllo | 13:10 |
quequotion | I'm looking for help with bug 1025535 | 13:14 |
ubot5 | Launchpad bug 1025535 in Unity "Unity 6.0 on precise: Unity panel visible over full-screen applications" [Undecided,New] https://launchpad.net/bugs/1025535 | 13:14 |
c10ud | quequotion, don't use unity6 in precise, i tried, i succeeded but from some specific revision on it won't work anymore | 13:18 |
quequotion | i was here some hours ago as well, and the discussion had gotten as far as this: unity is rendered independently of screen transformations, and therefore is visible while the cube is rotating or expo is enabled, etc; it is not known if this is related to my bug. No one else seems to have the same problem I do, but then no one has confirmed testing with unity-3d and compiz cube other than myself. | 13:18 |
c10ud | just remove all ppas and revert unity/nux/compiz to precise's stock versions | 13:19 |
quequotion | >>c10ud you have the backported repository yes? I haven't tried your packages yet. Are they built differently or are they the same as the Quantal versions? | 13:19 |
c10ud | quequotion, i deleted them as soon as i saw your bugreport mentioning my ppa :) they were just nux/unity from trunk built with precise packaging | 13:20 |
c10ud | (a few minor adjustments required) | 13:20 |
c10ud | i deleted them since they can break your os and make you lose time removing ppas and reverting packages | 13:21 |
quequotion | I could revert the packages, but I installed them for the purpose of fixing this bug. Since the updated packages for precise are not ready yet (?) I'd come right back to where I started. | 13:21 |
c10ud | did you try unity 5.14? | 13:21 |
c10ud | it's in precise-proposed iirc | 13:22 |
c10ud | also: if you revert compiz/nux/unity to precise-proposed ones i think there's more chance your bug will be fixed | 13:22 |
quequotion | I could look through my dpkg logs, but I believe I did try .14. The strangest thing about my problem is this: although no one seems to have my issue, I have never seen a different behavior since the inception of unity. | 13:24 |
c10ud | if you go back to a stock version there's more chance your issue can be reproduced --- check also compiz, try to revert to stock profile (i believe unity --reset) | 13:26 |
c10ud | maybe you messed up stuff with ccsm | 13:26 |
quequotion | I've been over that with Daniel van Vugt, and I know that my particular case is unsupported. I've had the same issue with all previous versions of unity, so I'm hoping to look at the issue from another angle. | 13:26 |
c10ud | quequotion, you also have some weird compiz | 13:27 |
quequotion | I have reverted, purged, and reinstalled numerous times. I even tried to search the file system to hunt down strays that dpkg didn't remove. | 13:28 |
quequotion | i have quantal's compiz, although perhaps a few revisions out of date now (the packages are installed manually, not from the repository) | 13:28 |
c10ud | i used synaptic to revert everything back to stock | 13:29 |
c10ud | but if i'm not wrong, now that i think of it i also had your issue....somewhere in time | 13:29 |
c10ud | when i was running bleeding edge stuff | 13:29 |
quequotion | that's a painful and risky way of reverting installs; i recommend using package pins whenever possible. | 13:29 |
c10ud | quequotion, also which nvidia drivers you're using? are those NVIDIA's betas? | 13:30 |
c10ud | i remember having unity showing weird glitches with those, so i reverted back to 302.17 | 13:30 |
quequotion | The issue was more common in the past, and at some point (pre unity 5.x i think) also affected Desktop Wall, and therefore many more users. | 13:31 |
c10ud | (i have a gt240, so not really different from yours) | 13:31 |
c10ud | quequotion, if you can tell me a testcase i can try and see if it works | 13:32 |
quequotion | I've been through several driver versions, with the same results from all of them. I've tried the stock, proposed, and updates versions. At the moment I've got 304.32 | 13:32 |
* didrocks remembers when c10ud didn't trust him when I was telling that trying unity 6 on precise was risky and will be incompatible :) | 13:34 | |
quequotion | The basic test case is this: Use the unity-3d desktop; enable the Desktop Cube (which requires disabling the Desktop Wall); run any application that is capable of full-screen; observe that unity is displayed over the full-screen window and that input is correctly directed to the full-screen window (clicks pass through unity) | 13:35 |
c10ud | quequotion, how to enable the desktop cube? | 13:35 |
quequotion | >>didrocks except for this one graphical annoyance it's been really stable and performs much better than 5.x > 5.16 | 13:35 |
quequotion | CCSM | 13:36 |
c10ud | didrocks, different thing, the guy here is running xorg-edgers with nvidia-beta along with some weird versions (duh!) :p you broke compatibility *on purpose* in order to use some c++11 weird stuff ;) | 13:36 |
didrocks | quequotion: right, but the code changes we didn't backport are risky and you just experienced one corner cases when it indeed regressed with the unsupported stack, I'm sure they are tons of them | 13:36 |
didrocks | c10ud: hum, we didn't use newer c++11 "stuff" | 13:37 |
didrocks | compared to precise | 13:37 |
c10ud | boost, sorry | 13:37 |
c10ud | i've been on vacation | 13:37 |
c10ud | :P | 13:37 |
didrocks | right, it's not on purpose though, it's the platform moving | 13:37 |
c10ud | in the end, if unity was a nice monolithic build building it for different platform wouldn't have been so difficult | 13:37 |
c10ud | but you have bamf, nux, libunitywat(?) etc. | 13:38 |
didrocks | c10ud: well, what's the point of having it running on platform we don't support? | 13:38 |
c10ud | then it's better to stay with canonical | 13:38 |
c10ud | nothing, we're hackers right? ;) | 13:38 |
quequotion | When you enable the Desktop Cube, ccsm will ask you to disable Desktop Wall. Since I'm-not-really-sure-how-long-ago, this no longer requires disabling unity and transitions rather smoothly. | 13:38 |
didrocks | c10ud: heh, yeah, if only there were already less bugs in what we support… ;) | 13:38 |
quequotion | I installed unity 6.0 and all it's dependencies from quantal (compiz, nux, bamf, libdee, etc) | 13:39 |
c10ud | i reckon lp is not that good for dismissing bugs as "not supported, won't fix" (at least, it's not comfortable for me to use :p) | 13:39 |
c10ud | eheh | 13:39 |
c10ud | quequotion, i'll now try what you ask | 13:39 |
c10ud | quequotion, now desktop cube is enabled | 13:40 |
c10ud | do i have to press something? | 13:40 |
quequotion | did compiz restart smoothly? | 13:40 |
quequotion | you need to open an application that full-screens | 13:41 |
c10ud | i didn't restart compiz, is that needed? | 13:41 |
quequotion | ie firefox (F11) | 13:41 |
quequotion | compiz should restart itself when you enable the cube | 13:41 |
quequotion | a video in totem would also be a good test | 13:41 |
quequotion | or manually starting the screensaver | 13:41 |
c10ud | quequotion, is there some key combination i can try so i see the cube is running? | 13:41 |
c10ud | then i'll try firefox | 13:42 |
quequotion | that's a good question... I don't know if there's a way to explicitly verify that specific compiz plugins are loaded/active, but I usually test the cube by switching viewports. | 13:43 |
c10ud | i am sorry but i don't know how to do that (duh!) i'm trying to search in ccsm | 13:44 |
c10ud | but if you tell me the defaults.. ;p | 13:44 |
c10ud | looks like i am the only one that *never* tried compiz' cube | 13:45 |
quequotion | well, it's going to dirty this test a little, but i don't know another way than enabling the Rotate Cube extension. then you can use CTRL+ALT+Left/Right to rotate viewports left or right. | 13:47 |
c10ud | quequotion, ok i see something is rotating | 13:48 |
quequotion | note that your desktop may not actually be very cubish, but two flat panels that will rotate (like a coin) this should be irrelevant to the bug. | 13:49 |
c10ud | yes exactly | 13:49 |
c10ud | quequotion, f11 firefox doesn't show the bug you reported, i'll now try with a full screen video | 13:50 |
c10ud | fullscreen totem, no artifacts or anything | 13:50 |
quequotion | no unity panel whatsoever in fullscreen? | 13:50 |
c10ud | nope | 13:51 |
c10ud | looks like didrocks wins, heh | 13:52 |
c10ud | but you really have too much bleeding edge stuff | 13:52 |
quequotion | I know this, but even without all the bleeding edge stuff I was having the same behavior. | 13:53 |
c10ud | quequotion, probably it was fixed with 5.14 | 13:53 |
c10ud | i just have precise-proposed/updates, etc. + https://launchpad.net/~ubuntu-x-swat/+archive/x-updates/ | 13:54 |
c10ud | xorg edgers is just too edgy | 13:54 |
c10ud | i don't know what you need from edgers or unity6, i needed performance improvements, but now that i convinced (?) the guys here to include it in 5.16 i'm good with stock+xswat | 13:55 |
quequotion | I needed unity 6 for this bug and xorg-edgers for improvements in the nvidia drivers. | 13:56 |
c10ud | the nvidia betas usually don't show improvements unless you mean bugs (here it was missing gl calls (!)) | 13:57 |
quequotion | yes, bug fixes. | 13:57 |
c10ud | anyway, i did this test for you, now you decide | 13:57 |
quequotion | thank you. i guess it means another full purge, reversion and reinstall. having done this before, I'm curious what I should do if it does not work? | 13:58 |
c10ud | file a bug! :D | 13:59 |
quequotion | lol | 13:59 |
quequotion | launchpad has probably had quite enough of my ranting for now. | 13:59 |
c10ud | double check with synaptic for bad packages (i saw some compiz ones survived the downgrade) | 13:59 |
quequotion | oh? | 14:00 |
c10ud | dunno if they were used or anything, they just were there (i'm talking about my experiments) | 14:00 |
quequotion | ah | 14:00 |
quequotion | i thought you meant from my dpkg log in the bug report. | 14:01 |
c10ud | nope, i didn't read that, i just saw nvidia beta, compiz trunk and unity6...that was enough :P | 14:01 |
quequotion | i used apt-pinning to downgrade before. the documentation is spotty to say the least, but it's a fairly sound method | 14:01 |
c10ud | i'd just remove ppas and reinstall stuff | 14:02 |
quequotion | check later in the bug report. I went all the way back to stock, baring the packages directly related to the bug report (unity and dependencies) | 14:02 |
c10ud | but i'm not really a dpkg xpert | 14:02 |
quequotion | i doubt that most anyone is, since the documentation is out-of-date and incomplete. | 14:03 |
quequotion | i suppose I'll get to work cooking up another batch of /etc/apt/preferences.d/downgrades | 14:05 |
c10ud | quequotion, if you already tried downgrading unity and friends i wouldn't trust xorgedgers and nvidia | 14:09 |
quequotion | i previously downgraded xorgedgers and other ppas. | 14:23 |
quequotion | I have also downgraded unity, but I'll give it another try and make sure everything is back to stock + proposed this time. | 14:24 |
sil2100 | Trevinho, andyrock: did you guys try running quantal unity-team staging? | 15:10 |
Trevinho | sil2100: no, sorry... | 15:11 |
sil2100 | Trevinho, andyrock: is it only me, or is the launcher strangely broken? | 15:11 |
sil2100 | Maybe related to bamf | 15:11 |
sil2100 | Since when I open up any application, it doesn't appear on the launcher | 15:11 |
sil2100 | Icons when pressed blink much much longer | 15:12 |
Trevinho | sil2100: is your bamfdaemon running? | 15:12 |
sil2100 | And the arrows only update 'once in a while' | 15:12 |
Trevinho | mh, so... well, yes it seems something related to libbamf / bamfdaemon | 15:12 |
sil2100 | Trevinho: but I noticed, that when I switch workspaces, then the icons get updated (arrows and icons) | 15:12 |
Trevinho | sil2100: mhmhm... well yes, it could be signal issue... maybe similar to the one that u-p-s had with the new dbus | 15:13 |
didrocks | Trevinho: shouldn't be the same dbus eavesdrop issue. Otherwise, we would have it in quantal | 15:15 |
didrocks | (and we don't) | 15:15 |
Trevinho | didrocks: ah, right... so it's not :) | 15:15 |
sil2100 | Trevinho: do you have a quantal machine somewhere? | 15:15 |
Trevinho | sil2100: I need to prepare one, but I won't be able to do it until Friday... | 15:16 |
sil2100 | Trevinho: but could you somehow try working on this? Since, if this is confirmed on other systems, it essentially blocks the release | 15:22 |
Trevinho | sil2100: ok, I'll give that a try | 15:23 |
sil2100 | Trevinho: thanks! | 15:25 |
quequotion | pins in place, beginning downgrade | 15:36 |
=== dandrader is now known as dandrader|afk | ||
=== dandrader|afk is now known as dandrader | ||
quequotion | I'm back, with unity 5.14.0, all packages downgraded to stock + proposed + x-swat-x-updates (as advised).. no change. unity panel still visible across the top in all full-screen applications... | 16:53 |
=== dandrader is now known as dandrader|lunch | ||
c10ud | quequotion, purge the beta driver | 16:59 |
quequotion | i did | 17:28 |
quequotion | i could go one more step, and purge x-swat-x-updates, but you weren't having any trouble with this ppa. | 17:29 |
=== dandrader|lunch is now known as dandrader | ||
=== _morphis is now known as morphis | ||
=== Cantide is now known as CanStudy | ||
Kyshtynbai | Hi guys! Are there any ways to make compiz scale plugin to show all windows in all desktops? | 19:28 |
=== CanStudy is now known as Cantide | ||
=== dandrader is now known as dandrader|afk | ||
johnjohn101 | when will unity support the old compiz snow plugin? | 20:59 |
=== morphis is now known as morphis|away | ||
=== dandrader|afk is now known as dandrader | ||
=== salem_ is now known as _salem |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!