Kubuntu 9.04 alpha6 panel corruption

[dotancohen]

Binary package hint: libplasma3

Hardware: Dell Inspiron E1505 with ATI video card:
01:00.0 VGA compatible controller: ATI Technologies Inc Radeon Mobility X1400

$ lsb_release -rd
Description: Ubuntu jaunty (development branch)
Release: 9.04

$ kde4-config -v
Qt: 4.5.0
KDE: 4.2.1 (KDE 4.2.1)
kde4-config: 1.0

$ apt-cache policy libplasma3
libplasma3:
  Installed: 4:4.2.1a-0ubuntu3
  Candidate: 4:4.2.1a-0ubuntu3
  Version table:
 *** 4:4.2.1a-0ubuntu3 0
        500 http://us.archive.ubuntu.com jaunty/main Packages
        100 /var/lib/dpkg/status

I am experiencing much panel corruption, see attached screenshot. I should mention that I did not touch the keyboard or mouse in the timeframe immediately preceding the onset of corruption. Also, the system remains usable, only the taskbar and desktop seem to be affected. The desktop is affected by the fact that the temp and CPU monitors have become transparent, and the remaining widgets (desktop folder, fuzzy clock, composting effects toggle) have disappeared.

I am using the panel widgets: Kmenu (the standard one), network manager, system tray, analog clock, task bar

(I am only guessing that the bug is in plasma, if not, then please reassign. Thanks)

[dotancohen (dotancohen)]

In what may or may not be a related issue, as soon as I posted the panel went blank and Kmenu is just a shadow. See attached screenshot. The black bar at the bottom is my panel, and it can be seen that it is blank. If you look closely at the lower right, you can see a rectangular shadow. That should be the Kmenu. Kolourpaint is segfaulting on me right now, so I cannot edit the screenshot to point it out better, sorry.

$ kolourpaint
Starting KolourPaint on a 24-bit screen...
Segmentation fault (core dumped)

[dotancohen (dotancohen)]

As kolourpaint is segfaulting, I looked in .xsession-errors and it is full of these:

$ tail .xsession-errors
QPainter::end: Painter not active, aborted
QPainter::begin: Paint device returned engine == 0, type: 3
QPainter::setCompositionMode: Painter not active
QPainter::end: Painter not active, aborted
QPainter::begin: Paint device returned engine == 0, type: 3
QPainter::setCompositionMode: Painter not active
QPainter::end: Painter not active, aborted
QPainter::begin: Paint device returned engine == 0, type: 3
QPainter::setCompositionMode: Painter not active
QPainter::end: Painter not active, aborted

I attach dmesg for good measure.

[Jonathan Thomas (echidnaman)]

This is most likely a graphics driver issue. Which drivers are you using? (fglrx or the "free" ones?)
The kolourpaint issue in unrelated.

[dotancohen (dotancohen)]

This is with the drivers as installed by a stock Kubuntu install. I have not touched the video drivers. I therefore assume "free".

How can I check to be sure which video drivers are in use? I checked xorg.conf but it is nearly empty!

[Jonathan Thomas (echidnaman)]

The output of lspci -vvnn (two v's, not w) is handy in this regard.

[dotancohen (dotancohen)]

Maybe it is the Radeon video driver:
jaunty@ubuntu:~$ lsmod | grep radeon
radeon 148512 3
drm 96424 4 radeon

How does one determine which video driver is in use on the new systems without xorg.conf?

[dotancohen (dotancohen)]

Thanks, but I still do not see where the driver is listed:

01:00.0 VGA compatible controller [0300]: ATI Technologies Inc Radeon Mobility X1400 [1002:7145]
 Subsystem: Dell Device [1028:2003]
 Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
 Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
 Latency: 0, Cache Line Size: 64 bytes
 Interrupt: pin A routed to IRQ 16
 Region 0: Memory at d0000000 (32-bit, prefetchable) [size=256M]
 Region 1: I/O ports at ee00 [size=256]
 Region 2: Memory at efdf0000 (32-bit, non-prefetchable) [size=64K]
 [virtual] Expansion ROM at efd00000 [disabled] [size=128K]
 Capabilities: <access denied>

The whole output is attached.

[dotancohen (dotancohen)]

I see that I forgot to add the screenshot in the original bug! Here it is.

Furthermore, the blanking of the panel (as mentioned in the first comment) happens even when KDE composting is turned off.

[Jonathan Thomas (echidnaman)]

OK, I think /var/log/Xorg.0.log should have driver infos.

[Jonathan Thomas (echidnaman)]

Yup, it's radeon.

[StephenKing (steffen-steffen-gebert)]

Have the same problem with nvidia!

The problem comes ~1 time per day and sometimes goes away after some time (minutes/hours).

Same versions as dotancohen, amd64. Updated from intrepid some days ago. I'm not 100% sure, but I think I have the nvidia drivers from ubuntu repository (had some problems with envyng).

If you need more info, just tell me (but I'm away till Tuesday, so you would have to wait a bit).

$ lspci -vvnn
03:00.0 VGA compatible controller [0300]: nVidia Corporation G94 [GeForce 9600 GT] [10de:0622] (rev a1)
        Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
        Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
        Latency: 0
        Interrupt: pin A routed to IRQ 16
        Region 0: Memory at fd000000 (32-bit, non-prefetchable) [size=16M]
        Region 1: Memory at d0000000 (64-bit, prefetchable) [size=256M]
        Region 3: Memory at fa000000 (64-bit, non-prefetchable) [size=32M]
        Region 5: I/O ports at dc00 [size=128]
        [virtual] Expansion ROM at fea80000 [disabled] [size=512K]
        Capabilities: <access denied>
        Kernel driver in use: nvidia
        Kernel modules: nvidia, nvidiafb

[StephenKing (steffen-steffen-gebert)]

Attached my /var/log/Xorg.0.log

Steffen

[lcampagn (luke-campagnola)]

I've been seeing similar issues on my machine. I have kubuntu intrepid with KDE4.2 via intrepid-backports.
I have attached an example image of my CPU monitor (plasma-widget-cpuload) corrupted (the mouse image which appears in the corruption came from the twitter plasmoid). Sometimes the other monitor widgets are corrupted, sometimes the panel at the bottom of the screen. Sometimes the corruption clears up after a few seconds, sometimes it stays around, and sometimes it immediately precedes a plasma crash (I haven't caught a crash backtrace yet). My machine is a Thinkpad X60 tablet, using the intel video driver.

My .xsession-errors is filling up very quickly with these (several lines per second):
QPainter::begin: Cannot paint on a null pixmap
QPainter::end: Painter not active, aborted
QPainter::begin: Cannot paint on a null pixmap
QPainter::end: Painter not active, aborted
QPainter::begin: Painter already active
QPainter::setCompositionMode: Painter not active
QPainter::setFont: Painter not active
QPainter::end: Painter not active, aborted

I have a suspicion that there are a few competing bugs in this mess, and that only some of them overlap with the original report..

System info:
$ lsb_release -rd
Description: Ubuntu 8.10
Release: 8.10

$ kde4-config -v
Qt: 4.4.3
KDE: 4.2.2 (KDE 4.2.2)
kde4-config: 1.0

$ apt-cache policy libplasma3
libplasma3:
  Installed: 4:4.2.2-0ubuntu1~intrepid1
  Candidate: 4:4.2.2-0ubuntu1~intrepid1
  Version table:
 *** 4:4.2.2-0ubuntu1~intrepid1 0
        500 http://ppa.launchpad.net intrepid/main Packages
        100 /var/lib/dpkg/status
     4:4.2.0-0ubuntu2~intrepid2 0
        500 http://ubuntu.media.mit.edu intrepid-backports/main Packages

[RK (kubuntu-rk)]

I can confirm this bug, also using nvidia proprietary drivers. However, if I don't restart plasma as soon as the corruptions start to appear, I will eventually also lose text in console windows, i.e. text in a certain color will
simply not be printed on the screen (not even on redraw forced by marking and unmarking the area). So I would guess that this is a bug either in plasma or even in qt.

[RK (kubuntu-rk)]

Try to reassign to plasma.

[lcampagn (luke-campagnola)]

I have fixed these issues by upgrading to Jaunty and then following this
guide (intel specific):
http://ubuntuforums.org/showthread.php?t=1130582

There were so many problems with Jaunty before I followed the guide that I
can not say whether this particular bug was fixed by upgrading or by
following the guide (or both).

Q

On Mon, Apr 27, 2009 at 23:15, Bryce Harrington
<email address hidden>wrote:

> ** Changed in: xserver-xorg-video-ati (Ubuntu)
> Status: New => Confirmed
>
> --
> Kubuntu 9.04 alpha6 panel corruption
> https://bugs.launchpad.net/bugs/345076
> You received this bug notification because you are a direct subscriber
> of the bug.
>

[Smurphy (smurphy-linux)]

Same issue here - using Nvidia proprietary Drivers on a T61p with KUbuntu 4.2.2 and 4.2.3 in 64bit mode.
If deeper details are required, let me know.

[jimisdead (ryan-spacemonkey)]

This issue is happening on both my Kubuntu 9.04 machines.

both are running clean installs of Kubuntu 9.04 amd64, and the binary nvidia drivers.
one 9600GTS card with an intel quad core, and the other has a 8600m card and an intel dual core laptop processor.

killing plasma, then restarting it fixes the problem.

[Jonathan Thomas (echidnaman)]

Driver issue, not a kdelibs issue.

[dotancohen (dotancohen)]

If this is a driver issue than why are Nvidia folks confirming, while I (I am the OP) have an ATI card?

[Jonathan Thomas (echidnaman)]

The corruption the nvidia users could very well be different corruption. Corruption in and of itself is a quite broad term.

[StephenKing (steffen-steffen-gebert)]

Looking at dotancohen's screenshot I can confirm that it's exactly the same curruption with my both PCs using nVidia cards!
Everything is getting transparent after a few hours.

[lcampagn (luke-campagnola)]

I'd agree--the intel drivers have had numerous updates in the past few
months that seem to have solved most of my corruption issues.

On Fri, Jun 12, 2009 at 13:07, Jonathan Thomas <email address hidden>wrote:

> The corruption the nvidia users could very well be different corruption.
> Corruption in and of itself is a quite broad term.
>
> --
> Kubuntu 9.04 alpha6 panel corruption
> https://bugs.launchpad.net/bugs/345076
> You received this bug notification because you are a direct subscriber
> of the bug.
>

[RK (kubuntu-rk)]

Here you can see my corruptions. I'm using nvidia (180.53 with this shot, now at 185.something). note:
* the window list has almost disappeared, from 3-4 windows only one icon of konsole is left
* several icons have disappeared from kicker (K icon, show desktop icon) and from the tray (kde4lirc)
* trash icon disappeared
* temp display corrupted
* amarok 1.4 icon _not_ corrupted
* not visible: browser hightlighted icon corrupted
* not visible: kde3lirc icon also was never corrupted

Using KDE 4.2 with qt 4.5.0, the corruption happened preditably every few hours. With KDE 4.3b1 and 4.3b2 using qt 4.5.1, it only happens every 8-9 hours. The previous nvidia driver update changed occasional hard freezes that often went away after a few minutes by very rare total freezes, but did nothing on the corruption front.

So while some intel driver corruptions that were mentioned in the comments have been fixed by intel driver updates, the one reported in this bug still occur.

[RK (kubuntu-rk)]

Reopen.

[Jonathan Thomas (echidnaman)]

Your screenshot looks nothing like the first few reported against ati, reclosing. Different corruptoin, different bug.

[RK (kubuntu-rk)]

> Your screenshot looks nothing like the first few reported against ati, reclosing. Different corruptoin, different bug.

What again are you smoking? You see _EXACTLY_ the same corruption as from StephenKing in comment #12. Note in particular how the only remaining icon in the left screenshot is from Opera, which is also qt3. And notice how in the right screen shot there's _NOTHING_ left, _EXACTLY_ like in the screenshot from the original report. Also when comparing with the original, see that my temp and cpu displays have been corrupted (as seen also in StephenKing's screen shot and as mentioned in the original report). You also see the big "Desktop Folder" widget not existing in the middle. I also have a composite-on-off switch, that also disappears. So you see it is EXACTLY the same kind of corruption, why are you then closing it? And yes, I could have waited somewhat longer until the remaining stuff has disappeared as well...

-> Reopen. Please do not close until something has been actually fixed. And remember that the original reporter also still has this problem, so even if the bug would be in several drivers at the same time, it still would not be closed.

[Jonathan Thomas (echidnaman)]

StephenKing's corruption looks *nothing* like the screenshots posted by the original bug reporter.
And note that I did not close the entire bug. There is the xserver-xorg-video-ati task open and confirmed right under it. :/

[RK (kubuntu-rk)]

Can you please re-read my notes regarding my screen shot and compare it to the original report? I already spelled it out in my last comment. It is obvious this is not a driver issue, at least not only, because both ATI and nvidia are affected in the same way. Two more reasons:
a) whenever I move the mouse coursor over anything that has disappeared, I get this on stdout:
QPainter::begin: Paint device returned engine == 0, type: 3
QPainter::setCompositionMode: Painter not active
QPainter::end: Painter not active, aborted
b) the (classic) K menu stops working, first there are no sub-menus, later it doesn't show up at all anymore.

If you STILL don't believe me that my corruptions are the same as those of the original reporter, then maybe the original reporter could create a screenshot with all of those elements on it that more or less survived corruption for me, that is: a digital clock; a minimized konsole window; the device notifier applet; program icons; the KDE 4.3B weather applet; the logoff/shutdown KDE 4.3B applet; kmix in the system tray. On the desktop another weather applet; a file monitor; cpu / temp monitor. Then we can compare better.

[Jonathan Thomas (echidnaman)]

It still would not be the fault of anything in kdelibs, as it does not do low-level drawing.

[dotancohen (dotancohen)]

Actually, none of the panel elements survive corruption on my system. Not even the Kmenu itself is safe, nor the desktop plasmoids.

[StephenKing (steffen-steffen-gebert)]

The same corruption with transparent window also occurs on my PCs (although it's not in my screenshot).

If it's not kde4libs related, could we find out what package is responsible instead of just setting the bugreport to invalid?
I'm a bit disappointed how such really annoying bugs are handled and blocked here.

Steffen

[RK (kubuntu-rk)]

Well, fact is that my corruptions using an nvidia card definately do not come from a driver bug in the ATI driver.

@dotancohen: I guess I can't convinve you to upgrade to KDE4.3B2 and check with precisely the same set of widgets as I have?

@Jonathan Thomas: Don't you think it would be more productive if you could help locate the bug, instead of just closing it? Even if that ends up not being in kdelibs itself?

[dotancohen (dotancohen)]

Actually, with the KDE 4.2.4 PPA enabled the corruption went away. Now that I've moved to KDE 4.3b2 the corruption has not returned. I think that implies that this is a kde4libs issue indeed. I intend on checking a Karmic alpha or early beta before closing, however.

I will, however, gladly check with your configuration. Tell me exactly which plasmoids to configure.

[Jonathan Thomas (echidnaman)]

Beta2 would have any of the fixes that 4.2.4 would have.

[Jonathan Thomas (echidnaman)]

It could just maybe be possible that this would be a Qt issue.

[RK (kubuntu-rk)]

I still do not see any reason at all that would make this a driver issue. Quite frankly, I only see the opposite. So please stop closing the bug for kdelibs unless you have provided evidence in your favour. Thanks.

[RK (kubuntu-rk)]

@dotancohen: please see comment #30 on which widgets I'm currently using.

I'd also like to know which versions of KDE stuff you're using - in particular packages still belonging to 4.2.2 and 4.2.3. For me, this is:

$ dpkg -l | grep 4:4.2.2 | grep ^i | awk '{print $2}' | tr "\n" " "; echo
konq-plugins konqueror-plugin-adblock konqueror-plugin-akregator konqueror-plugin-autorefresh konqueror-plugin-babelfish konqueror-plugin-crashes konqueror-plugin-dirfilter konqueror-plugin-domtreeviewer konqueror-plugin-fsview konqueror-plugin-imagerotation konqueror-plugin-khtmlsettings konqueror-plugin-kimgallery konqueror-plugin-minitools konqueror-plugin-rellinks konqueror-plugin-searchbar konqueror-plugin-uachanger konqueror-plugin-validators konqueror-plugin-webarchiver konqueror-plugins python-plasma
$ dpkg -l | grep 4:4.2.3 | grep ^i | awk '{print $2}' | tr "\n" " "; echo
libkorundum4-ruby1.8 libqt4-ruby1.8 libsmokekde4-2 libsmokeqt4-2 python-kde4

If the upgrade to 4.2.4 was really a reason to fix it, then we should see some differences here. Suspicious are the python and other bindings not being 4.2.90...

[StephenKing (steffen-steffen-gebert)]

wow.. it's really like in kindergarden here...

I woun't follow this "discussion" any longer.. maybe I'll switch to any other distribution.
Steffen

[StephenKing (steffen-steffen-gebert)]

Ah... I already DO USE another distribution... The second PC with this problems runs sidux and also has this bug.

If you need infos, just tell me.

[dotancohen (dotancohen)]

There, fix released. It certainly seems to be a kde libs issue, and it has been fixed.

@RK: Please upgrade to the 4.2.4 PPA as described here:
http://www.kubuntu.org/news/kde-4.2.4

[dotancohen (dotancohen)]

@StephenKing:
Is your sidux install on KDE >= 4.2.3? If so then please post a link to the sidux bug, either here or directly to me in private mail, so that I can forward this upstream to KDE. Thanks.

[RK (kubuntu-rk)]

@dotancohen: Thanks.

It updated two packages: python-kde4 and python-plasma ... I directly went to 4.3b1 (and b2) because 4.2.4 is kinda pointless then. So much about "Beta2 would have any of the fixes that 4.2.4 would have." kdebindings and konqueror-plugins are definately simply missing in b1 and b2.

If this fixes it, then the bugs for ATI and nvidia would be invalid, while this one would not be "Fix Released" because the fix is in no official repository. (I don't know the correct procedure to get it there, though.) Too bad kubuntu is unusable without private repositories since more than a year now...

[RK (kubuntu-rk)]

The packages in question that improved the situation are python-kde4 and kde-workspace (for python-plasma), so add those and mark bug invalid everywhere else.

Unfortunately, I cannot confirm the bug to be completely fixed with this. After 11h uptime (3h40min since last plasma crash), I get disappearing texts in the window list widget and elsewhere. In between I had the effect that clicking on a python plasma widget (essentially the hello world + hal event listener + auto mount) had the immediate effect of all visible text disappearing. Clicking on the syslog or messages text file monitor let it re-appear. I could reproduce it not everytime, but a few times these clicks had the mentioned effect. Now I'm back to the previous level of corruption.

Maybe Jonathan Thomas could do something useful for a change and provide some insight into how these things could be debugged?

[dotancohen (dotancohen)]

> Maybe Jonathan Thomas could do something useful for a change and provide some
> insight into how these things could be debugged?

Are you trying to piss him off by any way that you can? You are not a developer, and you go around confirming and invalidating components, and playing ping pong confirmed/invalid with a dev. Then you make a remark like that?!? I personally am angry at you right now. I need this bug fixed, and you are going to piss off the developer so that he leaves us _both_ here to rot.

Go away. File a different bug, you have every right to as the dev implies that your issue is not the same as the OP (yes, that's me). Piss the devs off there. But let them do their work here and help those of us who appreciate it.

[RK (kubuntu-rk)]

> You are not a developer

I don't know where you get your information from, but your source doesn't seem to be reliable in any way.

> and you go around confirming and invalidating components,

I changed the state according to the knowledge accumulated in this thread. The same you did. Or are you saying that maintaining the bug database is a bad thing to do?

> and playing ping pong confirmed/invalid

Whenever I re-opened the bug, I each time added an explanation and or provided more evidence. Whenever that guy closed it, it was always with a dogmatic "driver issue, don't pester me" against all evidence accumulated so far. I don't think you have to blame _me_ here for not accepting that guys misbehaviour.

> Then you make a remark like that?!?

Check what that guy contributed to the bug fixing. He checked which driver you were using and then just kept insisting on the dogmactic view that "corruption occured => must be driver issue". I do not see any constructive action to get the bug fixed. Not in any way.

> I need this bug fixed

According your comment #35 the bug was fixed for you. I moved the bug to the only components that were different between 4.3b2 and 4.2.4. I also presented further evidence that make it likely to be related to those components. Do you think that was wrong? Do you think you would get the bug fixed if the only thing being done is shifting the blame to some driver that, according to presented evidence, is not at fault here?

> File a different bug,

What purpose would it have to file a different bug that's obviously a duplicate?

> to as the dev implies that your issue is not the same as the OP

That guy implies (in fact, dogmatically claims) a lot. Contrary to all evidence. Just don't let reality get into way of your opinion. Feel free to check the evidence provided and feel free to argue to the contrary.

If you actually want to get this bug fixed, then get someone to figure out why the "QPainter" could become "not active". See your comment #3. This seems to be the core issue here. And that needs to debugged in KDE4 land, not in any video driver.

If you want to do something on your own, deinstall KDE4 python bindings and check whether this fixes the corruptions or not. If it does not, then those where the wrong packages to blame. If it does, then you even have an explaination why not more people are affected.

[dotancohen (dotancohen)]

> > You are not a developer

> I don't know where you get your information from, but your
> source doesn't seem to be reliable in any way.

If you are a dev then I take my harsh statements back. It looked to me like you were getting on Thomas's case. If it's in-fighting between equals then I won't get involved, go at it!

> According your comment #35 the bug was fixed for you.

On my personal machine. But I've got a bunch of other users who have various related corruption issues and I'd like to see it resolved. I'm not even installing 9.04 on any more machines, and 8.10 couldn't be installed either. Kubuntu 8.04 was not LTS so I won't install that soon-to-be-outdated version, I have nothing to install.

[dotancohen (dotancohen)]

> If you actually want to get this bug fixed, then get someone to
> figure out why the "QPainter" could become "not active".

I requested on the KDE list that someone come over and take a look.

> If you want to do something on your own, deinstall KDE4 python
> bindings and check whether this fixes the corruptions or not

I can't at the moment, but lets see what the KDE folk say.

[Jonathan Thomas (echidnaman)]

Well, he's not a Kubuntu developer, that's for sure. No Kubuntu developer in their right mind would blame python-kde4, a source package that hasn't been used for two releases for corruption issues. The default desktop doesn't even really use it. All it is are python bindings to the KDE api, and since I have to explain that its pretty clear that RK is not a developer. He probably is just trying to jerk my chain. If you ever actually contribute anything more than a few bug reports here and there I may start to respect your judgment, but as it stands it is pretty clear that you do not know what you are doing.

But, being the nice guy that I am, I did do some more research on the bug. (I am also a KDE developer) Previously, the consensus between KDE developers was that this was a graphics driver bug. (For example, https://bugs.kde.org/show_bug.cgi?id=189059 , https://bugs.kde.org/show_bug.cgi?id=187991)

In reality this is a somewhat complex issue, with many factors determining whether or not the user will actually see the issue. It should be noted that two issues have been reported in this bug report initially. 1) The disappearing of the Plasma elements and b) the actual corruption in the panel. The panel corruption likely is a driver bug, and the disappearance issue is the complex one.

Anyways, on to the issue. As it turns out, graphical Plasma elements start to disappear when the video card runs out of pixmap space. Normally this doesn't happen, unless something is leaking pixmaps and is not destroying them after they are done being used. This explains why people using different video cards see this, and why not all people using the same video card see this. it could even be said that one person using the same driver as another but having a card with less VRAM could see this more often than the one who had a card with more VRAM (if he saw it at all)

Recently such a leak was found in Qt that fits the ticket. (Documented here: http://bugs.kde.org/show_bug.cgi?id=183191) There was a leak in the digital clock applet that exasperated the issue further, although it was not the core cause of the bug. (And definitely even the KDE component of this bug was not a kde4libs issue.)

In conclusion:

- This is not, I repeat *not*, and never ever was at all a kde4libs bug as you so adamantly proclaimed.
- Not a pykde issue, though this goes without saying
- Not an issue with the python-kde4 package which has been unused since Hardy
- This is a Qt issue which is scheduled to be fixed in Qt 4.5.2
- The KDE portion of this (which affected kdebase-workspace, but was not the main bug) has already been fixed in KDE 4.3 beta2.
- It was understandable to have thought that this was a driver issue, as that had been the previous consensus
- RK's little shenanigans would have not really changed anything, and the issue would have gone away once everybody upgraded anyway. All you did was cause me unnecessary stress in an attempt to piss me off.

You are quite lucky that I even took time to research this answer and give you this information, though I suppose I did it for those who care more about the bug than for those who care more about pissing me ...

[RK (kubuntu-rk)]

> This is not, I repeat *not*, and never ever was at all a kde4libs bug as you so adamantly proclaimed.

I never proclaimed it was a kde4libs bugs, I only refused to accept the bug being shifted to the video driver(s) without explanation, and quite rightfully so. I'd say it would be quite appropriate if you would acknowledge that your insistance on shifting the blame to the drivers was wrong and that the reported corruptions were, in fact, the same.

Additionally, it would be very helpful for the project if you would step down the high horse here. It shouldn't matter whether I am a kubuntu developer (which I never claimed to be), a debian developer, a kde developer, a developer (which I am) or a user. You went to a great length to explain why I'd be stupid, and even stooped to claim I couldn't be a developer just because I might not know every detail of KDE programming intrinsacies. I did try my best to provide insight based on what I saw and could figure out. You simply did not, -- except for your last comment. If you'd presented your insight earlier instead of commencing a close/open war, you could have saved us a lot of time. You could also have resisted from your self-righteousness, considering how wrong you have been here.

And let's not forget that what you call "shenanigans" prevented this bug from rotting away as a driver issue...

That being said, I still thank you for (finally) providing the insight that explains what we see.