View previous topic :: View next topic |
Author |
Message |
Spanik Veteran

Joined: 12 Dec 2003 Posts: 1039 Location: Belgium
|
Posted: Sat Mar 01, 2025 1:18 pm Post subject: |
|
|
Changed those settings and went to kernel 6.6.52. That went ok. The message about the number of cpu's exceeding NR_CPU limit is gone. However the "kwin_x11[2921]: memfd_create() called without MFD_EXEC or MFD_NOEXEC_SEAL set" is still there. No change in behaviour once in Plasma.
Removed -wayland and -pipewire in USE. Emerge -puvUD @world will be for the rest of the weekend I fear:
Code: | daw ~ # emerge -puvDU @world
These are the packages that would be merged, in order:
Calculating dependencies... done!
Dependency resolution took 504.84 s (backtrack: 14/20).
......
Total: 574 packages (475 upgrades, 1 downgrade, 37 new, 30 in new slots, 31 reinstalls, 10 uninstalls), Size of downloads: 7,013,520 KiB
Conflict: 33 blocks (all satisfied)
WARNING: One or more updates/rebuilds have been skipped due to a dependency conflict:
|
Going to start that this evening. _________________ Expert in non-working solutions |
|
Back to top |
|
 |
pietinger Moderator

Joined: 17 Oct 2006 Posts: 5474 Location: Bavaria
|
Posted: Sat Mar 01, 2025 1:44 pm Post subject: |
|
|
Spanik wrote: | Going to start that this evening. |
I guess after that you should not have the error about kwin_x11 furthermore ... and hopefully more stability. (Please dont forget to do an "emerge -c" afterwards.) _________________ https://wiki.gentoo.org/wiki/User:Pietinger |
|
Back to top |
|
 |
Hu Administrator

Joined: 06 Mar 2007 Posts: 23156
|
Posted: Sat Mar 01, 2025 2:45 pm Post subject: |
|
|
Spanik wrote: | The message about the number of cpu's exceeding NR_CPU limit is gone. However the "kwin_x11[2921]: memfd_create() called without MFD_EXEC or MFD_NOEXEC_SEAL set" is still there. | I think that will be true whenever you run a kernel that can warn about that deprecated use and run a user process that performs the deprecated action. To fix this, you need a version of kwin_x11 that does not use the deprecated call. I do not know if such a version exists, or if it does, in what release this was first changed. |
|
Back to top |
|
 |
Spanik Veteran

Joined: 12 Dec 2003 Posts: 1039 Location: Belgium
|
Posted: Sat Mar 01, 2025 3:30 pm Post subject: |
|
|
I have been going a bit through what will be compiled... I'm worried:
12 times Rust
3 time python
2 times gcc
poppler
harfbuzz
All packages that I don't have good experiences with... _________________ Expert in non-working solutions |
|
Back to top |
|
 |
asturm Developer

Joined: 05 Apr 2007 Posts: 9361
|
Posted: Sat Mar 01, 2025 3:33 pm Post subject: |
|
|
Spanik wrote: | No change in behaviour once in Plasma. |
What change, exactly, do you expect from Plasma, right now?
As far as I can tell you have multiple, non-DE-related issues:
- "W10 VM's lock up" - without even telling us what kind of VM you're running
- "REW still doesn't start but now it complains about java." - what?
- "Audex still locks up the pc when starting a rip"
Then there were those:
- "changed scroll behaviour" - okay? is that a problem?
- "focus and focus stealing prevention working differently" - okay? is that a problem?
... both coupled with "But I have tried so many settings that there could be something still changed." - so, change the settings back to your liking?
Leaving this:
- "no session restore anymore" - does it work now, or what's your current status on that?
Spanik wrote: | All packages that I don't have good experiences with... |
not sure why, but...
Spanik wrote: | 12 times Rust
3 time python
2 times gcc |
There's no reason to have so many of those. It's just really excessive. Try clean up your system for a bit? |
|
Back to top |
|
 |
pietinger Moderator

Joined: 17 Oct 2006 Posts: 5474 Location: Bavaria
|
Posted: Sat Mar 01, 2025 4:12 pm Post subject: |
|
|
Spanik wrote: | [...] All packages that I don't have good experiences with... |
Maybe try an "emerge -cp" BEFORE your world-update ... because ... if you get a new gcc the switch to it is done automatically with this "emerge -c" OR you have to do it manually with "gcc-config" (or eselect) ... you can check the current used one with "gcc-config -l" ... (the same is true for rust and python) ... this is the reason why this is so important. _________________ https://wiki.gentoo.org/wiki/User:Pietinger |
|
Back to top |
|
 |
Spanik Veteran

Joined: 12 Dec 2003 Posts: 1039 Location: Belgium
|
Posted: Sat Mar 01, 2025 8:35 pm Post subject: |
|
|
asturm wrote: | Spanik wrote: | No change in behaviour once in Plasma. |
What change, exactly, do you expect from Plasma, right now?
As far as I can tell you have multiple, non-DE-related issues:
- "W10 VM's lock up" - without even telling us what kind of VM you're running
- "REW still doesn't start but now it complains about java." - what?
- "Audex still locks up the pc when starting a rip" |
Those worked fine in my DE before the change. Now they don't. And you keep suggesting it isn't related to the change in the DE?
Quote: | Then there were those:
- "changed scroll behaviour" - okay? is that a problem?
- "focus and focus stealing prevention working differently" - okay? is that a problem?
... both coupled with "But I have tried so many settings that there could be something still changed." - so, change the settings back to your liking? |
Yes, scroll is still changed, yes, it is an issue.
Yes, focus and focus stealing is (still) an issue.
I have changed them back, makes no difference.
Quote: | Leaving this:
- "no session restore anymore" - does it work now, or what's your current status on that? |
No, it doesn't work anymore.
Quote: | Spanik wrote: | All packages that I don't have good experiences with... |
not sure why, but...
Spanik wrote: | 12 times Rust
3 time python
2 times gcc |
There's no reason to have so many of those. It's just really excessive. Try clean up your system for a bit? |
Yes, the others in this thread do give good tips and suggestions. I will follow them and see where it leads. _________________ Expert in non-working solutions |
|
Back to top |
|
 |
Spanik Veteran

Joined: 12 Dec 2003 Posts: 1039 Location: Belgium
|
Posted: Sat Mar 01, 2025 8:41 pm Post subject: |
|
|
Hu wrote: | Spanik wrote: | The message about the number of cpu's exceeding NR_CPU limit is gone. However the "kwin_x11[2921]: memfd_create() called without MFD_EXEC or MFD_NOEXEC_SEAL set" is still there. | I think that will be true whenever you run a kernel that can warn about that deprecated use and run a user process that performs the deprecated action. To fix this, you need a version of kwin_x11 that does not use the deprecated call. I do not know if such a version exists, or if it does, in what release this was first changed. |
pietinger wrote: | Spanik wrote: | [...] All packages that I don't have good experiences with... |
Maybe try an "emerge -cp" BEFORE your world-update ... because ... if you get a new gcc the switch to it is done automatically with this "emerge -c" OR you have to do it manually with "gcc-config" (or eselect) ... you can check the current used one with "gcc-config -l" ... (the same is true for rust and python) ... this is the reason why this is so important. |
Ok I'll have my work cut out for tomorrow. I'll go through all that before the update. _________________ Expert in non-working solutions |
|
Back to top |
|
 |
Hu Administrator

Joined: 06 Mar 2007 Posts: 23156
|
Posted: Sat Mar 01, 2025 9:34 pm Post subject: |
|
|
A virtual machine lockup is far more likely to be a problem with the virtual machine software than with the desktop environment you use on the host, so yes, I think asturm is suggesting that is not related to the DE change, and I agree with that suggestion.
Likewise, if Audex really is causing the machine to hang, that cannot be a DE issue. Machine hangs are a hardware problem or a kernel problem.
Generally, you seem to be making it very hard for us to help you. Asturm hinted that knowing more about the VM might help us help you. You completely skipped answering that hint.
As regards Rust, that sounds like you're letting Portage build every Rust on the path to present. You probably don't need to do that. |
|
Back to top |
|
 |
asturm Developer

Joined: 05 Apr 2007 Posts: 9361
|
Posted: Sat Mar 01, 2025 9:37 pm Post subject: |
|
|
Spanik wrote: | Those worked fine in my DE before the change. Now they don't. And you keep suggesting it isn't related to the change in the DE? |
Correlation, not causation.
Spanik wrote: | Yes, scroll is still changed, yes, it is an issue.
Yes, focus and focus stealing is (still) an issue.
I have changed them back, makes no difference. |
When will you finally start telling us, *what* *exactly* *it is* you're seeing now that was not the case before. Why do *we* need to try so hard extracting every little bit of information out of you?
When you change an option, which one is it, from what setting do what other setting, what change do you expect and does it do something else than expected [what??] or nothing at all? |
|
Back to top |
|
 |
Spanik Veteran

Joined: 12 Dec 2003 Posts: 1039 Location: Belgium
|
Posted: Sun Mar 02, 2025 9:41 am Post subject: |
|
|
Ok. Now I suggest I first get this mess with the @world fixed and then see what is solved, what is still an issue. And open a specific thread for each separate issue if and when needed.
After the question from Pietinger about firmware and graphic cards I checked mine with the Nouveau page and this one doesn't need additional firmware. So that one is ok.
Hu wrote: | As regards Rust, that sounds like you're letting Portage build every Rust on the path to present. You probably don't need to do that. |
Looks like it. According to "eselect rust list" I only have 1.81.0 and that is selected. But it wants to build all from 1.74.0 to 1.84.0.
pietinger wrote: | Maybe try an "emerge -cp" BEFORE your world-update ... because ... if you get a new gcc the switch to it is done automatically with this "emerge -c" OR you have to do it manually with "gcc-config" (or eselect) ... you can check the current used one with "gcc-config -l" ... (the same is true for rust and python) ... this is the reason why this is so important. |
GCC on the other hand looks ok. According to gcc-config I have 2 versions installed and selected: AVR and X86_64 which fits. It wants to go to go to gcc-14.2.1 for both version so that covers both gcc builds.
Code: | daw ~ # emerge -cp
* Always study the list of packages to be cleaned for any obvious
* mistakes. Packages that are part of the world set will always
* be kept. They can be manually added to this set with
* `emerge --noreplace <atom>`. Packages that are listed in
* package.provided (see portage(5)) will be removed by
* depclean, even if they are part of the world set.
*
* As a safety measure, depclean will not remove any packages
* unless *all* required dependencies have been resolved. As a
* consequence of this, it often becomes necessary to run
* `emerge --update --newuse --deep @world` prior to depclean.
Calculating dependencies... done!
* Dependencies could not be completely resolved due to
* the following required packages not being installed:
*
* >=gui-apps/xwaylandvideobridge-0.4.0 pulled in by:
* kde-plasma/plasma-meta-6.1.5
*
* Have you forgotten to do a complete update prior to depclean? The
* most comprehensive command for this purpose is as follows:
*
* emerge --update --newuse --deep --with-bdeps=y @world
*
* Note that the --with-bdeps=y option is not required in many
* situations. Refer to the emerge manual page (run `man emerge`)
* for more information about --with-bdeps.
*
* Also, note that it may be necessary to manually uninstall
* packages that no longer exist in the repository, since it may not
* be possible to satisfy their dependencies. |
According to revdep-rebuild the system is consistent. Emerged xwaylandvideobridge again and then it listed this:
Code: | [All selected packages: =media-video/ffmpeg-chromium-126 =dev-qt/qtpaths-5.15.14 =dev-libs/qcoro5-0.10.0-r1 =kde-frameworks/kidletime-5.116.0 =media-libs/gavl-1.4.0-r3 =dev-build/autoconf-2.71-r7 =kde-frameworks/kimageformats-5.116.0 =dev-qt/designer-5.15.14 =kde-plasma/plasma-activities-5.116.0 =dev-libs/tinyxml2-10.0.0 =net-libs/accounts-qt-1.17 =media-plugins/gst-plugins-vpx-1.22.11 =media-libs/clutter-gst-3.0.27-r2 =media-libs/libzen-0.4.41 =gnome-base/gnome-desktop-44.0-r300 =media-video/movit-1.7.1 =kde-frameworks/countryflags-5.116.0 =kde-frameworks/kpeople-5.116.0 =kde-frameworks/kitemmodels-5.116.0 =kde-frameworks/kpty-5.116.0 =kde-frameworks/ktexteditor-5.116.0 =net-libs/libaccounts-glib-1.27 =virtual/perl-Math-Complex-1.620.0-r1 =kde-frameworks/kquickcharts-5.116.0 =media-video/mediainfo-23.10 =dev-libs/grantlee-5.3.0 =media-libs/libmediainfo-23.10 =kde-frameworks/kdelibs4support-5.116.0 =media-libs/gst-plugins-bad-1.22.11-r1 =media-libs/opencv-4.9.0-r2 =dev-libs/xapian-1.4.25 =kde-frameworks/purpose-5.116.0-r1 =dev-util/itstool-2.0.7-r2 =llvm-runtimes/compiler-rt-sanitizers-18.1.8-r1 =kde-frameworks/kfilemetadata-5.116.0 =kde-frameworks/kinit-5.116.0 =media-libs/sdl2-image-2.8.2 =media-video/gnome-video-effects-0.6.0 =kde-frameworks/threadweaver-5.116.0 =kde-plasma/plasma-activities-stats-5.116.0 =dev-perl/Mozilla-CA-20999999-r1 =kde-frameworks/kholidays-5.116.0 =net-libs/signon-oauth2-0.25_p20210102 =x11-libs/libXres-1.2.2 =kde-frameworks/prison-5.116.0 =kde-frameworks/kdnssd-5.116.0 =kde-frameworks/krunner-5.116.0 =media-libs/cogl-1.22.8-r3 =media-libs/mlt-7.24.0 =media-libs/rubberband-3.3.0-r1 =kde-frameworks/knotifyconfig-5.116.0 =sys-kernel/gentoo-sources-6.6.38 =llvm-core/clang-18.1.8 =dev-libs/flatbuffers-24.3.25 =kde-plasma/oxygen-sounds-5.27.11 =media-plugins/gst-plugins-v4l2-1.22.11 =dev-qt/qtsensors-5.15.14 =media-libs/gst-plugins-good-1.22.11 =llvm-core/clang-runtime-18.1.8 =media-libs/libebur128-1.2.6 =kde-apps/libkexiv2-23.08.5 =media-libs/clutter-gtk-1.8.4-r1 =dev-qt/qtmultimedia-5.15.14 =kde-frameworks/kunitconversion-5.116.0 =llvm-runtimes/compiler-rt-18.1.8-r1 =media-plugins/frei0r-plugins-1.8.0 =kde-frameworks/kjs-5.116.0 =kde-frameworks/qqc2-desktop-style-5.116.1 =virtual/perl-Math-BigInt-FastCalc-0.501.800 =kde-apps/khelpcenter-23.08.5 =virtual/perl-Math-BigRat-2.3.2 =media-plugins/gst-plugins-jpeg-1.22.11 =kde-frameworks/khtml-5.116.0 =dev-qt/qdbus-5.15.14 =dev-qt/qtquickcontrols-5.15.14 =kde-frameworks/syntax-highlighting-5.116.0 =media-sound/jamin-0.98.9_pre20170111 =media-libs/clutter-1.26.4-r1 =kde-frameworks/kemoticons-5.116.0 =perl-core/Test-Simple-1.302.196 =virtual/perl-bignum-0.670.0 =kde-apps/kaccounts-integration-23.08.5 =net-libs/signond-8.61-r100 =dev-qt/qtnetworkauth-6.7.2 =net-libs/accounts-qml-0.7_p20231028 =dev-python/ordered-set-4.1.0 =llvm-core/clang-toolchain-symlinks-18 =dev-qt/qtxmlpatterns-5.15.14 |
Seems mostly left over from kde/plasma 5 and qt5. Ran the depclean, unmerges xwaylandvideobridge again and ran emerge -puvUD @ world again.
Code: | Total: 574 packages (475 upgrades, 1 downgrade, 37 new, 30 in new slots, 31 reinstalls, 5 uninstalls), Size of downloads: 7,013,520 KiB
Conflict: 28 blocks (all satisfied) |
Still the same multiple Rust builds. Would it help to switch to rust-bin? At least it would reduce compile time. _________________ Expert in non-working solutions |
|
Back to top |
|
 |
pietinger Moderator

Joined: 17 Oct 2006 Posts: 5474 Location: Bavaria
|
Posted: Sun Mar 02, 2025 10:02 am Post subject: |
|
|
Spanik wrote: | Still the same multiple Rust builds. Would it help to switch to rust-bin? At least it would reduce compile time. |
Just let the world update run at night (like I do) and don't worry about rust ... when it's done, only a rust version will be updated in the future. _________________ https://wiki.gentoo.org/wiki/User:Pietinger |
|
Back to top |
|
 |
asturm Developer

Joined: 05 Apr 2007 Posts: 9361
|
Posted: Sun Mar 02, 2025 10:06 am Post subject: |
|
|
It appears your last emerge --sync was more than 2 months ago. So don't be surprised if your follow-up upgrade will contain updates to mostly the same packages in mostly the same total amount or even more.
@pietinger: Personally, I wouldn't like to build that many rust slot updates all for nothing, but hey ...
Last edited by asturm on Sun Mar 02, 2025 10:17 am; edited 1 time in total |
|
Back to top |
|
 |
Spanik Veteran

Joined: 12 Dec 2003 Posts: 1039 Location: Belgium
|
Posted: Sun Mar 02, 2025 10:17 am Post subject: |
|
|
I'll start it now. If some issues come up I'll see it before tomorrow. Next week is a busy work week, won't be behind the pc much. _________________ Expert in non-working solutions |
|
Back to top |
|
 |
logrusx Advocate


Joined: 22 Feb 2018 Posts: 2801
|
Posted: Sun Mar 02, 2025 11:07 am Post subject: |
|
|
You should be able to just emerge (one shot)rust-bin and unmerge regular rust. I don't support the suggestion to build all the rusts. It's insane amount of CPU cycles even if you were updating it regularly and amortized build time over a long period of time.
p.s. this is Portage & Programming, even though it started s Desktop Environments issue.
Best Regards,
Georgi |
|
Back to top |
|
 |
Spanik Veteran

Joined: 12 Dec 2003 Posts: 1039 Location: Belgium
|
Posted: Sun Mar 02, 2025 6:59 pm Post subject: |
|
|
Pc locked up solid about halfway. I let it like for another hour of 2 hoping it would come through but it didn't. Had to hard reset it. Suddenly Virtualbox now complains that it isn't in sync with the kernel version which is strange as the kernel is still the same as the last time it was booted. I now have sounds beeping and buzzing whatever I do. Something that is normally always off on my pc's. Looks as if some settngs have not been kept with updates.
Anyhow, it looks as it it locked up compiling boost. I'm starting it again but this time letting nothing else run. _________________ Expert in non-working solutions |
|
Back to top |
|
 |
dE_logics Advocate


Joined: 02 Jan 2009 Posts: 2300 Location: $TERM
|
Posted: Wed Mar 05, 2025 6:42 am Post subject: |
|
|
Wayland might have issues now, but it has a better architecture and is faster!! X will be obsolete someday. So a switch has to be done. _________________ My blog |
|
Back to top |
|
 |
dE_logics Advocate


Joined: 02 Jan 2009 Posts: 2300 Location: $TERM
|
Posted: Wed Mar 05, 2025 6:44 am Post subject: |
|
|
Spanik wrote: | Pc locked up solid about halfway. I let it like for another hour of 2 hoping it would come through but it didn't. Had to hard reset it. Suddenly Virtualbox now complains that it isn't in sync with the kernel version which is strange as the kernel is still the same as the last time it was booted. I now have sounds beeping and buzzing whatever I do. Something that is normally always off on my pc's. Looks as if some settngs have not been kept with updates.
Anyhow, it looks as it it locked up compiling boost. I'm starting it again but this time letting nothing else run. |
Seems like your PC is overheating. See sensors output. Temps must be below 95 generally speaking (for most processors). _________________ My blog |
|
Back to top |
|
 |
logrusx Advocate


Joined: 22 Feb 2018 Posts: 2801
|
Posted: Wed Mar 05, 2025 9:34 am Post subject: |
|
|
I honestly have never had overheating problems, but according to my experience this is consistent with going swap thrashing if the computer has HDD.
Best Regards,
Georgi |
|
Back to top |
|
 |
Spanik Veteran

Joined: 12 Dec 2003 Posts: 1039 Location: Belgium
|
Posted: Wed Mar 05, 2025 8:03 pm Post subject: |
|
|
Sorry for not replying earlier but busy week at the office and not so good health.
1: the update ran through the second time. Kernel updated as well. So far every application seems to work (haven't tried everything yet). Going to change to rust-bin before the next update.
2: no idea why it lock up during update. I was using the pc at the time so maybe something got updated and called that was not supposed to have changed in between? The pc has 64GB ram, swap is on an nvme ssd. Never had any heat related issue, fan are always at minimum speed. But has the pc didn't react to anything it was not possible to check.
3: lots more xruns on jack since I changed config_no_hz_idle = y and the config_hz = 300Hz. I had the habit (good or wrong) to set the timer to 1000Hz and not to use tickless from the days I recorded and edited more audio. Seems this still has influence even with the more powerfull cpu's these days.
4: session restore not working: did some google searching on this. And it looks like this is a deliberate choice of kad/plasma6. Previously they used X functionality for this. But wayland does not support this so they changed to a "fake session restore". They just launch every application again at startup. Said to be "work in progress". But this is just as worse as no session restore. You loose the point you were working/looking and all applications are just relaunched in the first desktop, all on top of each other.
5: W10 VM in Virtualbox locking up. This is a strange one. The XP VM's don't lock up, just work fine as they always did. But there is a different driver used for graphics between the XP VM and the W10 VM. I'll first try to transplant the W10 VM to my laptop that is still running kde/plasma5. That should tell if the machine is the problem or virtualbox under kds/pasma6 on the pc.
6: as for the focus and scrolling, it looks as if those have been design changes in kde/plasma6. Still haven't tried out all combinations of settings but so far it looks as if whatever combination I try it hardly changes. So I'm not happy as it makes everything harder to use FOR ME. YMMV.
7: audex... it is what it is. When it see a new cd in the player it just locks up the rest. Strangly it hold up kde applications, not others. So Dolphin cannot open a new instance or switch directories, Kate cannot open another .txt. But Claws or Vivaldi just putter along.
dE_logics wrote: | Wayland might have issues now, but it has a better architecture and is faster!! X will be obsolete someday. So a switch has to be done. |
Can't say about the architecture but I cannot say that it is faster. I take more time now to do the same things because of changed design choices because of the change of focus to wayland. I cannot help it that it is more a case of programmers wanting to scratch an itch and do shining new stuff.
But I wonder if when the change has been done we will not discover use cases that had always been available with X and then have to be added on with some string and rubber bands. And after that history will repeat itself and it will be superceded by something that implements X all again. Only there will be nobody around to remember that the new shiny thing is just old stuff in a new jacket.
Going to end it here. At the moment it is usable enough. I hope that the daily plasmashell crashes will now be gone. Going to try a bit more on the W10 VM and if it doesn't get solved, open a thread just for that.
Thanks for all your thoughts and input. _________________ Expert in non-working solutions |
|
Back to top |
|
 |
asturm Developer

Joined: 05 Apr 2007 Posts: 9361
|
Posted: Wed Mar 05, 2025 8:29 pm Post subject: |
|
|
Spanik wrote: | 4: session restore not working: did some google searching on this. And it looks like this is a deliberate choice of kad/plasma6. Previously they used X functionality for this. But wayland does not support this so they changed to a "fake session restore". They just launch every application again at startup. Said to be "work in progress". But this is just as worse as no session restore. You loose the point you were working/looking and all applications are just relaunched in the first desktop, all on top of each other. |
You're talking as if you switched to Wayland session. Why didn't you keep using X11 session then?
Spanik wrote: | 6: as for the focus and scrolling, it looks as if those have been design changes in kde/plasma6. Still haven't tried out all combinations of settings but so far it looks as if whatever combination I try it hardly changes. So I'm not happy as it makes everything harder to use FOR ME. YMMV. |
You still have not used a single word to describe what it actually is that changed for you.
Spanik wrote: | 7: audex... it is what it is. When it see a new cd in the player it just locks up the rest. Strangly it hold up kde applications, not others. So Dolphin cannot open a new instance or switch directories, Kate cannot open another .txt. But Claws or Vivaldi just putter along. |
That's it? You're just accepting a clear anomaly that no one else seems to have?
Spanik wrote: | Going to end it here. At the moment it is usable enough. I hope that the daily plasmashell crashes will now be gone. |
And this is even the first time you are talking about a plasmashell crash in here ... |
|
Back to top |
|
 |
Spanik Veteran

Joined: 12 Dec 2003 Posts: 1039 Location: Belgium
|
Posted: Thu Mar 06, 2025 3:32 pm Post subject: |
|
|
asturm wrote: | You're talking as if you switched to Wayland session. Why didn't you keep using X11 session then? |
Because I am running an X11 session? I don't know what I have to do to convince you I start an X11 session but I'm out of ideas.
https://quirinus.one/gentoo/kde6.jpg
https://quirinus.one/gentoo/kde6_3.txt
Interesting bits from the kwin info:
Code: | Version
=======
KWin version: 6.2.5
Qt Version: 6.8.2
Qt compile version: 6.8.2
XCB compile version: 1.17.0
Operation Mode: X11 only
Build Options
=============
KWIN_BUILD_DECORATIONS: yes
KWIN_BUILD_TABBOX: yes
KWIN_BUILD_ACTIVITIES: yes
HAVE_X11_XCB: yes
HAVE_GLX: yes
X11
===
Vendor: The X.Org Foundation
Vendor Release: 12101015
Protocol Version/Revision: 11/0
SHAPE: yes; Version: 0x11
RANDR: yes; Version: 0x14
DAMAGE: yes; Version: 0x11
Composite: yes; Version: 0x4
RENDER: yes; Version: 0xb
XFIXES: yes; Version: 0x50
SYNC: yes; Version: 0x31
GLX: yes; Version: 0x0
.....
Output backend
==============
Name: KWin::X11StandaloneBackend
....
Compositing
===========
Compositing is active
Compositing Type: OpenGL
OpenGL vendor string: Mesa
OpenGL renderer string: NVD9
OpenGL version string: 4.3 (Compatibility Profile) Mesa 24.3.4
OpenGL platform interface: GLX
OpenGL shading language version string: 4.30
Driver: Unknown
GPU class: Unknown
OpenGL version: 4.3
GLSL version: 4.30
Mesa version: 24.3.4
X server version: 1.21.1
Linux kernel version: 6.12.16
Direct rendering: Requires strict binding: yes
Virtual Machine: no
OpenGL 2 Shaders are used
....
|
asturm wrote: | You still have not used a single word to describe what it actually is that changed for you. |
If you insist on it:
- when you open Dolphin and you have a list of 2000 files. You now put your mouse in the scrollbar halfway and click. With kde5 it would act as a "page down" so you could scroll fast through the files. With kde6 it jups to where your cursor stands, to halfway down the list. This is not limited to Dolphin but all applications showing a scrollbar. Both horizontal and vertical scrollbars.
- I use "focus follows mouse" and focus stealing prevention "high". In kde5 when you launch an application from the menu it comes to the front of the window stack. In kde6 in goes to the back of the stack. Even windows like hen you power down or restart.
- if you middle-click on an active application in the taskbar to open another instance the new instance comes at the back of the window stack, not the front like kde5
- if you open an application in a virtual desktop and you switch desktop before it has opened completely then it opens in the last desktop, not the one it was started in
- I have set that I only want to see in the taskbar the windows of the desktop I'm working on. But it still shows those from other windows if they pop up on another desktop
- even when an application pops up a dialog box it ends up on the bottom of the window stack. This is mostly visible in Gimp where dialog boxes (like if you want to rotate at open, or what to do with the color profile) when you open a file end up not noticed because they are hidden behind the main application.
All in all a bunch of minor things but they make a break from how kde has always been behaving.
asturm wrote: | That's it? You're just accepting a clear anomaly that no one else seems to have? |
Yes. I find very little about Audex, there seem to be not much users at all. and I'm not happy about a few other things as well. But it is so far the only ripper I found that treats characters with all sorts of accents without issues. And as it is only when you put in a new cd to be ripped it is managable.
So yes, I put up with it.
asturm wrote: | And this is even the first time you are talking about a plasmashell crash in here ... |
It happened maybe 3 times in several months. And mostly I did noticed it even because it was in a window I found behind several others (see the focus issue). The screen didn't even flicker so I only knows it ever happened because I found a window telling me it occured.
As for kde/plasma and wayland, it is very obvious they have chosen to go that way and abandon x11. From a developper of kde:
Quote: |
fusionfuture
KDE Developer
The X11 session is basically untested by developers nowadays. I happened to hit the bug and submitted a fix but I didn’t expect the destructiveness. |
See post 6 here: [url]https://discuss.kde.org/t/plasma-6-3-update-a-lot-of-problems-with-sessions-x11/29930/6[/quote] The whole thread goes about issues with x11 sessions. _________________ Expert in non-working solutions |
|
Back to top |
|
 |
keekkenen n00b

Joined: 05 Oct 2024 Posts: 43
|
Posted: Thu Mar 06, 2025 4:50 pm Post subject: |
|
|
I'd only one problem with plasma. It's any apps freeze if they use file systems.
What's trouble? I'm using samba mount NAS shared folders and if I switch to another Wi-Fi, it means changing network and after that, any actions with Kate, Dophin, etc, I mean any file operation blocked the application for a long time. Its behavior could more smart. _________________ 7950x3d / x670e MSI Tomagawk / Sapphire RX 7800XT 16Gb / G.Skill 64Gb 5600 / A-Data Legend 960 2Tb (x2), A-Data SX8200PNP 256Gb |
|
Back to top |
|
 |
asturm Developer

Joined: 05 Apr 2007 Posts: 9361
|
Posted: Thu Mar 06, 2025 4:52 pm Post subject: |
|
|
Spanik wrote: | asturm wrote: | You're talking as if you switched to Wayland session. Why didn't you keep using X11 session then? |
Because I am running an X11 session? I don't know what I have to do to convince you I start an X11 session but I'm out of ideas. |
I *had* to ask since you were talking about Wayland session specifics that are not valid in X11, where the old session save/restore code is still present.
Iirc you previously described some kind of window position saving bug also reported here: https://bugs.kde.org/show_bug.cgi?id=486265
... at least thanks to your screenshots and output we know that you have finally upgraded to current stable Plasma 6.2.5.
Spanik wrote: | asturm wrote: | That's it? You're just accepting a clear anomaly that no one else seems to have? |
Yes. I find very little about Audex, there seem to be not much users at all. and I'm not happy about a few other things as well. But it is so far the only ripper I found that treats characters with all sorts of accents without issues. And as it is only when you put in a new cd to be ripped it is managable.
So yes, I put up with it. |
Well, it's not normal, as it works fine for me.
Spanik wrote: | asturm wrote: | And this is even the first time you are talking about a plasmashell crash in here ... |
It happened maybe 3 times in several months. And mostly I did noticed it even because it was in a window I found behind several others (see the focus issue). The screen didn't even flicker so I only knows it ever happened because I found a window telling me it occured. |
Wait:
Spanik wrote: | I hope that the daily plasmashell crashes will now be gone. |
Which is it now, "daily" or 3 times in several months?  |
|
Back to top |
|
 |
Spanik Veteran

Joined: 12 Dec 2003 Posts: 1039 Location: Belgium
|
Posted: Thu Mar 06, 2025 6:21 pm Post subject: |
|
|
keekkenen wrote: | I'd only one problem with plasma. It's any apps freeze if they use file systems.
What's trouble? I'm using samba mount NAS shared folders and if I switch to another Wi-Fi, it means changing network and after that, any actions with Kate, Dophin, etc, I mean any file operation blocked the application for a long time. Its behavior could more smart. |
Could be. It is always at the beginning when it is reading the content of the cd and then again when it has to write a ripped file to the disk. I have checked cpu usage but that is at 0.2%. And it is indeed typically Kate and Dolphin locking up until audex has done reading/writing.
asturm wrote: | Spanik wrote: | asturm wrote: | You're talking as if you switched to Wayland session. Why didn't you keep using X11 session then? |
Because I am running an X11 session? I don't know what I have to do to convince you I start an X11 session but I'm out of ideas. |
I *had* to ask since you were talking about Wayland session specifics that are not valid in X11, where the old session save/restore code is still present. |
I have been telling you that I'm using x11, see my third and fourth post which contain the same info as my screenshot. The issues I find are "since kde/plasma made the switch to wayland", forcing me to enable the wayland use flag.
That is something you have to read about in the KDE sites: the old x11 session restore isn't used anymore even in an x11 session because the "fake session restore" they have devised is "working there as good in x11". Their words, not mine. Kde has made the switch to wayland and x11 isn't even tested anymore by the (main) developers. So until wayland is really on par with x11 or all the missing functionality is re-written, kde isn't a good idea if you want to keep using x11. Probably other DE are similar, depending on which features you really like/want/need.
asturm wrote: | Wait:
Spanik wrote: | I hope that the daily plasmashell crashes will now be gone. |
Which is it now, "daily" or 3 times in several months?  |
Ok, that daily was an exageration. Less than daily, I don't use the pc daily the last weeks. Job got in the way. I just clicked them away as everything kept working.
It more or less works now. Only since kernel 6.12.16 I have more xruns in jack and virtualbox is having issues. So I'm now going to try to get that ironed out. _________________ Expert in non-working solutions |
|
Back to top |
|
 |
|