QV version 240425 pre-alpha uploaded

[Federico]

After installing the KDE meta package kde5-6.0.4_2 Konqueror still does not work, but at least for us it's not that important. The important point is Dolphin with root access and that was my fault: I did not read fully the instructions from PKGget. So, sorry, I apologize.
To get system wide root access, one can just execute /usr/bin/dolphin.bin0 .
Executing /usr/bin/konqueror.bin0 does not seem to make any difference regarding Konqueror.

Now installing AnyDesk with the flatpak from flathub.

[Federico]

@ Caramel

/usr/lib/libpam.so.0 is a symbolic link to libpam.so.0.85.1. I renamed it and rebooted. This time AnyDesk still does not work, but I do get some errors:

Code: Select all

# ./anydesk

(process:5806): GLib-GObject-CRITICAL **: 17:21:47.008: invalid (NULL) pointer instance

(process:5806): GLib-GObject-CRITICAL **: 17:21:47.008: g_signal_connect_data: assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed

(process:5806): GLib-GObject-CRITICAL **: 17:21:47.008: invalid (NULL) pointer instance

(process:5806): GLib-GObject-CRITICAL **: 17:21:47.008: g_signal_connect_data: assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed

(process:5818): GLib-GObject-CRITICAL **: 17:21:47.060: invalid (NULL) pointer instance

(process:5818): GLib-GObject-CRITICAL **: 17:21:47.060: g_signal_connect_data: assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed

(process:5818): GLib-GObject-CRITICAL **: 17:21:47.060: invalid (NULL) pointer instance

(process:5818): GLib-GObject-CRITICAL **: 17:21:47.060: g_signal_connect_data: assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed
# 
(process:5835): GLib-GObject-CRITICAL **: 17:21:50.346: invalid (NULL) pointer instance

(process:5835): GLib-GObject-CRITICAL **: 17:21:50.346: g_signal_connect_data: assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed

(process:5835): GLib-GObject-CRITICAL **: 17:21:50.346: invalid (NULL) pointer instance

(process:5835): GLib-GObject-CRITICAL **: 17:21:50.346: g_signal_connect_data: assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed

# ./anydesk

(process:5988): GLib-GObject-CRITICAL **: 17:26:25.184: invalid (NULL) pointer instance

(process:5988): GLib-GObject-CRITICAL **: 17:26:25.185: g_signal_connect_data: assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed

(process:5988): GLib-GObject-CRITICAL **: 17:26:25.185: invalid (NULL) pointer instance

(process:5988): GLib-GObject-CRITICAL **: 17:26:25.185: g_signal_connect_data: assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed

(process:6000): GLib-GObject-CRITICAL **: 17:26:25.236: invalid (NULL) pointer instance

(process:6000): GLib-GObject-CRITICAL **: 17:26:25.236: g_signal_connect_data: assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed

(process:6000): GLib-GObject-CRITICAL **: 17:26:25.236: invalid (NULL) pointer instance

(process:6000): GLib-GObject-CRITICAL **: 17:26:25.236: g_signal_connect_data: assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed

I tested with the flatpak

@ Caramel

Finally it works. Thank you.

[Caramel]

Method to use the binary anydesk in QV.

First : Neutralize PAM (see viewtopic.php?p=119911#p119911)

Second : After downloading and extracting the archive (for Generic Linux (64 Bit)) from anydesk.com , make sure that the binary anydesk is owned by root and that its permission is 755 (right to read and execute to all, right to write to root)

Third : Move (or copy) the binary anydesk to /usr/bin. Move (or copy) anydesk.desktop to /usr/share/applications

Fourth : Open the Login and Security Manager ("Configurer votre identifiant " in French, "Anmelde- und Sicherheitsmanager" in German) in the system section of the menu) and use it to set anydesk to run non-root (It will run as the anydesk user)

Fith : With PKGet, install gtkglext

PS : not install at the same time that the flatpak

Anydesk shoud work

[Federico]

@ Caramel

It did not only run, we have used it for an hour without even the smallest issue. The connection between Germany and Italy ran also silk smooth, without any stuttering or disconnection, even if the internet connection was only average fast.

So, basically, QV already works quite well. The biggest problem remaining seems to be this X server issue and I hope that it isn't related to some strange bug of the kernel 6.8 (similarly to what happened recently on Kirkstone with the Kernel 5.15).
My father is now going to also try QV 240506 on his laptop (AMD Picasso) and will report on the situation. So we'll be able to at least discern if the problem is general (kernel related) or only concerns the Intel driver.

Remaining questions are obviously also the absence of any update mechanism, bluetooth support, maybe hibernation (for the ones who need it for some reason) and I honestly cannot think at anything else at the moment.

Overall, I have to say that we were quite well impressed by QV. We also liked that fact that any app (even Ungoogled Chromium) can be easily executed as root by using an alternative executable file, and this is even true for flatpaks and appimages. And also obviously by the increased performance of the Linux kernel 6.

[Federico]

@ all
So, I restarted the system and, as usual, the X server did not start:

- Black screen
- System fan spinning fast
- Command prompt still available through Ctrl+Alt+F2

at this point I was going to format the usb stick once again. But I thought: why not to try it on my desktop first (Nvidia RTX 3080)?
Well, incredibly it booted up, perfectly, in 4K!
I decided to immediately try it again on my laptop, for sorting out somehow what the culprit of the issue could be. And, incredibly, it booted ok again on my laptop. But why?
So, I rebooted again on my laptop, and this time the X server started, but only in 800x600 pixel.
Restarting once again, it booted up normally.

So, at the end of the day, I think that I can safely say that at least for the video chip of my laptop (Intel Iris Xe - driver Intel i915 -) the detection of the screens is buggy. This should have nothing to do with the kernel.

I have also tried to boot and reboot on the desktop PC. The first time it booted, the second time I got black screen (and now I'm back on the laptop, on which it booted fine). So it looks like the X server always works when the driver is switched (from Intel to Nvidia or vice versa), and never works more than twice when the system is rebooted on the same machine. Therefore, another explanation may be that there's something wrong within the shutdown procedure.

[williwaw]

Sorry, I might be a bit slow to respond to QV bug reports for a few days.
I'm finally giving EasyOS Kirkstone some TLC. A complete recompile with updated packages.

I understand scarthgap is the focus right now.
just some notes for when you get back to QV

epdfview not installed, menu still shows evince
remove snapshot in boottime menu fails with syntax error line 44

[Federico]

We just discovered on the new EasyOS Scarthgap thread that the issue with the X server is just caused by the german locale. Just set english as locale and the issue disappeared. I haven't tested other locales jet. The keyboard is not a problem (you can set it as needed according to its physical layout) but the system locale must be set as english, otherwise you are going to face severe problems with the X server.

[BarryK]

I am intending to return to QV development, after working on Scarthgap.

Have read through these forum posts, and my blog posts, to refresh my memory.

QV has an extremely interesting structure, and being based on Void packages has access to a large repository.

The btrfs developers are still not interested in taking fscrypt any further, which is a shame.

I don't like the Void rolling release model, but will just have to accept it.

One important thing left undone is an update mechanism. I posted before that you would only have to download the big .img file once, then there could be small .pet packages to perform updates. Actual package updates are taken care of by the xbps package manager.

The kernel is 6.8.1 in the last release of QV, so first thing will compile the latest 6.8.x

[FeodorF]

Hallo @BarryK .

While you compiled the new kernel I hope you didn't miss this.

Ref.: Chromebook3 touch pad problem (solved)

https://forum.puppylinux.com/viewtopic. ... 68#p130868 (dimkr)
https://forum.puppylinux.com/viewtopic. ... 33#p131033 (rockedge)

Forum member dimkr fixed this problem.
'CONFIG_CROS_EC_LPC=m'

https://www.kernelconfig.io/config_cros_ec_lpc

config_cros_ec_lpc - kernelconfig.io
If you say Y here, you get support for talking to the ChromeOS EC over an LPC bus, including the LPC Microchip EC (MEC) variant. This uses a simple byte-level protocol with a checksum. This is used for userspace access only. The kernel typically has its own communication methods. To compile this driver as a module, choose M here: the module will be called cros_ec_lpcs.

Regards,
Feodor

[williwaw]

I am intending to return to QV development, after working on Scarthgap.

Have read through these forum posts, and my blog posts, to refresh my memory.

QV has an extremely interesting structure, and being based on Void packages has access to a large repository.

The btrfs developers are still not interested in taking fscrypt any further, which is a shame.

I don't like the Void rolling release model, but will just have to accept it.

I am curious if the renewed interest is more about btrfs or the void repos?

[BarryK]

I am intending to return to QV development, after working on Scarthgap.

Have read through these forum posts, and my blog posts, to refresh my memory.

QV has an extremely interesting structure, and being based on Void packages has access to a large repository.

The btrfs developers are still not interested in taking fscrypt any further, which is a shame.

I don't like the Void rolling release model, but will just have to accept it.

I am curious if the renewed interest is more about btrfs or the void repos?

I suppose both.

The way that QV is setup for snapshots is very interesting, probably has a lot of potential.

And of course a good size package repo. I recall Federico being happy because KDE Dolphin file manager worked.

[BarryK]

Hallo @BarryK .

While you compiled the new kernel I hope you didn't miss this.

Ref.: Chromebook3 touch pad problem (solved)

https://forum.puppylinux.com/viewtopic. ... 68#p130868 (dimkr)
https://forum.puppylinux.com/viewtopic. ... 33#p131033 (rockedge)

Forum member dimkr fixed this problem.
'CONFIG_CROS_EC_LPC=m'

https://www.kernelconfig.io/config_cros_ec_lpc

config_cros_ec_lpc - kernelconfig.io
If you say Y here, you get support for talking to the ChromeOS EC over an LPC bus, including the LPC Microchip EC (MEC) variant. This uses a simple byte-level protocol with a checksum. This is used for userspace access only. The kernel typically has its own communication methods. To compile this driver as a module, choose M here: the module will be called cros_ec_lpcs.

Regards,
Feodor

Right now compiling the 6.6.52 kernel, for both EasyOS and QV.

Have set both of these as modules:

Code: Select all

CONFIG_CROS_EC=m
CONFIG_CROS_EC_LPC=m

[BarryK]

Sorry, I might be a bit slow to respond to QV bug reports for a few days.
I'm finally giving EasyOS Kirkstone some TLC. A complete recompile with updated packages.

I understand scarthgap is the focus right now.
just some notes for when you get back to QV

epdfview not installed, menu still shows evince

Yes, evince is the choice for pdf viewer.
Is epdfview better?

remove snapshot in boottime menu fails with syntax error line 44

...I don't know what file this is referring to.