If you use a package that depends on pyside2, then you might be stuck. You could try building pyside2 with a modified PKGBUILD. This seems like a bug/difference in behavior with paru: https://github.com/Morganamilo/paru/issues/1377

If you’re adventurous, you could use another package management tool. Not familiar with garuda though.

Or maybe wait until paru developers possibly change their tool, or the AUR package happens to change to work around paru behavior?

Seems like there’s an issue with pyside2 and paru: https://aur.archlinux.org/packages/pyside2

If you don’t need it, remove pyside2?

Good to hear. I think jellyfin could be improved so that it doesn’t completely clear migrations.xml every time a new migration is applied when it starts up (switch from a Create to an OpenOrCreate on the file). Although a low disk situation can be difficult for software to deal with typically anyway.

Can you check your config folder (seems like it’s /etc/jellyfin for you) for migrations.xml and system.xml? I’m guessing if you were low on disk, jellyfin tried to write to them, but couldn’t, so it left empty files which it can’t now read from.

Sounds like https://bugs.kde.org/show_bug.cgi?id=465864

Do you see an option to import other images in your dashboard? For ubuntu, you’d get those from https://cloud-images.ubuntu.com/

Is that from the deltarpms? It’s a pretty common problem with them. You can disable deltarpms if you don’t mind installing the full package by default. Though if you’re concerned about the cost of the data, deltarpms are probably exactly what you want 🫤

The modularity thing seems too gimmicky to me. Like the Motorola phones with the same concept. I do think it’s a good direction for repairability though.

Do either have a touchpad as good as the ones on macs or xps?

I was under the impression that like Oracle, they used RHEL sources as their base (from git.centos.org). But it appears that they now (as of 2022) only use fedora sources and maintain other sources on top of fedora, so they’ve deviated from RHEL compatibility as far as I can tell.

I’m curious to see if oracle, amazon, or suse will try to absorb some of the RHEL derivatives like alma and rocky. Right now there seems to be a lot of fragmentation in RHEL derivatives. Not to say they are trying to compete with Red Hat, but Amazon and Oracle seem like they would try to do so this way.

Reminds me of LEDs used to extract data out of airgapped networks: https://threatpost.com/blinking-router-leds-leak-data-from-air-gapped-networks/126199/ but different because that one requires software within the airgapped network to exfiltrate.

Didn’t there used to be network equipment that was vulnerable to data exfil from data status lights, or did I dream that up? Most data LEDs now just consistently blink.

Surprisingly readable. Some of the letters are really close to each other, and multiple capitals together look odd. I will try it 🙂

Slacks problem isn’t with generating things to say, it’s with managing attention spans and navigating between conversations easily. If the integration can help with that, maybe it will make sense. But if it just helps people write, that is nice but I’d be concerned that it would lead to longer and less direct messaging.

I’d check /var/log/pacman.log for anything you recently updated which might give a hint. Maybe check that you can reach your local router via a browser or network devices using ping?

Have you tried the tiling assistant or gtile extensions? Not quite the same as a regular tiling window manager, but tiling assistant works pretty we’ll for just snapping and moving windows.