KDE Linux/Calls

Where?

There: https://meet.kde.org/b/har-wqp-vt6-taa

Next Call

per-homedir Btrfs subvolumes

Person: Hadi

https://invent.kde.org/kde-linux/kde-linux/-/issues/393

/usr/local

Person: Hadi

https://invent.kde.org/kde-linux/kde-linux/-/merge_requests/347#note_1379173

UKIs

Person: Hadi

https://invent.kde.org/kde-linux/kde-linux/-/issues/470

Past Calls

2026-02-06

Turn on delta updates?

https://invent.kde.org/kde-linux/kde-linux/-/issues/304

Conclusion: do it!

Mirror syncs are a bit problematic, but that's is a problem anyway.
Update size for discover: push the MR in discover and the branch in kde-linux-sysupdated forward

Remove default Distrobox home override?

https://invent.kde.org/kde-linux/kde-linux-containers/-/merge_requests/9#note_1394032

Conclusion: follow upstream, remove override. Maybe drop distrobox at some point in the future when its replacement is ready. it offers a footgun in all scenarios imaginable :(

Does it matter that our default web browser defaults to Google as the search engine?

https://invent.kde.org/kde-linux/kde-linux/-/issues/201

Conclusion: irrelevant because we can't change it without violating Firefox's distribution policy or removing their branding.

TODO: figure out if we are in violation of the firefox distribution policy "You may not add to, remove, or change any part of the software, including ... not add any extensions to Firefox, change default settings, or alter search codes."
TODO: someone needs to research what makes our browsers better than any other option. The devs need to be blown away by e.g. Falkon before we can reasonably ship it by default

Use per-homedir Btrfs subvolumes for snapshotting purposes?

https://invent.kde.org/kde-linux/kde-linux/-/issues/393

Conclusion: wait for Hadi to come to a call or coordinate with him.

Plan for non en_US keyboard layouts with FDE?

https://invent.kde.org/kde-linux/kde-linux/-/issues/303#note_1384896

Conclusion: needs more research. Rebuilding the UKI is not an option. Look into other ways of supplying the information, such as sysext, cmdline, otherwise take this upstream for discussion. vconsole configuration needs to be unsigned somehow.

Stop including Snap?

https://invent.kde.org/kde-linux/kde-linux-packages/-/merge_requests/69

Conclusion: do it. Harald can use VSCode from the tarball, since it self-updates. Remove AppArmor; make someone present a use case first for it or SELinux. Then we can discuss it.

Eventually produce ISOs?

https://invent.kde.org/kde-linux/kde-linux/-/issues/287

Conclusion: no political objection, someone just needs to do it, and it's hard and annoying. "Patches welcome". But this won't fully fix VM woes since they typically default to BIOS-only and KDE Linux is UEFI-only. Needs talking to VM devs about better support for modern OSs

Have writable /usr/local?

https://invent.kde.org/kde-linux/kde-linux/-/merge_requests/347#note_1379173

Conclusion: for xdg overrides we can use xdg_data_dirs to inject a new dir. let Hadi chime in though before we conclude no.

Long-term plan for enthusiast/developer system extension?

https://invent.kde.org/kde-linux/kde-linux/-/issues/443

Conclusion: still no decision. Lasath presented Kapsule; see https://blog.lasath.org/2026/02/kapsule-completing-kde-linux.html. It's a promising option, but needs further development.

Misc/questions

etc should be versioned alongside the /usr so switching back to an older version also picks up the older etc.
Use flatpak preinstall once it's in a stable release.
Packages pipeline: move back to kde-builder. Ignore overlapping files for now. If we want this solved it needs solving at a different level (l10n? plasma? don't know). Make sure we don't fall behind on maintaining the yaml file!
TODO improve kde-builder distro deps so it doesn't pull all deps up front but only the ones it needs per project

2026-03-13

When? Fill in your availability at https://timeful.app/e/b5fab

Roadmap to user builds

Person: Nate

https://invent.kde.org/kde-linux/kde-linux/-/work_items/396

Conclusion: Yes let's focus on it, but we should first sort out our enthusiast extension story and the base

Long-term plan for enthusiast/developer system extension

Person: Nate

https://invent.kde.org/kde-linux/kde-linux/-/issues/443

Conclusion: go all in on Kapsule; things you install with it should be first-class citizens and then we strip out as much from the base OS as possible. Make it tiny!

Switch the base

Person: Nate

Conclusion: don't consider openSUSE, only consider Arch (status quo), Fedora, BuildStream. Current thoughts are that we shrink the base, migrate things to a Kapsule container, and then see what the win is if we switch the base to something else.