* chore: cosmic is stable now, don't hide it behind --advanced flag
* fix: remove --advanced check in GreeterType
* chore: remove unused import to make linter happy
* feat(applications): add CUPS installation support
* fix: use translation for print service preview_action
* fix: incorrect action for print service menu item
* chore: refactor naming, printer -> print service
* fix: commit untracked file
* chore: fix formatting to make linter happy
* On horrible hardware this makes it so that the "Graphics drivers" section loads directly.
By initializing it with launch instead of on the fly.
* Init after logs
* This cache the property of graphics drivers properly.
Instead of trying to hack early init.
* Add host-to-target (H2T) installation mode detection
- Add running_from_host() function to detect if running from installed system vs ISO
- Function checks for /run/archiso existence (ISO mode) vs host mode
- Add clear logging of installation mode on startup
- Skip keyboard layout changes when running from host system
- Fix Pyright type error in jsonify() by using Any instead of object
- Update README to mention installation from existing system
This enables archinstall to be run from an existing Arch installation
to perform host-to-target installs on other disks/partitions.
* match existing style
* rem debug
* info -> debug
* Add Kurdish language
This pull request introduces full Kurdish language integration into archinstall, allowing Kurdish speaking users to navigate and use the installer entirely in their native language.
* Fix translation for timezone selection prompt
* Menu now filters and sorts using priority, improving UX.
* Refactor: improve logic, removed redundancy
* Refactor: improve logic, removed redundancy
* Improved logic when getting view items in menuItems
* Improved logic when getting view items in menuItems
As mentioned by @svartkanin on #3950, given we now have a way for the user to
explicitly specify if they want to install to the removable location, having a
fallback like this seems undesirable.
On top of that, as mentioned by @correctmost on the same PR, the code that said
PR introduced was bugged and would always raise an exception anyways.
Commit c095eb56d8 was supposed to introduce logic
such that if the `grub-install` command failed with a `--removable` flag, then
another attempt would be made with such flag removed.
This was broken because the `--removable` flag was kept in both cases (likely a
copy-paste mistake). This has been an issue since, in all future iterations of
the code.
What this commit does is fix this logic, but also invert the cases tested:
first test without `--removable`, then add it should that case fail, as this is
the most sensible thing to do.
- Allows for white space in between groups, aligning better with displayed example.
- Removed unneeded | symbol, which was checking as literal rather than working as "or %"
* Do not create BLS and Limine entries for fallback initramfs
Fallback initramfs seem to no longer be built by default.
* Remove initramfs variant logic altogether