On USB-C-only laptops, multiple power_supply udev events fire at once on
resume, and each tries to create the same --unit=omarchy-power-profile
transient unit, so all but the first fail with "unit already loaded" and log
an error on every wakeup. Drop the fixed unit name so each event gets an
auto-generated one; powerprofilesctl set is idempotent, so concurrent runs are
harmless, and --collect still garbage-collects each unit on exit.
Closes#6091.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adapts basecamp/omarchy#6200. The sof-audio-pci-intel-* driver family
(Arrow Lake, Meteor Lake, Tiger Lake, Alder Lake, and more) needs
sof-firmware to boot the DSP, but mainline linux only optdeps it, so
fresh installs came up with no audio (Dummy Output only).
Replace the Panther/Wildcat Lake GPU checks with omarchy-hw-intel-sof,
which detects any Intel PCI audio controller. Drop the XPS exclusion
since omarchy-pkg-add is idempotent and the check is harmless there.
Remove the now-orphaned omarchy-hw-intel-wcl detector and backfill
existing installs via migration.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* Configure Dell XPS haptics directly
* Use packaged Dell XPS touchpad haptics service
* fix: keep one Dell haptics migration
* We already have a dedicated Trigger > Hardware menu
* Name for action
* Moved
* Need to make the package available for offline
* Simplify
---------
Co-authored-by: David Heinemeier Hansson <david@hey.com>
`DefaultLimitNOFILESoft` is not a valid systemd configuration key.
systemd silently ignores it and emits this warning on every boot,
for both the system manager (PID 1) and the user manager:
Unknown key 'DefaultLimitNOFILESoft' in section [Manager], ignoring.
The consequence is that the script's stated goal of raising the soft
fd limit from 1024 to 65536 has not actually been happening: dev
tools (VS Code, Docker, dev servers, databases) continue to run with
systemd's default 1024 soft limit.
The valid key is `DefaultLimitNOFILE`, which accepts either a single
value (sets both soft and hard) or `SOFT:HARD`. Using `65536:524288`
raises the soft limit while preserving systemd's typical 524288
hard-limit default — using the bare `=65536` form would lower the
hard limit, which would be a regression.
After this change and `systemctl daemon-reexec` on both managers,
`systemctl show | grep DefaultLimitNOFILE` reports:
DefaultLimitNOFILE=524288 (hard)
DefaultLimitNOFILESoft=65536 (soft, derived display attribute)
and the boot-time "Unknown key" warnings disappear.
(The typo likely originated from someone reading `systemctl show`
output, where `DefaultLimitNOFILESoft=` appears as a read-only
display attribute for the soft component of the parsed limit.
It looks like a config key but isn't.)
* Adjust condition to also match non-X series Panther Lake GPUs.
* Add migration to install GPU acceleration on non-X Panther Lake systems.
* Update comment with Non-Arc Panther Lake GPUs mentioned.
* Simplify conditions in the migration
* Extract ASUS Expertbook B9406 display backlight fix and apply it for Zenbook ux5406aa too
* Add migration for Panther Lake Asus display bakclight fix
* Check if limine-update exists before executing it in the migration
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
---------
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
* Add bass speaker fix for Lenovo Yoga Pro 7 14IAH10
The ALC287 codec on Yoga Pro 7 14IAH10 requires the alc287-yoga9-bass-spk-pin
quirk to properly route audio to both amplifier speakers. Without this fix,
only one speaker works and bass output is missing.
This adds:
- Hardware detection script (omarchy-hw-lenovo-yoga-pro7-bass)
- Audio fix that creates /etc/modprobe.d/lenovo-yoga-pro7-bass.conf
Tested on Lenovo Yoga Pro 7 14IAH10 (Intel Core Ultra 9 285H).
Reference: https://wiki.archlinux.org/title/Lenovo_Yoga_9i_2022_(14AiPI7)
* Add Lenovo Yoga Pro 7 bass fix to install flow
Addresses review feedback: the hardware fix script was not invoked
during installation. Add it to install/config/all.sh so it runs
automatically on matching hardware.
* Fix implementation
---------
Co-authored-by: David Heinemeier Hansson <david@hey.com>
* Add Transcode entry to Nautilus context menu
Right-click on image or video files in Nautilus to send them to
omarchy-transcode in a presentation terminal, mirroring the existing
LocalSend integration.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* Use shlex.quote for shell-safe command construction in transcode.py
Quote both the resolved binary path and file paths via shlex.quote
instead of a custom escaper. Addresses Copilot review on #5635.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* Style changes
---------
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Co-authored-by: David Heinemeier Hansson <david@hey.com>
* Install sof-firmware on Intel Panther Lake for DSP audio
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* Need to have the sof-firmware package available during original offline install
---------
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Co-authored-by: David Heinemeier Hansson <david@hey.com>
* Fix power profile auto-switching on USB-C only machines
* Small nits
* Stick with single unit name
---------
Co-authored-by: David Heinemeier Hansson <david@hey.com>
* Fix disable-while-typing on ASUS ROG Flow Z13 detachable keyboard
* Replace lsusb with DMI detection for Z13 keyboard
* Use omarchy-hw-match instead of custom z13_present function
* Add ASUS ExpertBook B9406 display and touchpad fixes for Panther Lake
* Drop B9406 migration
Per DHH's review: the migration's audience is essentially nobody.
Anyone running Omarchy on B9406 today is doing so via nomodeset
manually at every boot, not via a working unmodified install. Fresh
installs after this PR ships will pick up the fixes through the
hardware-match install hook; no migration needed.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* Disable PSR1 on XPS IPS and OLED. IPS uses PSR2
* Improve conditional flow
* Simplify the migration logic
* We don't actually need this
We are using += everywhere
---------
Co-authored-by: David Heinemeier Hansson <david@hey.com>
