Repository navigation
Merge upstream quattro (screenshot chord, NVIDIA apps, ARM recovery) - #81
Merged
Merged
Conversation
* Take a screenshot with Super + Shift + S Print Screen is the only screenshot key, and Apple and Logitech keyboards do not have one. Super + Shift + S is the chord Windows uses for the same job, and it now takes the identical shot. That chord opened Google Maps, which moves one modifier over to Super + Shift + Alt + S, alongside the other secondary web apps. No tiling binding changes. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01HhEFSASvbeUD1AHAEATwCg * Retire uncommented copies of the old Super + Shift + S example The bindings.lua template offered Super + Shift + S -> screenshot as its Logitech MX Keys example. Now that the same binding ships as a default, an uncommented copy in a user's bindings.lua stacks a second bind on the chord, and the two screenshot instances race each other's picker. Comment the known example back out on existing installs; the chord rebound to any other command is the user's own and stays untouched. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01HhEFSASvbeUD1AHAEATwCg * Let a personal Super + Shift + S binding win, and keep symlinks intact Two problems with the migration, both raised in review. It rewrote the user's own binding. Commenting out an active bind takes away the only screenshot shortcut anyone running with `omarchy_default_bindings = false` has, and it only recognised the chord when it still ran the screenshot command — a chord rebound to something else was left to stack with the new default, so both fired. Unbinding first is the pattern this repo documents everywhere else, and it fixes both cases at once: the user's line stays exactly as they wrote it and simply becomes the only one that runs, whatever they bound to it. It also detached symlinked configs. `sed -i` replaces a symlink with a regular file, and a dotfile-managed `bindings.lua` is usually a link into a repo — the migration would have quietly cut the config off from the source the user actually edits. `--follow-symlinks` writes through the link, as the tmux-alert migration already does. The tests now cover the kept binding, a chord rebound to another command, an existing unbind (not duplicated, and idempotent across reruns), the commented-out template, unrelated bindings, and a symlinked config whose edit has to land in the repo it points at. * Run the migration guards the way omarchy-migrate does omarchy-migrate executes migrations with `bash -euo pipefail`; it does not source them. The top-level `return 0` guards die there ("can only `return' from a function or sourced script"), so every user without a Super + Shift + S binding of their own -- almost everyone -- had the migration fail and rerun on each update. The guards are one conditional now, and the test invokes the script exactly like the runner instead of sourcing it, which is how this slipped through. The unbind also only recognised o.bind. A chord taken with o.bind_toggle or bare hl.bind stacks with the new default just the same, so all three forms count now. * Drop the Super + Shift + S bindings.lua migration The package swaps the default binding files on update, so users' own bindings.lua doesn't need rewriting. * Keep the current Omasnap description in the screenshots manual * Drop the screenshot binding test * Drop the screenshot binding comments and the Mac keys claim --------- Co-authored-by: Claude <noreply@anthropic.com> Co-authored-by: TripleU613 <tripleutech@gmail.com> Co-authored-by: Ryan Hughes <ryan@heyoodle.com>
|
Important Review skippedBot user detected. To trigger a single review, invoke the ⚙️ Run configuration
You can disable this status message by setting the Use the checkbox below for a quick retry:
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Cherry-picks upstream omacom/omarchy
quattroonto this fork'squattro(2da6fb23, which already has #80). Upstream moved from8a787b6ato33783b17. 3 commits are applied oldest-first, one fork commit per upstream commit. The fork-only file set is still 167 files.Screenshot chord (omacom#8827)
Super + Shift + Stakes the same screenshot asPrint Screen. Apple and Logitech keyboards have no Print Screen key, and this is the chord Windows uses for the same job. Google Maps moves one modifier over, toSuper + Shift + Alt + S, next to the other secondary web apps. No tiling binding changes.The bindings template's commented Logitech MX Keys example of that chord is gone, so an uncommented copy cannot race the new default. A personal
bindings.luais left alone: the package replaces the default binding files on update, and a chord the user rebound stays theirs. The manual names both keys for the shot and the new Maps chord.Installed NVIDIA applications (omacom#12723)
Hardware setup now runs
install/hardware/nvidia-apps.shafter the driver script. Whennvidia-ai-workbenchis already installed,nvidia-ctk runtime configure --runtime=dockerwrites the Docker runtime into omarchy-settings'daemon.json(later defaults arrive as.pacnew). Whendgx-dashboardis already installed, its service is enabled, and a deferred first boot starts it without blocking. A refused start does not fail that boot. Ollama's own service management is left alone. Nothing here installs the applications; the Spark profile already does that.ARM recovery packages (omacom#12722)
Migration
1789928586marks an installed Limine recovery stack explicit so orphan cleanup keeps it:limine,limine-mkinitcpio-hook,limine-snapper-sync, andsnapper. It runs only on aarch64 when a Limine config is present, and it only retags packages that are already installed as dependencies. x86 is unchanged, a missing Limine config is a no-op, and a failed package-database write leaves the migration pending. A partial stack marks only the packages that are there.Fork-only lines kept
These commits do not touch the fork overlay. Setup/Remove > Security > Face, and Update > Grok Bot / Cursor, stay, including in the English menu catalog. Cursor is still installed from Cursor's official tarball, not through mise.
omarchy-theme-set-vscodestill does not theme Cursor.omarchy-powerprofiles-setstill writesActiveProfilethroughbusctl, and the menupower-profilesprovider stays out.Tests
test/shell.d/nvidia-apps-test.shandtest/shell.d/arm-recovery-migration-test.shpass. The NVIDIA test printspop_var_contextnoise from its exported setup function and still exits 0.