You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This commit was created on GitHub.com and signed with GitHub’s verified signature.
Improvements
The timeline is now a map of your footage: each video is laid out whole, every clip sits over its own stretch of it, and what you cut stays on the timeline, dimmed. Trimming never rescales the timeline or moves other clips, split pieces pulled apart keep the gap between them, and dragging a handle back over dimmed footage restores it. Only zooming changes the scale (#61)
Omacut follows your Omarchy language, translated into 49 languages, with right-to-left layouts (#49, #50)
Corners follow your windows: sharp on Omarchy's default, rounded when your Hyprland theme rounds them, and live as that changes (#47)
Confirmations answer from the keyboard: the arrows and Tab move between the buttons, Enter presses one, Alt and a button's underlined letter picks it, and Escape backs out (#58)
Quitting while an export is running asks first (#48)
Moving a clip's end to the playhead is now Ctrl+Shift+Space, since Alt+Space is Omarchy's dictation key (#63)
Fixes
Ctrl+Space and Ctrl+Shift+Space keep the playhead on its frame, so pressing one twice no longer trims twice, and they no longer cancel a play that's waiting for the video to load (#56)
A trim handle stops at the next piece of the same video instead of running into footage that piece already shows (#61)
The end of the filmstrip shows the video's last frame instead of a blank (#52)
Pieces that play as one can be joined even when a drag leaves them a hair apart (#53)
The play button's tooltip no longer says Pause while the video is loading (#55)
A clip's × no longer stays up after a drag ends off the timeline (#57)
Omacut builds with CMake, and GitHub Actions runs the tests on every pull request (#46)
The tests run offscreen, fail on any QML warning, and include a randomized timeline test that checks every edit against a freshly opened editor, down to the pixel (#60, #61)