Repository navigation
v3.0.0
A responsive reader for audio and transcripts
- Scroll audio and Markdown together with the mouse wheel, arrows, Page Up/Down, and Home/End. Home returns to the waveform; T jumps to the transcript without stopping audio.
- Long documents use indexed wrapping, bounded screen output, and deduplicated frames. Fast keyboard input is no longer discarded; paused views stop repainting.
- Tail cuts and completed playback keep the player open. SPACE replays completed audio.
- Stopping an interactive recording opens playback immediately. Transcribe and format open the Markdown reader;
rec view path.mdreopens saved notes. - Copy the complete Markdown with Y or
rec share. ChatGPT, Claude, and Gemini destinations copy the document and open the website for pasting; nothing is automatically submitted. - The whole playback waveform stays colored, with selection anchors and a visible playhead.
Breaking change
Transcript bodies now preserve each utterance as [HH:MM:SS–HH:MM:SS] Speaker N: text instead of merging turns into prose-only paragraphs. Update tools that parse the previous body format. Frontmatter remains available and formatting templates receive the timestamps.
Validation
Unit tests on macOS, Linux, and Windows; five cross-platform builds; and a PTY regression suite covering 1,000 synthetic paragraphs, scrolling, mouse events, key bursts, resize, idle output, clipboard capture, and tail cuts.
Update with rec update.
Full changelog: v2.0.0...v3.0.0