Skip to content

CHANGELOG.md

Latest commit

 

History

History
83 lines (61 loc) · 5.69 KB

CHANGELOG.md

File metadata and controls

83 lines (61 loc) · 5.69 KB
Raw

Changelog

All notable changes to decibri-cli will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

Unreleased

Fixed

  • decibri play no longer hangs if the output device is lost mid-playback or when Ctrl+C arrives while playback backpressure is applied; it now stops promptly and, on device loss, exits 4 with a clear message.

0.2.1 - 2026-07-11

Changed

  • Release plumbing only: republish 0.2.0's content under 0.2.1 so the crate and npm packages ship together through trusted publishing. No functional change from 0.2.0.

0.2.0 - 2026-07-11

Added

  • --device-id flag on capture and play: select a device by its stable per-host ID (exact match). Mutually exclusive with --device.
  • Version-parity test (tests/version_parity.rs): cargo test fails if the crate version and the npm wrapper version diverge.
  • Contributor License Agreement section in CONTRIBUTING.md describing the CLA check that runs on every pull request.
  • id field in devices --json output: the stable per-host device ID (WASAPI endpoint ID on Windows, CoreAudio UID on macOS, ALSA pcm_id on Linux). Empty string when the host cannot assign one.

Changed

  • decibri library updated from 3.0.0 to 5.0.0.
  • Capture output now honors --rate on every device: the device is opened at its native rate and the library resamples to the requested rate, so a 48 kHz microphone recorded with --rate 16000 produces a correct 16 kHz WAV.
  • --channels on capture accepts only 1; capture is mono only. Any other value is rejected as an argument error (exit 2).
  • dropped_chunks in the capture --json completion payload reports the real count of capture buffers dropped while the writer could not keep up (previously always 0). A nonzero count also prints a stderr warning. Capture completes rather than aborting when the writer falls behind.
  • Device loss is reported with the underlying failure from the library: capture exits 4 with the partial recording preserved, and playback cut short by an output-device failure exits 4 instead of reporting success.
  • Minimum supported Rust version raised to 1.88.
  • Documentation corrected for 0.2.0 behavior (versions, mono-only capture, --device-id, resampled capture pipeline) and cleaned of roadmap and tool-comparison content.
  • Release verification instructions corrected: provenance attestations cover the extracted decibri binary, not the release archive.

Removed

  • The writer-lag watchdog that stopped capture with exit 4 when buffering exceeded about 16 seconds. The library now bounds its capture buffer and drops the newest audio when the consumer stalls; the drop count is reported via dropped_chunks.

0.1.0 - 2026-04-12

Stable release of the same feature set as 0.1.0-alpha.1. No functional changes since alpha; shipped to the default latest npm tag after alpha round-trip testing succeeded on Windows, macOS, and Linux.

0.1.0-alpha.1 - 2026-04-12

Added

  • decibri version subcommand with human-readable and JSON output modes. The version --json schema is locked at v0.1.0: {decibri_cli, decibri, audio_backend, target, rust_version}.
  • decibri devices subcommand listing audio input and output devices as a table or as JSON. Supports --input, --output, and --json flags.
  • decibri capture subcommand for WAV recording from a microphone.
    • Flags: --output, --duration, --rate, --channels, --device.
    • Device selection by case-insensitive name substring or numeric index.
    • Default configuration is 16000 Hz mono (voice/ASR preset); --rate 44100 --channels 2 is the music preset.
    • Duration accepts bare seconds (5, 10.5) or humantime strings (10s, 1m30s).
    • Records until the specified duration or until Ctrl+C.
    • Ctrl+C produces a valid truncated WAV, not a corrupt file.
    • Watchdog protection: if the disk writer falls more than ~16 seconds behind, capture stops cleanly with exit 4 and the partial recording is preserved.
    • Clean device-unplug handling: loss mid-capture exits 4 with a partial WAV.
    • Output format is always 16-bit PCM WAV (universally compatible).
  • decibri play subcommand for WAV file playback.
    • Flags: <FILE> (positional), --device.
    • Supports 16-bit PCM int and 32-bit float WAV inputs.
    • Unsupported formats (24-bit, 8-bit, non-PCM codecs) exit 1 with a clear error.
    • Ctrl+C mid-playback exits 0 with "interrupted": true in JSON output.
  • Global flags --json and --quiet on all subcommands.
  • Hidden decibri completions <shell> plumbing via clap_complete. The subcommand is wired but not exposed in help output.
  • Documented exit code table: 0 success, 1 generic error, 2 invalid arguments, 3 device not found, 4 IO error.
  • Cross-platform release pipeline building Windows x86_64, Linux x86_64, Linux aarch64, macOS x86_64, macOS aarch64, and a macOS universal2 binary.
  • npm distribution via npm install -g decibri-cli. The postinstall script downloads the platform binary from the matching GitHub Release, verifies its SHA256 against the release manifest, and places it on the user's PATH.
  • SLSA provenance attestations on every release binary via GitHub Actions.
  • SHA256SUMS manifest attached to every release for integrity verification.