replaydiff

Replay the same ROS 2 recording (MCAP) through two versions of your code and find out whether the behavior changed.

replaydiff runs a recording through your stack at two git refs, records the outputs, compares them topic by topic against a per-topic tolerance file, and exits 0 (same behavior within tolerance), 1 (regression) or 2 (can't tell: the replay is too noisy for the tolerance you set, or something failed). It is built to run in CI on pull requests.

Status: design only. Nothing is implemented yet. The design is in docs/design.md.

Planned first release:

• Comparators for geometry_msgs/Twist, nav_msgs/Odometry, nav_msgs/Path, sensor_msgs/JointState and /tf.
• A YAML tolerance file with absolute/relative limits, matching modes and warmup windows.
• A measured noise floor (the base version replayed several times), so timing jitter can't pass as a regression and can't hide one either.
• A GitHub Action that posts the report on the pull request.

Replay and recording use replay_testing.

License

Apache License 2.0. See LICENSE.