Proposer: Sachin Chitta
Present at review:
- List reviewers
What to review
This is a review of the stack layout and its ROS API. The stack layout is presented on the main stack wiki page here while the ROS API is in two parts:
- yaml interface for setting up filters
ROS service interface for calling the filters using the FilterJointTrajectory message.
For more details on the yaml interface, please see the tutorials for an example. The service definition for the ROS service call is in the motion_planning_msgs package.
- stack.xml description needs to be filled in
- why do you specify joint limits? This should be determined using the robot model. If you want to specify artificial limits then there should be a trajecotyfilterwithlimits or something so that people don't need to constantly fill in this yaml and possibly get it wrong.
- the yaml for making the filter chain is the same as the filters library but what does the config for the individual filters look like if there is one and where is it documented?
- in general a lot more documentation is needed on what the filters do, how they work, and math behind them
Question / concerns / comments
Enter your thoughts on the API and any questions / concerns you have here. Please sign your name. Anything you want to address in the API review should be marked down here before the start of the meeting.
To be filled out by proposer based on comments gathered during API review period
Stack status change mark change manifest)
Action items that need to be taken.
Major issues that need to be resolved
- Rename trajectory_filter to trajectory_filter_server - DONE
- Default behavior loads position limits from the urdf but takes up additional limits from the yaml file.
- Put in more documentation for each individual filter
- Look into templating for filters - also need two different nodes, one for collision aware filters and one for filters - DONE
- trajectory_unnormalizer - goes into joint_normalization_filters