API review
Proposer: Blaise Gassend
Present at review:
- List reviewers
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.
Blaise
- Is there any use case in which the target frame is distinct from the image frame?
- BG: Haven't found one, but isn't hindering anything.
- Should the node be outputting a new camera_info, or is this just for visualization?
- BG: Changed output topic to rotated/image so that it will be easy to add camera_info at some point
Brian
(minor) I think that image should be a subscribed topic.
- BG: Fixed
Stu
- I would prefer target/frame_id, target/x, ...
- BG: Won't work with dynamic_reconfigure.
- I think the description of target_frame_id might not be correct
- BG: Renamed reference_* to source_* after discussion with Stu and Wim.
Patrick
Should only subscribe to image if something is subscribed to image_rotated, like what other image_pipeline nodes do.
- BG: Done.
You're using the frame_id in the image header, which is arguably dangerous. Some time ago Vijay, Jeremy and I hashed out some conventions on the exact meanings of frame_id in Image and CameraInfo. One conclusion we had was that image.header.frame_id doesn't have a well-defined meaning and that info.header.frame_id should always be used instead. We talked about having some arbitrary convention for image.header.frame_id but never actually implemented that in the drivers. So, I recommend using a CameraSubscriber and pulling the frame id out of the CameraInfo.
- BG: I added a use_camera_info option that defaults to true to make your request the default. I also added an input_frame_id option that allows explicit setting of the input frame.
I also slightly prefer target/x, etc. But I'm guessing that won't play nice with dynamic_reconfigure.
- BG: Won't work with dynamic_reconfigure.
Doc suggestions while I'm at it:
- In the package description state why this is useful, e.g. for viewing images from a camera with arbitrary motion.
- BG: Done in trunk.
Emphasize that image_rotated is intended only for visualization purposes.
- BG: Done in trunk. I assume you mean here that it would make no sense to use the rotated image for computation, correct?
The description of ~output_image_size makes it sound like an enum, so say something about interpolating for fractional values.
- BG: Done
Meeting agenda
To be filled out by proposer based on comments gathered during API review period
Conclusion
Package status change mark change manifest)
Action items that need to be taken.
Major issues that need to be resolved