Closed KangarooKoala closed 2 months ago
I think removing is the right choice since that is what was done in #6673.
Once we decide which of removing and deprecating we want to do, this would be a great first PR for somebody looking to start contributing to WPILib!
I'm interesting in taking this on
I think removing is the right choice since that is what was done in #6673.
I'm not sure I get what you mean by this- #6673 removed all usages of WheelPositions
in our kinematics, odometry, and pose estimator classes, but it didn't touch WheelPositions
itself.
That said, WheelPositions
should only be used by kinematics implementations for custom drivebase types, so it's probably fine to remove it without warning (like we did to SwerveDriveWheelPositions
). @calcmogul what do you think?
I'm interesting in taking this on
Great! Looks you've already opened a wpilib PR, but let us know if you have any questions! (e.g., how to run formatting and compile locally)
Ah - I misunderstood the scope of the change. I understand what you are looking for now.
Yea, WheelPositions
was only used by us, so it can be removed now that nothing references it.
6673 migrated
Kinematics
,Odometry
, andPoseEstimator
to not use theWheelPositions
interface (Java)/concept (C++). The only remaining uses are as superinterfaces ofDifferentialDriveWheelPositions
andMecanumDriveWheelPositions
. We probably could completely remove the interface/concept since it should only impact kinematics/odometry/pose estimator implementations for custom drivetrain types who would be impacted by #6673 already, but I can also see the case for deprecating first for a year.Once we decide which of removing and deprecating we want to do, this would be a great first PR for somebody looking to start contributing to WPILib!