Skip to content

kinematics modules / switchKins #5

Open

Description

@jallwine

What is the state of kinematics modules in EMCApplication? It looks like significant effort was put into extracting the trajectory planner and kinematics modules from the motion module in MachineKit. How were these changes merged into EMCApplication (or were they backed out)?

Specifically, I'd like to have a better understanding to be able to experiment with rebasing the dgarr/s_kins branch from LinuxCNC that allows switching kinematics modes from within a G code program.

Activity

  1. zultron commented on Oct 6, 2020

    @zultron
    Contributor

    Well, the Loadable Trajectory Modules work remains in the legacy Machinekit repo, and hasn't been merged into any LCNC mainline branch. If someone wants to save that work, best is to pull it out and submit a new PR against LCNC.

    The EMCApplication repo is a fork of a recent LCNC master branch with only minimal build system changes applied for building against an external HAL. As such, the Loadable Trajectory Modules work isn't in EMCApplication at all.

    As for the dgarr/s_kins branch, most likely you could rebase it onto EMCApplication as easily as you could rebase onto LCNC master. The Loadable Trajectory Modules work won't get in the way, of course, since it's not in that tree.

  2. jallwine commented on Oct 6, 2020

    @jallwine
    Author

    Gotcha, I just noticed a vtable include with some extra types defined that conflict during the rebase: https://github.com/machinekit/EMCApplication/blob/machinekit/master/src/emc/kinematics/kinematics.h

    If those aren't used at all, that should be straight forward to resolve. Thanks!

  3. zultron commented on Oct 6, 2020

    @zultron
    Contributor

    Gotcha, I just noticed a vtable include with some extra types defined that conflict during the rebase: https://github.com/machinekit/EMCApplication/blob/machinekit/master/src/emc/kinematics/kinematics.h

    If those aren't used at all, that should be straight forward to resolve. Thanks!

    Oh yes, I'm sorry, there were just a few small changes to the C code, but one of them was the vtables change, exactly what you've run into.

    I'm weak on that area of the code, but I remember those changes aren't rocket science. You can get a handle of exactly what's different by looking at PR #4, merged today: git diff -w 15211fd4 d0d1f2fe

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions