Remove boost::math dependency

traversaro commented 4 years ago

As discussed in https://github.com/ros-industrial/abb_libegm/issues/70#issuecomment-584000173 :

Remove the boost::math dependency by implementing the quaternion conjugate operation, which should be straight-forward.

It turns out that it was necessary also to implement a couple more operators, so the PR is ready but it is mostly open just to receive feedback, in particular:

I added a test to ensure that the estimateVelocity continued to work exactly as it did with boost::math . For the time being I did not use any test library, if you have any preference I can migrate the test to your favored framework.
To ensure that the code remained real-time safe with no allocation as it was before, I could not define functions that operated on the abb::egm::wrapper::Quaternion structure, but I had to introduce a small private class to contain internal quaternion operations. I intentionally hided this object from the public headers to avoid to maintain them in the future, let me know if this is ok.

gavanderhoorn commented 3 years ago

@jontje: would this be something you could check and review?

Would be great to get this merged.

Sorry @traversaro :S

traversaro commented 3 years ago

Sorry @traversaro :S

No problem, I know how it works with open source software maintenance. : )

The fact that ABB's even accepts downstream contributions for their EGM reference implementation is still mind-blowing w.r.t. to the kind of support/openness provided by some other industrial robotics vendors as of 2020 (we can always hope this will change in the future : ) ).

gavanderhoorn commented 3 years ago

The fact that ABB's even accepts downstream contributions for their EGM reference implementation is still mind-blowing

It's just a trick: whenever something doesn't work, we check git blame and then forward complaints to whoever contributed that change ;D

(just kidding of course, I don't work for ABB)

jontje commented 3 years ago

@jontje: would this be something you could check and review?

Would be great to get this merged.

I agree it would be great, and I have been meaning to get around to this. I will try to do it as soon as possible.

Sorry @traversaro :S

@traversaro, I am also sorry for the slow response rate 😔

traversaro commented 3 years ago

I saw that the PR has no some conflicts, I could fix them know but perhaps it could make sense to wait for a first review, to avoid solving other different conflicts in the future? In any case, the conflicts are minor and are not part of the core of the PR.

traversaro commented 1 year ago

I saw that the PR has no some conflicts, I could fix them know but perhaps it could make sense to wait for a first review, to avoid solving other different conflicts in the future? In any case, the conflicts are minor and are not part of the core of the PR.

I am trying to cleanup the PRs that I have in https://github.com/pulls, so given that this seems stale, I am going to close this. If anyone is interested in this in the future, feel free to ping me and I would be happy to rebase it on the top of the latest master, thanks!

ros-industrial / abb_libegm

Remove boost::math dependency #103