Closed jhidalgocarrio closed 5 years ago
Hi @jhidalgocarrio. the branch also contains other unrelated things.
Generally speaking I would agree, but then how do we deal with backward compatibility ? Is that a problem that everyone has noticed ? I'm puzzled that you're the first to notice ...
Hi @jhidalgocarrio. the branch also contains other unrelated things.
My bad. Somehow I messed up with a merge from another open PR recently closed. It should be fixed by now.
Is that a problem that everyone has noticed ?
No problem so far. The code comes from iMoby times.
How do we deal with backward compatibility ?
My guess. It would not have compatibility with some old iMoby code, but I doubt this code is currently in used (I don't think it even compiles with the current base/types and base/logging).
IMO let be practical and implement the rock coordinate convention in base/types.
Something I don't get. I had a look at the code, and the arrow is actually created pointing towards Y and then rotated. The comment (that you removed) actually points out that the rotation is because of Rock's convention.
If I visualize a WP with a heading of zero I get this
A.k.a. what I would expect. Would you be mixing NWU / ENU conventions for the world frame ? Or some problems related to the world frame of your pose estimation ?
Ping
Given the state of the discussion, I'm closing this. Reopen if you have more comments.
For some reason the waypoint visualization is artificially rotated -90deg to point X to the right. In rock convention X is forward.