Joint positions might incorrect after RESET

[dmronga]

One more (:
When I do a RESET, the joint positions might be flipped by 2pi. The problem occurs when I reset the simulation while one or more of the joints are located at position >pi or <-pi.
This is due to the absolute position conversion done here:

simulation-orogen-mars/tasks/Joints.hpp

Line 43 in a6ea2a4

double updateAbsolutePosition( double v )

The position is shifted by 2pi if the absolute difference to the last position is bigger than pi.
Is there a good reason for doing such a conversion? If not I would suggest to remove it.

[planthaber]

afaik this was done because the motors in the simulation do not support multi-rotation for position controls, so in normal operation the command values need to be between -pi and pi

[malter]

The simulation on the develop branch handles the joint position correctly, even for multi-turn motors. I also added a develop branch for the orogen mars component where the 2pi conversion is removed. It would be great if you can test the develop branch of mars and orogen/mars to check if the simulation behaves correct for you?
We will merge the develop branch into the master until we tested all new features a bit more. ;-)

[dmronga]

I finally had time to test this feature on the develop branch, seems ok to me

[dmronga]

Can you merge this feature to master?

[malter]

It should already be in the master branch since 10th of January ;-)

[dmronga]

Hmm, from what I see it is not. See here. Compared to that the develop branch looks like this. It is just that I recently switched to master and now I have the same issues as before, that sometimes joints are reset to the wrong position...

[malter]

Ahh, I see, I only checked the branch of MARS not of the orogen component ;-)

[malter]

Merged :-)