Hi,

I have a really generic ROS design question regarding the ROS interface for our arm / hand.

My current implementation is (well if I simplify a bit):
 - a publisher that periodically publishes all the joints informations on ONE topic
 - a subscriber that subscribes to my own msg which takes a vector of joint_names + associated targets, loop on this vector, looks up the name in the hash map and send the target to the relevant joint in our hardware.

When I look at other interfaces, it often seems to be:
 - a vector of publishers publishing data, one publisher per joint
 - a vector of subscribers for the commands, one subscriber per joint

I can see the advantage of having one topic per motor as it is then easy to send a command to a couple of joints, ...

I just wanted to know if there's a big overhead having lots of publishers / subscribers compared to having one. If you have any good advices regarding this design, I'm all ears :)

Cheers,

Ugo

--
Ugo Cupcic | Shadow Robot Company | ugo@shadowrobot.com
Software Engineer | 251 Liverpool Road |
need a Hand? | London N1 1LX | +44 20 7700 2487
http://www.shadowrobot.com/hand/ @shadowrobot