Improve JTA goal validation to also check for duplicate joint names about motoros2

[ted-miller]

Fix is in #162

[gavanderhoorn]

In order to not delay releasing 0.1.1, I propose we move this to the 0.1.2 milestone.

@ted-miller: thoughts?

[ted-miller]

Yeah, I'm fine with bumping this out.

Additionally, this type of feature would add more overhead to the message processing. I've been getting a lot of complaints lately about latency. So we should decide if this is really important. Personally, I don't think it's super needed. I don't envision a motion-planner outputting a joint that contradicts itself.

[gavanderhoorn]

I've been getting a lot of complaints lately about latency.

in MotoROS2?

I don't envision a motion-planner outputting a joint that contradicts itself.

it's really not that difficult to misconfigure things.

If one additional loop over 6 const char*s is where things become too slow, we probably have bigger problems.

[ted-miller]

in MotoROS2?

No. Just in general.
But we already have an issue with the publish rates in MotoROS2. We don't know where that bottleneck is yet.

If one additional loop over 6 const char*s is where things become too slow, we probably have bigger problems.

Well it'd be more than one loop... but your point is fair.

it's really not that difficult to misconfigure things.

Ok. We can keep it. It shouldn't be hard to implement. I'm just pretty bogged down until FabTech, otherwise I'd have done it already.

[gavanderhoorn]

@ted-miller: would this be something we include in 0.1.2? I'm thinking we should not delay releasing it too long, as it includes fixes for things like #144 and #55 which would be nice to release.