G.A. vd. Hoorn

While I understand the reason for the change you propose, I'm not sure removing the check is something we necessarily want to do. > An alternative suggestion is to instead...

I think either @jrgnicho or @JeremyZoss ran into the IKFast-solving-for-whatever-you-give-it issue some time ago. Not sure what the context was anymore.

@TheDash wrote: > Before, the plugins would be loaded and limited by a defined arm_prefix parameter, which can only be set to one prefix obviously. The real problem is this:...

@gavanderhoorn wrote: > I think either @jrgnicho or @JeremyZoss ran into the IKFast-solving-for-whatever-you-give-it issue some time ago. Not sure what the context was anymore. Apparently that was @Jmeyer1292, he even...

> I can check that the group of joints contains all of the arm strings, which should fix all issues here. I just dont want it to check for prefixes,...

I'm really not happy about removing the prefix just because it's harder to get the configuration right. But: 1. I'm not a maintainer here, so ultimately it's up to @ipa-fxm...

@TheDash wrote: > [..] all it does is check/confirm that the chain contains links for a UR robot, which one can assume outside of this code that it is being...

> Ok - general concensus seems to be to add some kind of checking/warning to the code while keeping it generic that would be ideal, yes. But again, I'm just...

> Have you tried using multiple functional UR arms on a platform? This is where the issue stems from - since multiple loading of the move_group in the RViz node...

First: I agree that this is probably the easiest work-around, taking the effort required into account (but that doesn't mean I need to like it :) ). But: > However...