Jonathan Gammell

Should we rebase/flatten the commits in this PR a bit?

As I understand it, the goal of RRT*-Smart is the same as Inf. RRT* (to use sampling to improve the convergence rate of RRT* once a solution is found) but...

I actually spent yesterday looking at `RRTstar.cpp`, so I added some comments to the proposed changes. I hope you don't mind that I'm a little pedantic.

I meant create a new class derived from `InformedSampler`, just like was done for `PathLengthDirectInfSampler` and `RejectionInfSampler`. See: http://ompl.kavrakilab.org/classompl_1_1base_1_1InformedSampler.html A path-biased sampler wouldn't exactly meet the current description of `InformedSampler`,...

I don't think we are sitting on any patch-level changes at the moment. @marlinstrub would know best if we have anything, but he's very busy at the moment, so you...