Improve failure reporting

[rhaschke]

I had to debug a MoveRelative stage that was failing for no visible reason. We should improve failure reporting by:

• showing the failing pose (this is currently pruned away in Cartesian and joint interpolation planners)
• naming and highlighting collision pairs if a collision is the reason for a failure
• naming and highlighting joint limit violations

This PR is just a (tiny) starting point into that direction. It (partially) served me for debugging my problem. However, this is not generally applicable yet: The achieved_fraction returned by the Cartesian interpolation methods is not the fraction we actually need to evaluate for the stage... I'm tempted to drop the min_fraction parameter from the planners as this should be a stage property only.

[codecov[bot]]

Codecov Report

Merging #306 (50f364a) into master (d6f68f9) will decrease coverage by 0.02%. The diff coverage is 55.56%.

@@            Coverage Diff             @@
##           master     #306      +/-   ##
==========================================
- Coverage   51.85%   51.84%   -0.01%     
==========================================
  Files         101      101              
  Lines        7561     7567       +6     
==========================================
+ Hits         3920     3922       +2     
- Misses       3641     3645       +4     

Impacted Files	Coverage Δ
core/src/solvers/cartesian_path.cpp	76.75% <55.56%> (-7.03%)	:arrow_down:

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update d6f68f9...50f364a. Read the comment docs.

[rhaschke]

I'm tempted to drop the min_fraction parameter from the planners as this should be a stage property only.

@v4hn, any comments on this?