1.2.0 release

[bourumir-wyngs]

This release is mostly about URDF parameter parsing.

[gavanderhoorn]

Nice feature.

Can you say anything about whether this depends on conventions in the URDF/XACRO, or is it mathematically robust against different approaches to modelling a particular robot's kinematic structure?

[bourumir-wyngs]

For all robots we tested, the joint offsets in URDF follow the clear pattern matching OPW parameters. There are ways to obfuscate this with fixed joints in between potentially representing offsets, or by describing the whole robot as rotated by unusual angle (say 45 degrees) but I have never seen such an URDF. It is not obvious how much it is a problem.

[gavanderhoorn]

ah, ok. Thanks.

if you've mostly looked at robots modelled by people in ROS-Industrial repositories I fear conventions made this work well.

We have/had a rather strict approach to modelling kinematic chains, which (by coincidence) aligned well with how the OPW paper does it.

(I've been one of the reviewers for many of those models)