search

tesseract-robotics / tesseract_ignition

https://tesseract-ignition.readthedocs.io
7 stars 5 forks source link

Comparing SRDFs #8

Open mpowelson opened 2 years ago

mpowelson commented 2 years ago

It would be nice to just see what changed in a concise way. This could be a tool that shows you, but regardless we should probably use the same SRDF ordering as the moveit setup assistant. It is currently difficult to compare.

Levi-Armstrong commented 2 years ago

Can you provide more details on what you are looking for or should change?

mpowelson commented 2 years ago

The allowed collisions are in seemingly random order. So if you use this to regenerate an srdf created with moveit (or this too), the diff is incomprehensible. If we change joint limits and want to rerun this it is difficult to see what actually changed and make sure manually added pairs don't get lost.

Levi-Armstrong commented 2 years ago

Should they always be sorted?

mpowelson commented 2 years ago

Being consistently sorted would fix the problem. Then the diff would make sense. Ideally I think they would be sorted like the moveit srdf. I assume that is alphabetical by first then second, but I'd have to double check.

mpowelson commented 2 years ago

I looked into this today. It looks like we just need to sort the output in Tesseract_srdf bool SRDFModel::saveToFile(const std::string& file_path) const