Update README with information on how it works

[EzraBrooks]

Closes #5.

@traversaro is this the sort of information you were looking for?

[traversaro]

Closes #5.

@traversaro is this the sort of information you were looking for?

Yes! I suggest something more but more and less it is what I needed.

[EzraBrooks]

I have some ideas cooking along the lines of formal definition of an interface (I saw your fingerprints on the same over in ros/urdfdom#200 and related issues), but I'm trying to clean up the upstream schema before making one that derives from it. (ros/urdfdom#227, ros/urdfdom#228, ros/urdfdom#229)

[traversaro]

Yes, I think what would also useful at urdf level is the formalization of of a uri schema for relative paths w.r.t. to the urdf file location. That is already quite useful and used by many existing URDF parser, but it is not formalized now, and I do not if the official ROS tooling actually process it.

fyi @saikishor

[traversaro]

Apparently I added a suggestion but it remained in pending mode (that it was the "more" I was suggesting in my comment, sorry for the confusion. @EzraBrooks

[traversaro]

Suggested change

	URDFZ treats your URDF as an archive manifest for your URDFZ archive. It looks up all the URIs listed in the URDF, resolves them to absolute paths, and copies them into the archive. It then creates a copy of your URDF that has its URIs rewritten to a `urdfz://` scheme that points to the asset's path within the archive, and places that modified URDF at the root of the archive.
	URDFZ treats your URDF as an archive manifest for your URDFZ archive. It looks up all the URIs listed in the URDF, resolves them to absolute paths, and copies them into the archive. It then creates a copy of your URDF that has its URIs rewritten to a `urdfz://` scheme that points to the asset's path within the archive, and places that modified URDF at the root of the archive.
	Once the URDFZ is extracted, the `urdfz://` URIs are replaced with a `file://` containing the absolute path to the extracted resource file.

I am not sure if what I added is correct, but it would be great to describe what happens at extraction phase.

[saikishor]

Yes, I think what would also useful at urdf level is the formalization of of a uri schema for relative paths w.r.t. to the urdf file location. That is already quite useful and used by many existing URDF parser, but it is not formalized now, and I do not if the official ROS tooling actually process it.

fyi @saikishor

Good to know. We should definitely tackle this ;)