inertia: Should explicitly state that no inertia == massless (inertialess) link?

[EricCousineau-TRI]

Motivated by this discussion: gazebosim/sdformat#199 (comment)

Empty links in URDF are assumed to be massless and get converted into SDFormat frames

To substantiate this, I wanted to find a first-hand source, but trying to find one directly in specs / code is a bit harder than it should be?
http://wiki.ros.org/urdf/XML/link

Walking through some various points that connect parsing to usage in KDL:

urdfdom/urdf_parser/src/link.cpp

Line 432 in 0da4b20

link.inertial.reset(new Inertial());

https://github.com/ros/urdfdom_headers/blob/e31d66ea4a1f753fcca05e2d046b3af20a3234b3/urdf_model/include/urdf_model/link.h#L196
https://github.com/ros/kdl_parser/blob/24597c21301c53b66a5e0ae22619210261f1a28f/kdl_parser/src/kdl_parser.cpp#L147-L151

I've stopped my search, but was going to dig into KDL itself:
https://github.com/orocos/orocos_kinematics_dynamics/tree/master/orocos_kdl

Ultimately, it seems like "optional inertial" all the down to KDL means "zero-mass + zero-inertia", which is fine and sounds obvious if you say it out loud, but seems like it could be ambiguous.

It would be lovely to seethe XML page for it to look like:

inertial: (optional: defaults to zero mass and zero inertia)

[EricCousineau-TRI]

Edited Wiki page:
http://wiki.ros.org/action/info/urdf/XML/link?action=diff&rev2=70&rev1=69

@scpeters When you have a chance, can you confirm this looks OK? If so, then we can close this.