UCD vocabulary revision : first iteration of the maintance procedure

[Markus Demleitner]

Hi Semantics,

On Mon, May 21, 2018 at 12:30:29AM +0200, Mireille Louys wrote:
> The Semantics has proposed a Maintenance procedure in oder to update the UCD
> list ( and more generaly ivoa vocabularies)
> as stated in the WD
> http://www.ivoa.net/documents/UCDlistMaintenance/20171018/
> 
>   S | pos.spherical | Related to cylindrical coordinates

This would have to be "spherical" instead of cylindrical.

>   Q | pos.spherical.r | Radial distance or radius (spherical coordinates)
>   Q | pos.spherical.azi | Azimuthal angle (spherical coordinates)
>   Q | pos.spherical.colat |Polar or Colatitude angle (spherical coordinates)

What makes me a bit queasy here is that we already have two
"instances" (if you will) of spherical coordinates in UCD, pos.eq and
pos.galactic.

In hindsight, it would probably have been better to have a single
set of pos.spherical terms and rely on external annotation (i.e.,
COORDSYS or, once it's there, STC+VODML-mapping) to convey the
reference frame.

Now that we have several 1e4 pos.eq annotations, I don't think we
should touch them.  But if all three sets of terms exist, we should
make clear that pos.eq and pos.galactic (together with phys.distance
or pos.parallax as applicable) is preferred over pos.spherical. I'm
not sure where to put language to do that.  Perhaps:

S|pos.eq                               |Equatorial coordinates (use in preference to pos.spherical where applicable)
S|pos.galactic                         |Galactic coordinates (use in preference to pos.spherical where applicable)


Also, pos.eq has both pos.eq.dec and pos.eq.spd.  Hence, I think
there should be both pos.spherical.colat and pos.spherical.lat (say)
here.

         -- Markus