VOTable 1.2

[Mark Taylor]

On Tue, 3 Mar 2009, Francois Ochsenbein wrote:

> >   1. Why change INFO syntax at all?  There may be good reasons,
> >      but I don't recall this proposal, or the motivation for it,
> >      being discussed on the VOTable list.
>
> ==> That was the result if a discussion in an IVOA meeting,
>     effectively not discussed on the mailing list

I see.  This doesn't answer my question - what was the motivation?

In any case, I would like to request that in future such changes are
discussed, or at least presented, on the mailing list.  Even for 
interested parties who attend IVOA Interops, they may be unable 
to attend VOTable sessions.  Airing such issues on the mailing list 
both gives all interested parties a chance to contribute, and 
provides a record of any views which were expressed.

> >   2. If INFO is only for strings, why does it have a 'unit' attribute?
>
> ===> Being made similar to PARAM, there is no reason to forbid the
>      unit attribute -- after all strings may also represent things
>      with attached units, like time intervals.

Please give an example.  If you're talking about ISO-8601 strings,
I don't think a unit is appropriate, since the units information such
as it is is implicit in the format.

More to the point - surely units only make sense for quantities.
If you want INFO to be able to represent quantities, surely it would
make more sense to go the whole way and allow it to represent 
numeric values, which presumably means give it identical syntax to 
PARAM.  I'm not necessarily saying that having two elements with 
identical syntax and different usage is a good idea, I just don't 
understand the thinking behind the existing proposal.

> >and another one I've just thought of
> >
> >   3. If INFO is only for strings, why can it contain a VALUES element?
>
> ===> same as 2. above -- could be useful
>
> Do these modifications prepresent problems for STILTS ?

Not insuperable problems, but changes in the syntax of an element
between different VOTable versions complicates the implementation.
If there is some good reason for the change, I don't mind making the
additional effort.  But if it just seemed like a nice idea to the
handful of people at a VOTable session a few interops ago and
nobody can remember why, I'm not so enthusiastic.

Mark

-- 
Mark Taylor   Astronomical Programmer   Physics, Bristol University, UK
m.b.taylor at bris.ac.uk +44-117-928-8776 http://www.star.bris.ac.uk/~mbt/