SODA1.0 WD release

[Mark Taylor]

On Mon, 11 Jan 2016, François Bonnarel wrote:

> On 08/01/2016 23:39, Mark Taylor wrote:
>
> > arraysize values:
> >     sec 3.1.1: "In addition ... its datatype 'char', with arraysize '*'".
> >     I don't like mandating arraysize as "*".  There's no reason it
> >     shouldn't have an arraysize which is (e.g.) the exact length
> >     of the string.  Also, mandating datatype of char may come back
> >     to bite us if we do something datatype-dependent to accommodate
> >     unicode characters in the future.  I'd be inclined just to say
> >     it's a character array.  Similar remarks apply to POL (sec 3.2.4),
> >     which could reasonably have arraysize="1".
> I am waiting for additional advices on this
>
> > Finally (at least for now), it's not obvious to me from this document
> > how to actually use a SODA service.  Possibly that's because I'm
> > not familiar enough with Datalink or other associated standards,
> > but I may not be the only one...  Presumably (in view of the
> > language about xtypes, datatypes and arraysizes) you have to present
> > a VOTable to the service including the relevant parameters, but
> > it's not clear to me how to go about this.  Can you add a short
> > example section illustrating how you'd present a query to a SODA
> > service?
> Hummm... The basic usage is not by presenting a VOTABLE but by using an URL
> with the standard parameters, as HTTP parameters. The description with VOTABLE
> attributes is for the "service descriptor", as defined in DataLink spec, which
> is indeed a VOTABLE, and can help clients to create their interfaces on the
> fly.
> An exemple section is surely needed. Moreover, we plan to have an IVOA
> implementation note for combining ObstAP or SIAV2, DataLink and SODA in the
> general perspective of "Data cube discovery, description and access"

OK ... good that it's just application/x-www-form-urlencoded for
query submission, that's simple and well understood.

But I think this underlines my concerns about the detailed mandating
of the VOTable PARAM attributes (arraysize, datatype, xtype).
Since most of that detail either cannot or should not be enforced
by the SODA query processing implementations, then I don't see
much benefit in writing these prescriptions into the standard -
it just looks like providing extra ways for the service descriptor
to be wrong.  But if I'm missing the point, by all means put me right...

Mark

--
Mark Taylor   Astronomical Programmer   Physics, Bristol University, UK
m.b.taylor at bris.ac.uk +44-117-9288776  http://www.star.bris.ac.uk/~mbt/