SSA UTYPES

[Doug Tody]

Hi Alberto -

On Tue, 19 Dec 2006, Alberto Micol wrote:
> The paragraph 3.3.6 Query Response Metadata of the SSA0.97 document states
> that the leading "spectrum." in the UTYPE for a Spectrum data model attribute
> should be omitted, the reason being that SSA metadata do not need to be 
> specific to spectra.
>
> In such case, shouldn't SSA have its own model/name space?
> I do not think we want to have UTPYEs not associated to a model, do we?

The motivation for this is that we would like to use most of the
same metadata for other DAL interfaces (at this level very little is
specific to spectra), plus at this level we prefer to have the UTYPEs
rooted in each component data model.  Hence, a UTYPE is something like
"ssa:Target.Name".

As for SSA needing its own data model and namespace, probably it does
anyway, as at level of the protocol additional metadata needs to be
defined which is not relevant for Spectrum.  But the intention is
that most of the metadata in SSA and Spectrum shares a common model.

I like to make a distinction between a data model as an abstraction, and
an application of a data model in some context.  Two different applications
can share the same data model, without necessarily sharing something in
the implementation such as an XML schema.


> To recap, depending on whether if I follow :
>
> 1. the statement in SSA 3.3.6,
> 2. the SSA Appendix,
> 3. the SDM v1 rel cand 1 rev 2
>
> I can come up with three (or maybe four) different utypes for
> the same field:
 	[...]
> Could you please clarify?
>
> Thanks,
>
> Alberto
> PS: BTW, I support Randy's comments that the examples given in the Appendixes
> are really important to the developers. It would be so nice to have them
> fully thought through, so to act as approved references that one could use
> and adapt to his/her case... It would speed up development!

As noted earlier, we merely haven't updated the Appendix yet.
The UTYPEs will be as stated in the earlier sections of the document.

I strongly agree that an example VOTable such as this (once updated)
is an important part of the specification.  We are working on some
actual SSA implementations already (also a Java toolkit which will
deal with most of this!), which will also help to clarify all this.

 	- Doug