IVOA Footprint Service Protocol

[Rob Seaman]

Hi François,

As someone with no pony in this race, I might be bold enough to comment:

>     Sorry , I contest that 1 is "in the IVOA standard" and 3 is  
> not. We are definitly using VO standards and Technology for a  
> footprint interoperable APPLICATION.
>            1)  The general VOTABLE container is a standard
>             2) SIA ans SSA to which our footprints can be attached  
> are standard VO technology
>             3) the extension mechanism which is used for this  
> attchment is described (shortly indeed but anyway described) in the  
> SSA recommendation
>             4) the utype concept and attribute is described in the  
> VOTABLE standard. It is extensivly used and referred by SSA, SDM  
> and Char standards.
>              5) there is a Note by Jonathan suggesting how STC  
> utypes could be written and a Note by F Ochsenbein about  
> integartion of STC Coord systems abd utypes in catalog votables.
>              6) It is true that none of these have the satus of a  
> recommendation, but it is the same for all the utypes, simply  
> because the utype discussion is not OVER!

By this description there is no need for a VOEvent standard, for  
example, since we rely on an even longer list of previous (and  
subsequent) IVOA standards.  Rather, the question is what role the  
new standard (or lack of one) plays in the broader architecture.  The  
sense of last week's NVO meeting was that footprints are a key enough  
technology to require care in design and usage, and that this added  
design effort requires one or more new standards.

The alternative is to convey usage through an ad hoc recipe - "stir  
so much VOTable with this much SIA, add a dash of SSA, a pinch of  
utypes, a jigger of STC, and bake until a toothpick comes out  
clean".  More to the point, interoperability won't be achieved unless  
the cook is a Julia Child.

Sounds like the natural role of a standard to me.

>       So it is no more a prototype it is a set of operational  
> services using VO standards.
>
>        We never pretended to have built a STANDARD bu to use  
> EXISTING standards
>      for actual INTEROPERABLE work.

The proof of an interoperable pudding is whether the OTHER guy thinks  
interoperability has been achieved :-)

- Rob