[SpectralDM] - UTypes changes

[Louys Mireille]

Bonjour à tous ,
Hi all,


In response to Omar, I think this needs some correction because it gives 
an ambiguous vision of the work that has been done in the IVOA DM 
working group all over the years.
>
> I disagree with this. During the discussion whether or not we should 
> have reopened ObsCore to make the suggested changes, it was stated 
> that ObsCore *is not* a data model, but a view on an unspecified data 
> model. As such, it is part of the specification of a DAL interface, 
> which is fine.
>
Here it is important to avoid confusion:
  ObsTAP, is a DAL protocol based on the TAP protocol : it defines TAP 
schema and tables columns according to the data model classes defined in 
ObsCore: Observation Core Components Data Model.

> Part of Mark's effort is to make sure that we also have at least a 
> basic ObservationDM for the benefit of Cubes, Spectra, and SEDs, 
> because we indeed want to have all DMs to be modern, interoperable, 
> and to derive them from a common set of core models.
>
Core models ...which have been historically , Resource Metadata, STC, 
Characterisation and Spectrum and finaly...ObsCore.

This is also the reason why  it was decided last fall , after several 
discussions at the Heidelberg meeting, in May 2013, and in the summer 
that ImageDM should rely on the ObsCore kernel.

The stitching of the final "big picture " data model that Mark has 
produced , which federates different use-cases , with different data 
products Image , Cube, Spectra , SED , etc... is a real challenge.
I think it requires however constructive adjustments with respect to 
what already exists.

Best wishes, Mireille.
Icube & Observatoire de Strasbourg , Strasbourg University.