UTYPES

[Alberto Micol]

Dear All,

I think we have to really clarify the different roles of UCDs and UTYPEs.
(well, if not everybody, at least *I* do need it).

It looks to me that many people even on the DM list think that UCDs
are a DM thing, used to refer to complex concepts.

For me UCDs are solely a controlled dictionary of basic words,
while UTYPEs are a way to map to higher level concepts.


To clarify my position, I here give an example I mentioned in a
private email:

 > The fact that a sky coordinate refers to a particular coordsys
 > is not relevant to the observation data model per se.
 > So, the coordsys should be expressed by a UCD not by a UTYPE.
 > The UTYPE tells it's a sky coord, the UCD tells of which kind.


In a sketchy way I would say that:


  UCD -> Vocabulary
         list of *words* referring to basic definitions


UTYPE -> Encyclopedia
         list of *headwords* referring to complex concepts


The confusion lies in the fact that any headword -obviously- is also
listed in the vocabulary ... (but try to buy the Encyclopedia Brittannica
at the price of a vocabulary; the salesman will teach you the difference).


For what I just said, and as opposed to my previous email
( http://www.ivoa.net/forum/dm/0456.htm ),
I just figured out it is probably correct not to mention UTYPEs in the
Q-DM, since Q holds very simple information.
It should be a higher level DM (e.g., the OBS one) to define UTYPES
to be used in conjunction with Quantities when needed.

(... and maybe we should ask that salesman to develop utypes for us ...)


Alberto
PS: Is it really just me the one needing some discussion on this topic?