On Jun 24, 2005, at 5:53 AM, Andrea Preite Martinez wrote:
> Of course the discussion is not over, but unfortunately we need a
> rather "stable" version soon.
>
I'm concerned about the general nature of the UCDs in the list. It seems to me that the most useful quantities are those such as em.opt. [UBVRI] that have the broadest or most empirically determined definitions. Steve points out the shortcomings of several of the more precisely definable quantities. Many of these require some more complicated data structure to even attempt to describe the underlying physics. A simple scalar will never suffice. For instance, Steve picks apart several of the WCS related UCDs.
What is the precise benefit to the VO or its users of selecting UCDs to convey these quantities rather than STC elements? Note that I'm not seeking to reopen the "UCDs are not XML" discussion - rather, in a loose analogy in which XML represents a variety of data types and UCDs represent names, what is the point of providing names for scalar quantities that are of no utility standing by themselves? I suppose you could attach a UCD to a more complicated XML (or other kind of) data structure - but how would that work exactly? Wouldn't this possibility need to be reflected within the UCD mechanism itself?
The other thing about "em.opt.B" is that it represents a quantity whose definition is completely within the control of the international astronomical community. Other than additionally specifying exactly what it is whose measured Johnson B magnitude is being quoted (some object through some aperture), there is no intrinsic ambiguity. Some of the other quantities "belong" to non- astronomical constituencies at least as much as they belong to us.
Not trying to make trouble - please point me to a document or discussion thread if such carping is old news.
Rob Seaman
NOAO
Received on 2005-06-29Z05:25:59