I think the right word is semantic meaning. I LOVE UCDs, but they
are not part of what we call Schema. They are metadata. And we have
been fuzzing the otherwise very clear distinctions between those.
Again, a similar thing is with units. They are part of a broader meaning,
but in order to parse an item, I do not need to know its units, but
I need to know its datatype.
We can of course shove it all to the same place, but as we keep adding more metadata (UCD2, more verbose explanations, annotations, etc) it will be harder and harder to separate where the strict schema ends and the metadata begins.
I do not have a clear solution, but this is at the core of the issue.
Cheers, Alex
-----Original Message-----
From: owner-votable-at-eso.org [mailto:owner-votable-at-eso.org]On Behalf Of
Tony Linde
Sent: Wednesday, January 21, 2004 3:44 PM
To: votable-at-ivoa.net
Subject: RE: V2
Yes, I was coming to the same conclusion. We have data, contextual information (where the data came from and what it is about) and representational metadata (how this particular set of data is represented in this place).
I guess the confusion arises because it is all being referred to as 'metadata', which may or may not be correct but to a computing professional is misleading.
There's no problem with the data and representational metadata but we need a schema for the contextual information.
And is it valid to have FIELD elements in the contextual information since they are really representational, aren't they? Is a UCD part of the representation of the element, like its datatype, or part of the element's context?
(And is 'context' the right word?)
Cheers,
Tony.
> -----Original Message-----
> From: owner-votable-at-eso.org [mailto:owner-votable-at-eso.org] On
> Behalf Of Martin Hill
> Sent: 21 January 2004 19:12
> To: votable-at-ivoa.net
> Subject: Re: V2
>
>
> Tony said:
>
> > Surely it would be much simpler to accept that VOTable is simply a
> > mechanism for transporting a table along with its metadata
> but that to
> > use standard xml tools on the data you need to split them
> into their
> > normal separate
> > components: data and metadata. (But you could leave them
> together if you
> > really want to work on the combined file.)
>
> As I understand it so far, there is no mechanism for putting
> astronomical
> metadata (eg UCDs, related passband info, etc) into schemas.
> We actually have
> two layers of metadata - one describing the structure and
> overall naming of the
> data (schema), and another with values eg as above (votable
> header) that give
> context to the data values.
>
> Martin
> --
> Software Engineer
> AstroGrid @ ROE
> Tel: +44 7901 55 24 66
> www.astrogrid.org
>
>
Received on 2004-01-21Z23:04:32