Error column in VOResource

[Martin Hill]

Thanks Roy and Ray...

Roy Williams wrote:

>>Would it be a good idea to include a reference to the associated error
>>column in a column's description?  eg:
> 
> 
> But what happens when there is an error ellipse, RA, Dec, and the three
> components RA.RA and Dec.Dec and RA.Dec of the covariance.

Hmmm yes.  Ouch.  Is this peculiar to location, or does it happen with 
others too?

> 
> What we have thought of in VOTable is an adjunct structure written in RDF that
> expresses relationships between the columns, betwwen the columns and UCDs and
> other relationships.....
> 
> In fact, looking at your XML, may I politely ask why you have chosen not to use
> the already established VOTable schema? I had always assumed that the reason
> VOTable is included in VOResource is so that table metadata could be understood
> by the same software that reads the tables themselves.

Another new one to me - I hadn't realised VOTable was included in 
VOResource... How do VOTable and VODataService relate?

If you mean that a service returns a VOTable, and so we can ask for a 
VOTable header, that rather implies that you know already which columns 
to ask for, as the resultset will not (I believe?) give all the columns 
every time.

What I am looking for is a way of describing a dataset fully enough that 
not only can a software tool discover suitable datasets but also build 
suitable queries on the found ones.  This might mean adding metadata to 
a datacenter that doesn't get put in the Registry - but I rather feel 
that any metadata might as well go to the Registry to make 
discovery/querying easier...





-- 
Martin Hill
www.mchill.net
07901 55 24 66