Elements to add to TabularSkyService

[Gretchen Greene]

Hey Martin,

If you develop an extension that uses some of these new concepts as Ray
suggest,  it is along the lines of what we've discussed before with
schemas that are essentially a step beyond the standard resource
metadata.  To me it's much like the object-oriented approach to sw
practices in general,  components.....

Bear in mind though that much of what we are planning to do with the VO
registries is the be able to harvest 'between' the varying repositories.
I certainly understand the tedious development steps from our own
prototyping.  Again,  if you can present an extension it is not so
difficult for testing on our end and to support harvesting not sure what
all needs to be in place. 

The problem that I see growing is how the population and curation of the
metadata for resources will evolve once schema advance so far.  There's
a lot of variation between say datacenters registering managed datasets,
vs harvesting other centers and supporting independent publishing.  

Not sure what you mean by being "far away" from document based
descriptions of metadata.  I haven't seen any practicing astron archive
that has all these features in place.  We use technology which has more
XML features than I can imagine at this point.  Like I mentioned in the
earlier message ...when the ideas are being thrown around with far and
few between prototypes,  the technology is just waiting for the chance
to be demonstrated.

-Gretchen


-----Original Message-----
From: owner-registry at eso.org [mailto:owner-registry at eso.org] On Behalf
Of Martin Hill
Sent: Thursday, July 01, 2004 3:46 PM
To: registry at ivoa.net
Subject: Re: Elements to add to TabularSkyService


Gretchen Greene wrote:

> Martin,
> 
> I wouldn't rule-out NVO using TabularSkyService at all.  We are in the

> process of populating and testing new schemas just the same.  Our 
> prototypes were based on much earlier concepts and then yes...we have 
> advanced services which have their own similar set of 'elements' such 
> as skynodes.  This doesn't mean we are not interested in using these 
> newer schema elements,  it is simply a process of development that we 
> are exploring.

OK dok... however it sounds like you're some way away from using this 
document-based way of describing metadata (skynodes do their own thing 
with calls to methods that return lists of tables, lists of columns 
etc). So I was really wondering if we/Astrogrid can 'take ownership' of 
it for a while (viz ADQL by NVO) to develop it into a real-world usable 
document, and then present the results to the registry group for 
discussion/dismemberment/etc.  Rather than going through the rigmarole 
of discussing each small change!

However I think Ray has suggested a way of doing this but using a 
completely different extension, which will save us from interfering with

each other.

> With all the discussion on the theoretical aspects of the schema dev,

> I encourage you to share some instance examples.  It really helps 
> clarify how the schemas apply.  Project drivers always differ yet real

> examples go a long way.  Throw some out there!

Yers - the examples I have now are only from discussion with a couple of

innocent data providers who are about to be VO'd (poor things :-).  I'll

definitely put up some real examples when I have some.  I suspect before

then there will be lots of questions from me and them about other 
aspects of VOResource!

Is this the right forum to use to ask questions about 
ambiguous/undocumented elements in the VOResource document?  Perhaps 
with a special subject marker [SupportRequest] ?

Cheers!

MC


-- 
Martin Hill
www.mchill.net
07901 55 24 66
0131 668 8100 (ROE)