This discussion of VOQL and DAL seems of broader interest so I am forwarding
this to the architecture WG. Doug
Tony -
You seem to be arguing for a larger presence for SkyNode+ADQL in the DAL WG agenda. I originally had it in at the same level as the other services but it was squeezed out by the interest in having a joint session with the UCD WG, plus all the current activity in this area has been happening in VOQL. DAL really needs a working ADQL before we can worry too much about more sophisticated query capabilities in the DAL services. In the meantime adding access to new classes of astronomical data such as spectra and time series is more productive. Nonetheless, you have a good point. Perhaps we should expand the future topics session and increase the emphasis on catalog access and query language integration into the DAL services.
I agree that a generalized QL-based catalog search (SkyNode) is a high priority. The problem is the next step requires an advanced query lanaguge capability to address. Regarding DAL, there is no clear distinction between tabular and gridded data, so it would be a mistake to try to split the two. Similarly, the data query is a major part of the DAL services so we can't ignore this area either. DAL needs QL capabilities, but the major focus of DAL should be on integrating such technology (as with data modeling) to provide uniform data access services.
It would seem that the area of query language and catalog queries is broader and deeper than just what is required for data access services. The full query language problem includes query portals, a major VO user interface. The generalized catalog access problem includes distributed queries and joins, and large scale queries, both of which are very hard problems. Hence I think it makes sense to have a separate group develop the query technology.
For the above reasons I think it makes sense to have separate DAL and VOQL working groups. The main issue is adding a basic (nondistributed) SkyNode to the DAL services, and integrating query language capabilities into all the DAL services once this technology is sufficiently developed.
On Thu, 13 May 2004, Tony Linde wrote:
> Hi Doug & Masatoshi,
>
> The VOQL list have discussed the SkyNode protocol but it probably belongs
> under DAL in the architecture. We ought to consider whether there is some
> common interface from which SkyNode, SSA and SIA ought to inherit. If we do
> keep the split as it is now, maybe we need to rename the groups Image Access
> and Data Access.
>
> I think ownership of the standards should be resolved before the Boston
> Interop and then driven forward at that meeting. I'm worried that SkyNode
> might slip between the cracks while VOQL argues about ADQL and DAL about
> SSA/SIA. Yet SkyNode is probably the most important protocol we need right
> now.
>
> Cheers,
> Tony.
>
> > -----Original Message-----
> > From: owner-dal-at-eso.org [mailto:owner-dal-at-eso.org] On Behalf
> > Of Doug Tody
> > Sent: 13 May 2004 04:53
> > To: dal-at-ivoa.net
> > Subject: Proposed agenda for DAL session at the May Interop meeting
> >
> > Dear DAL folks -
> >
> > The Data Access Layer discussions will be held 08:30 - 12:00
> > Tuesday and Wednesday mornings, May 25-26, in Jefferson room
> > 256. We may also try to have some smaller impromptu focus
> > meetings during the week to work out details of the
> > interfaces currently in the definition stage.
> >
> > What follows is a draft agenda for the DAL sessions with some
> > suggested topics in each area. If you would like to suggest
> > a new topic or contribute in any of these areas please let me
> > know. Those of you who have actively worked to develop
> > specific material in the past six months are encouraged to
> > present the relevant topic. The time allocations will be
> > determined later as the agenda is developed.
> >
> > Any comments or other suggestions are most welcome!
> >
> > - Doug
> >
> >
> > 1. SSA, time series
> > - Service classes
> > - Query interface
> > - SSA data model
> > - Data representation (VOTable, FITS, XML, text)
> >
> > We are still working on the SSA interface and data model documents
> > and hope to have these available in advance of the workshop.
> >
> > 2. SIA V1.1
> > - General dataset metadata (identification, characterization)
> > - Logical dataset identifer proposal (logical groupings)
> > - Ranking proposal
> > - UCD normalization is discussed separately; see below
> > - Anything else?
> >
> > SIA V1.1 is a limited scope upgrade to the current
> > interface. The goal
> > is the bring the interface up to current standards and
> > add some modest
> > new capabilities.
> >
> > 3. UTYPE and UCD (Joint with UCD WG!)
> > - When and how to use UCD, UTYPE
> > - UCDs for SIA V1.1
> >
> > This topic has received a great deal of discussion over
> > the past year.
> > Maybe this time we can nail it down, at least as regards usage in
> > SIA V1.1, and SSA, which will follow the same pattern.
> >
> > 4. Roadmap and Future Topics
> > - Catalog access (plans for ADQL usage in DAL, updated
> > catalog access)
> > - Exanded image access capabilities
> > - Interferometry and event data
> > - Other
> >
> > What is required to support the applications we plan to build over
> > the next year?
> >
> >
>
>
Received on 2004-05-13Z20:03:09