Dear Doug et al.,
As I agree with almost everything you said I won't repeat it - except to say that whilst I agree that the present models are not general, they are the best starting point to make sure that we solve real problems. Peter Lamb has a very evolved model for ATCA data which is being developed as part of establishing the on-line ATCA archive for existing data (see his dm@... and radiovo-at-ivoa.net messages) and which we were planning to generalise for the IVOA under Peter's leadership. Anyone who was at the Strasbourg DM session after ADASS may remember Robert Power's presentation on this.
Studying the ATCA model and looking at the presentations on the ALMA model has made me think that we need to develop models for a purpose, and it seems to me that there are three main applications: 1) Access to data via a VO interface for the astronomer
They may be the PI or someone after public domain data. In general
they will want products which are ready for science or at least for
processing with common tools.
2) Data handling at the observatory.
3) Planning observations (not detailed scheduling).
We should concentrate on how to provide the models and metadata for 1) and 3) which will arise from a sort of front-end to 2); 2) probably needs some kind of common skeleton but if we try and standardise it even for every main observatory's products we will never finish, I fear. We should provide support for observatories (etc.) to develop 2)'s suitable for supporting 1) and 3).
It would be nice to see the radiovo mailing list take off, perhaps when the ATCA and ALMA models can be openned up for discusion? and we could also have telecons if we can find a time to allow participation from at least 3 if not 4 continents.... but you can't beat a face-to-face meeting. If anyone is planning on attending the AVO Demo at Garching in January, maybe we could meet the day after (not before, for the sanity to AVO people involved in the demo...)? At the IVOA meeting in Cambridge early this year we tried to have an informal meeting which was useful but not sufficient, and the DM cabal at ADASS was even more elusive, so I think that we do need to set aside a decent length of time which doesn't clash with anything major. Garching people, is this a possibility please?
thanks
Anita