Open Bonnarel opened 1 month ago
On Tue, Oct 22, 2024 at 10:21:00AM -0700, François Bonnarel wrote:
I think it would be unambiguously the cutout dataset. Because in
But this means that the identifier has to include the cutout information, because that's all the datalink service knows.
This is certainly doable; my gut feeling would be the standard pattern could be using the fragment identifier, as in
ID=ivo://example.edu/~?dataset/identifier#CIRCLE 23 24 3
But I can't say this doesn't feel like impending doom.
Just telling people "use SODA for cutouts" really is not an option?
During Sydney DAL-2 session (https://wiki.ivoa.net/internal/IVOA/InterOpMay2024DAL/DAL-2-notes), Gregory D-F made the following remark : if we would use the new proposed RETRIEVEMODE=cutout mode in combination with DataLink, what is the meaning of the "this" item in the DataLink response ?
I think it would be unambiguously the cutout dataset. Because in this case the discovered dataset is the cutout, not the full dataset, and so DataLink semantics applies to this cutout.
If the data provider wants to give access to the full dataset as one of the link in the DataLink response, the semantics will be "progenitor", because the service is creating the "this" cutout by processing the full dataset.