Characterising URL columns
Mark Taylor
m.b.taylor at bristol.ac.uk
Fri Feb 2 11:08:55 CET 2018
On Thu, 1 Feb 2018, François Bonnarel wrote:
> Dear Mark,
>
>
> Le 01/02/2018 à 16:23, Mark Taylor a écrit :
> > Francois,
> >
> > On Thu, 1 Feb 2018, François Bonnarel wrote:
> >
> > > I gave talks on this topic both in ShanhaI and Santiago meetings,
> > > following
> > > the ASTERICS discussion.
> > >
> > > See
> > > http://wiki.ivoa.net/internal/IVOA/InterOpOct2017DAL/DataLinkSolution.pdf
> > Thank you, that's an interesting discussion. I obviously wasn't
> > thinking hard enough about datalink in Santiago or Shanghai to take
> > it in.
> >
> > > We prefer either the LINK solution exposed in the talk or if there is
> > > variation of the output the ObsCore-like solution with two Fields and
> > > utypes.
> > The LINK solution however, if I understand correctly, is not
> > standards-based. It appears to rely on the LINK substitution syntax
> > described in VOTable Appendix A.1, annotated:
> >
> > "The definitions enclosed in this appendix are *not* part of
> > VOTable 1.1, but are considered as candidates for VOTable
> > improvements."
> >
> > I believe this non-standard syntax is understood by Aladin, but not by
> > other VOTable clients.
> Yes you are right. We need small changes in VOTABLE anyway (also because
> content-type doesn't admot the datalink value at the moment).
>
> The templating could be pushed to the normalized section if everybody agrees
Hm. This templated LINK thing does seem to be solving the same problem
that is solved by [DataLink section 4] Service Descriptors attached
to generic [non-DataLink section 3] VOTables.
Agreed it's more verbose, but I don't follow your point on slide 4
of the presentation above:
"Not really adapted For One single constrained PARAM (ID=...) which
is more like a templated URL"
- it seems perfectly feasible to use service descriptors like that.
--
Mark Taylor Astronomical Programmer Physics, Bristol University, UK
m.b.taylor at bris.ac.uk +44-117-9288776 http://www.star.bris.ac.uk/~mbt/
More information about the dal
mailing list