Discovering Data Collections Note

[Markus Demleitner]

Hi Pat,

On Wed, Nov 09, 2016 at 12:02:22PM -0800, Patrick Dowler wrote:
> But stepping back, I have to say I am not really a fan of documents
> other than the standard defining these. Basically,
> ivo://ivoa.net/std/TAP#sync-1.1 says that the concept of sync-1.1 is
> defined in that standard.... so ivo://ivoa.net/std/TAP#sync-aux-1.1
> would be a big fat lie unless sync-aux-1.1 is defined in that

Well, neither StandardsRegExt nor any other document says what's
defining the keys in these records.  But we don't have to decide this
now, I agree that in the future standards should define their aux
ids.  When this note is endorsed, the creation of the extra standards
keys in the legacy records is just a one-off measure.

> PS-Could we request/issue errata to add aux standardID(s) to existing
> standards as they are needed (eg the vospace example above) rather
> than wait until the next version?

Well, for the aux ids given in discovercollections, I'd propose to
just consider the endorsed note as a "collective erratum".  The
creation of the StandardKeys appears an issue too minor to actually
require a flood (5, I think) of errata.

If people disagree, I'd produce them, of course.

As to changing the standardIDs of existing standards: No, I don't
think we should do this.  They're hardcoded in quite a number of
places, and cosmetics in my eyes certainly don't justify breaking
all that.

This, by the way, will become an issue for TAP 1.1, and I've not yet
quite figured out what I could even suggest as a solution.  But
changing the standardID is actually a compatibility problem severe
enough to question if 1.1 is a dot release, given how service
discovery is being done so far (which, of course, was the main reason
for reforming standardIDs).  But that's a discussion for DAL.

        -- Markus