[Dachs-support] Clarifications about new coverage and limits

Markus Demleitner msdemlei at ari.uni-heidelberg.de
Wed Feb 16 12:56:23 CET 2022


Hi Yannick,

On Wed, Feb 16, 2022 at 12:19:24PM +0100, Yannick Roehlly wrote:
> So I suppose we still need to have the <coverage> tag with the
> inner <updater> tag.  Am I right?

Yes -- DaCHS needs to be told where to get the resource coverage
from, and that's this updater thing.

> In some cases, I have several tables in a resource that spans
> different locations in the sky (different fields).  What is the
> best approach here? 

In that case, keep things as they are, i.e., define the coverage
manually.  There is no way to tell DaCHS to get coverage from
multiple tables, and my gut feeling is that I don't want to teach it
how to do that.  Multiple updaters would be an option, and their
results could reasonably be merged, but... well, I'd rather if we
didn't have to go there.

> Before I was manually computing the MOC of the spatial coverage
> because having two <updater> in the coverage was not working, is it
> still the case?

Yes.

> If I understand correctly, the spatial coverage is associated to
> the resource, not to the service; so I have a resources with two
> tables on two field and two associated cone searches, the
> information page of each cone search can only see the full
> coverage.  Am I right?

Yes -- you *could* define two VO resources, but at this point this
would require that the tables sit in two different RDs.  As long as
two VO resources come from from the same RD, they'll have the same
coverage.

Lifting that, i.e., allowing coverage elements (presumably) inside of
publish elements is something I would consider, though.  Push me and
I'll look into it.

> 
> PS: A note for users having troubles displaying the coverages on the web 
> pages. You need to have python3-healpy installed which is only a 
> "recommendation" for DaCHS packages and may not be installed depending on how 
> you install them.

Right.  Perhaps that should be stated more clearly... somewhere.

Thanks,

          Markus


More information about the Dachs-support mailing list