WG Review period for TAPRegExt-1.0

[Mark Taylor]

Pat, Markus et al.

On Fri, 18 Nov 2011, Patrick Dowler wrote:

> I would like to start the (informal) WG review of the TAPRegExt now so that I 
> can begin the RFC in January.
> 
> The latest WD is here: 
> 
> http://www.ivoa.net/Documents/TAPRegExt/

I'm mostly happy with the current draft.  I have one or two comments
and corrections.


Sec 2.3, discussion of language features:
-----------------------------------------

Where the features-adqlgeo key is defined its IVORN is quoted wrong,
it says:

    ivo://ivoa.net/TAPRegExt#features-adqlgeo

when it should say:

    ivo://ivoa.net/std/TAPRegExt#features-adqlgeo

(compare with the features-udf key above).

Having noticed this I did a 'grep ivo://ivoa.net/TAPRegExt *' in 
the volute projects/registry/TAPRegExt dir and found quite a few
matches.  I think most of them don't find their way into the 
built document, but I don't understand the build well enough to
work out if there's a good reason they are in those files.
Maybe Markus could do a similar grep and make sure the std-less
IVORN is not getting used anywhere.

In the same section, the features-adqlgeo key has an example <feature>
element but the features-udf key does not.  Probably the example isn't
essential, since examples of both feature types appear in Appendix B,
but it would make sense to have it consistent?  Especially since the
UDF example is more complicated than the Geom one.


Appendix B (Example document):
-----------------------------

ADQL version ivo-id is incorrect:

    <version ivo-id="ivo://ivoa.net/std/ADQL#2.0">2.0</version>

should read

    <version ivo-id="ivo://ivoa.net/std/ADQL#v2.0">2.0</version>

The non-standard uploadMethods listed here, while not incorrect,
may be a bit confusing.  How about removing the 
"ivo://ivoa.org/tap/uploadmethods#"* entries.


Other than that, the document looks good to go to me.


> PS-If/when people implement some portion of this in the VOSI-capabilities 
> resource of their TAP service, please let me know (base service url would be 
> good) as I need a list of "implementations". Thnx.

The most recent version of the taplint TAP service validator in STILTS
performs fairly comprehensive validation of Capabilities documents 
which implement TAPRegExt.  Validation tools are required or at least
encouraged alongside implementations by the IVOA Document Standards,
so you should mention this as well (it's a kind of implementation in
any case).  taplint refs:

   http://www.star.bristol.ac.uk/~mbt/stilts/sun256/taplint.html  (docs)
   ftp://andromeda.star.bris.ac.uk/pub/star/stilts/pre/       (download)
      
Note it's possible to use it standalone to validate a TAPRegExt document
which is not deployed in a TAP service as well, though this is not
documented or obvious - I can explain how if anyone wants to know.

Mark

--
Mark Taylor   Astronomical Programmer   Physics, Bristol University, UK
m.b.taylor at bris.ac.uk +44-117-928-8776 http://www.star.bris.ac.uk/~mbt/