resuming progress on TAP

[Doug Tody]

Hi Arnold -

Then why do you suppose none of the existing (already fully functional)
NVO TAP prototypes which implemented PQ this past summer, did not
follow this approach you suggest and do it on top of ADQL?  Native SQL
and local spatial indexing mechanisms were indeed used (very easily),
but ADQL, no.  What you describe is merely theoretically possible.
In an actual service implementation it is much easier and more
robust to have both at the same level and sharing the same back-end
processing.  There are many issues, for example would you really want
to require a full up ADQL-based table metadata query capability just
to be able to provide the basic, much more constrained table metadata
queries such as PQ defines?

 	- Doug


On Tue, 10 Feb 2009, Arnold Rots wrote:

>> From a relative outsider who has been watching this for a while:
>
> If I think about TAP from an implementation point of view and
> consider that AQ support is mandatory, PQ support optional,
> I would be inclined to first implement the AQ protocol and then
> add the PQ protocol by writing a PQ->AQ converter.
>
>> From that perspective, it makes sense to first make sure the AQ
> standard is well-established and then work the PQ in - which, if I
> understand the conversation correctly, is precisely what Keith is
> proposing. So, (I think) I agree with Ray and wonder where the current
> debate is heading.
>
> But don't let this stir up another parallel discussion.
> My 25c.
> Cheers,
>
>  - Arnold
>
>
> Ray Plante wrote:
>> Hi Guys,
>>
>> I guess what I don't understand is what exactly the motivation is behind
>> this suggestion to either take TAP into this two step process or to split
>> it into separate documents, besides that you like the idea.  What problems
>> is it meant to solve?  We have these last two integrated vesions, and as
>> I understand it some effort to further tweak the organization back around
>> early December.  Why, then, do we need to take a step backward?
>>
>> It's not that I don't think there are issues--we know there are.  But what
>> are they?  I can't see how these new proposals will solve problems when
>> the problems have not been spelled out.  (I made my own attept to pull
>> some out in a previous message, but I was hoping others who have concerns
>> would explain what they are.
>>
>> One concern of mine is that these proposals are effectively about *not*
>> talking about the issues.  If we can write down what they are, we can talk
>> about them.  If we can talk about them, we can propose solutions and
>> discern which direction to go.
>>
>> cheers,
>> Ray
>>
> --------------------------------------------------------------------------
> Arnold H. Rots                                Chandra X-ray Science Center
> Smithsonian Astrophysical Observatory                tel:  +1 617 496 7701
> 60 Garden Street, MS 67                              fax:  +1 617 495 7356
> Cambridge, MA 02138                             arots at head.cfa.harvard.edu
> USA                                     http://hea-www.harvard.edu/~arots/
> --------------------------------------------------------------------------
>