About the draft itself Re: WD-SIA-2.0

[François Bonnarel]

Le 25/11/2013 18:27, Patrick Dowler a écrit :
>
> I have uploaded the current (incomplete) WD to the document repository:
>
> http://www.ivoa.net/documents/SIA/20131115/
>
> This is far fro m complete but I wanted to get an initial round of 
> comments so we could stabilise something and get on to implementing it.
>
>
Hi all,
    I am coming back to the SIAV2 draft and proposes some changes or ask 
some questions here.
0 ) page 5. the last sentence in the introduction is incomplete . "For 
access to datacubes SIA...."

I ) 1.2.4 and 1.2.5 . (page 7) I would merge these two paragraphs and 
write the 1.2.5 sentence ahead because the main motivation to have 
AccessData resource with operations  is because  datacubes are too large.

II ) 1.2.4 (page 7) and 2 intro (page 8). There is some confusion 
between "Data Access" as in "Data access layer" which is general and 
"AccessData" resource wich as far as i understand is there to "Access 
inside the data" or "extract from the data". So "a low level data access 
service" should be rephrased " a low  level AccessData service" (1.2.4) 
and "plan and execute data access operations" should be "plan and 
execute AccessData operations"

III ) I don't understand "alternate authentication schemes". Can you 
give an example ?


IV ) POS, BAND and TIME

       IN case of shapes and intervals what is the relationship between 
this shape and the range of FOV of the dataset ?
BAND states clearly that "searching using an interval will find any data 
with energy coverage that overlaps the specified interval". (page 10)
"THe TIME parameter defines the time interval to be extracted from the 
data". Does that mean the same behavior?

The POS behavior for ROI should be the same by default for consistency. 
But it is not stated in the text.

By the way   other behavior (find out datasets where coverage is totally 
included in the ROI/ parameter ranges, or alternatly intersecting with 
them) is also  interesting. Should we reintroduce the SIAV1 INTERSECT 
parameter for that behavior description ? Or do we wait the next version 
of SIAV2 ?

V ) POL .  Pat has asked a question about the meaning of having several 
POL parameters . OR or AND ? (page 12)
      I think AND will be also needed. Stokes parameter is an ordered 
list according to FITS  standard = I,Q,U,V
     So consistenly with other parameters I/V or Q/V would mean the 
whole range of Stokes paremeters from I to V or Q to V

VI ) 2.2 metadata resource. "The output format is an XML document 
defined by the ImageDM" . Will it be an XML document ? a VOTABLE one ? 
Or something else ? Should SIAV2 version 1.1 have to say a world about 
the output format ?



cheers
François