SED data model v0.92

[Gilles DUVERT]

Congratulations for the IVOA SED Data Model , this is an impressive 
piece of work.

If I may risk three comments, two on the draft itself, the other on 
comments of comments...
Sorry for the rather direct style, and for my probable misunderstanding 
of some parts of the document...

1) Ed Shaya wrote:

> Hopefully, it is rare that one only has upper limit info.  The 
> observers certainly should provide the measured value even if it is 
> below the noise.  There is non-zero information in that value.

Well, although having only upper limits may seem pitiful, a number of 
sound scientific results have been drawn in the past from 
undetectablilty results. Upper limits are really a result for SEDs, and 
used as such by astronomers. Since they are used, they are published and 
they will go  in the  VO...
 I just wanted to comment that a value *measured* cannot be *below* the 
noise. Only the noise level is meaningfull in this case (the text notes, 
and this is customary indeed, that authors usually "choose to render 
measurements as upper limits if the flux value is less than some 
multiple (e.g. 3) of the lower error" (note: shouldn't  be *Upper* 
error? ) ). The risk would be, if any *measured* value is set, that is 
is taken at face value, when only the noise level has sense. Could you 
use some kind of blanking value for the measured value in this? (or is 
there a general concept of upper limit that would go in the 
Quantity::Accuracy data model?)

2) in http://hea-www.harvard.edu/~jcm/vo/docs/spec.html, Errors are 
OPTIONAL?? Errors are part of any measurement, and should be enforced by 
the data model. Or is there provision for a sort of common error for all 
measurements? Because a measurement without errorbars should simply not 
exist  (but the reverse may exist, see above)!

3) http://hea-www.harvard.edu/~jcm/vo/docs/spec.html use  non standard  
old 'ergs'  (cgs) units... All documents produced in the 21st Century 
should promote MKSA units, and conversion to the dear cgs units of our 
grandfathers should be implicit and computer-assisted... ;^)

Best,

Gilles

-- 
----------------------------------------------------------------------------
Gilles DUVERT, Astronome        | Gilles.Duvert at obs.ujf-grenoble.fr
Laboratoire d'Astrophysique     | Phone: +33 (0)4 76.51.48.85
Observatoire de Grenoble        | Fax:   +33 (0)4.76.44.88.21
BP 53X  F-38041 GRENOBLE CEDEX  | LAOG (http://www-laog.obs.ujf-grenoble.fr)
Dir. Technique JMMC(http://mariotti.ujf-grenoble.fr) Resp. Informatique LAOG
----------------------------------------------------------------------------