SED Data Model: Questions and Comments

[Gilles DUVERT]

Hi,

Thanks for your comments on my *deliberately provocative* mail!

If I sumarize:
- Igor suggests to close a version 1 of a  restricted and simplified SED 
DM that does not take into account all the peculiarities/oddities of ALL 
possible SED measurements lying around. This leaves time to 1) test this 
'simple case' DM in practice and 2)  think twice about version 2.
I could not have said better.
- Ed acknowledges the need of making (meta)data as independent as 
possible from the expertize of one data provider, an (usually mortal)  
group or person. This can, I think, be done easily in *most* cases by 
some work from the originating group/person to fit/convert/translate/tag 
his data into a given DM framework. The simpler the DM, the better.
- Anita is (rightly) preoccupied by the fact that a 'restrictive DM' 
would prevent inserting numerous valuable data in the VO, and provides 
three mechanisms that could suffice, I quote:
1)  We must ensure that data (...) have a warning flag when, for 
example, errors are not given.
2) Data could be published albeit with warnings since Resource Metadata 
v1.0 has 3 grades of overall quality, from totally calibrated to unknown;
3)  there could be a mechanism for users to add feedback if they deduced 
more accurate metadata.
For point (1), doesn't it mean that Flux.Quality is  not 9  only  when  
the group sed:Segment.Points.Flux.Accuracy is not void?.
I was not really aware of (2) and I guess it should suffice for a start 
but this does not imply that the DM goes very far in trying to match all 
possible kind of data...
and (3), if it is feasible, is really brilliant!

If 'Quantity' is, or will be, able to deal with such things as flux 
conversions from/to air wavelength vs. humidity and frequency,etc..  and 
Janskys from/to Cousins/Johnson magnitudes, at least to a large extent 
and great accuracy, all the best. In that case all units or combinaisons 
thereof supported by Quantity should be authorized in, e.g., the SED DM. 
But at the same time one could use the Quantity DM to convert the data 
beforehand to specific quantities that would be the only  ones referred 
to in the SED DM, such as basic  MKSA units. And make the DM simpler.

All the best,

Gilles
PS: naturally it was not my intention to use DMs to enforce any 'law', 
simply that DMs will be useful if rigorous but, as any esperanto, used 
only if kept simple.