Identification model in VO Spectrum Model

[Alberto Micol]

Dear Pavlos,

The most recent version, in case you do not know,
is actually Version 1.0 Release Candidate 1 Rev 2,
and you can find it here:
http://hea-www.harvard.edu/~jcm/vo/docs/spec100/specrc1.pdf

Dear Doug, dear all,

> When it comes to what is mandatory, recommended, or optional, one  
> thing
> to bear in mind is that this depends upon the context.

I have heard this tons of time, in various contexts (VO Char DM, etc)
and never agreed. That is, I do agree that it depends on the context,
but I do not agree that this be the reason not to have more MAN fields.

On one hand, leaving freedom to the data providers is calling for  
troubles:
many parameters might end up being missing for proper interpretation
at the client side.

On the other hand, leaving too many parameters at a non-MAN level  
makes the
data providers unsecure: "should I do my best to put this parameter  
in or
not? and if so, how do I define it in my particular case?"
A lot of assumptions will be made by the implementors, and not
necessarily the correct ones.

I think that anyone willing to adopt a given protocol/DM should be  
guided
as much as possible in his/her own context. Reading better (read  
"stronger"
if you wish) instructions it does help.

Wouldn't, hence, be possible to specify (in the protocol/DM)
the  different available contexts (adding for example an enumerated
CONTEXT attribute), and to define for each of them the proper
mandatory/optional/whatever fields?

With that, clients will be able to specialise on a given context
based on solid grounds, and I find this possibility extremely valuable.

Someone could argue that we cannot imagine all the possible usages
of a DM a priori; true! but on the other hand, as soon as someone comes
up with a new context, it will be not difficult to include it in the
next release.

My two cents ... of pesetas.

Alberto


On Oct 25, 2006, at 21:10, Doug Tody wrote:

> On Mon, 23 Oct 2006, Pavlos Protopapas wrote:
>> I hope this is not being already discussed tons of times but
>> in the latest document of VO Sectrum DM I have (0.98d Rev 4)
>> in the data identification model (sec 5.5) states that all dataid  
>> fields are
>> optional. Shouldn't we have at least the Curation.PublisherDID be  
>> MAN ?
>
> When it comes to what is mandatory, recommended, or optional, one  
> thing
> to bear in mind is that this depends upon the context.  The  
> defaults in
> Spectrum mainly address what is needed to work with an individual  
> dataset.
> When the data model is is used for a data query however, the SSA  
> protocol
> defines what is mandatory or recommended, and this overrides what is
> specified in the underlying data model.  Hence in reviewing what is
> mandatory or optional, one really needs to look at both documents, and
> decide whether the use-case is data discovery and selection, or  
> analysis
> of an actual dataset.  (Note also, there is a distinction between  
> DataID
> and Curation).
>
> 	- Doug
>