[Missing Practical Stuff]

[Matthew Graham]

Hi Francois,

> <>    Just to summarize what it is all about:
> <> 
> <>     Brian's "private" proposal about External Resource, including
> <>                 Format Mime/types of the Files
> <>                 computer encoding ....
> <> 
> <>     Transform encoding / compression
> <>     File size
> <> 
> <>     Pointers to real data and "Location" of relevant data in the dataset
> <
> <None of the above are to do with modelling data!  Nevertheless they are 
> <important, and I wonder if they would be better 'opened' in the interop group? 
> <Since they seem to be about locating & describing data instances in general? 
> <Some other groups have already considered some of these aspects - such as 
> <Registry and 'pointers'.
> <
> 
>   Suppose you use any kind of serialization of Observation + Quantity DMs
> (VOTable or "standard XML") in order  to describe your Observational data.
> At some point you have to describe in this serialization where the data
> really are, what kind of format, if they are compressed or not. Your software
> will have to take appropriate decisions according to the actual values of 
> these attributes. So of course it has to be there in the data model for
> description. This does not prevent us to take into account the work done by
> the other groups.

It strikes me that there is overlap here with the scope of metadata
that should be in the registry, although it might not necessarily be there
yet.

	Cheers,

	Matthew