content, format, ctype, or xtype ?

[Rob Seaman]

On Mon, 4 May 2009, Patrick Dowler wrote:

>> This is not intended to be a metadata concept at all, just an  
>> indication of the format a value is expressed in.

This is an instance of "data about data" - the very definition of  
metadata.  I presume the distinction here is between something like  
science-metadata and ordinary-metadata.

On May 4, 2009, at 11:14 AM, Doug Tody wrote:

> If we want a simple way to specify the format of a field, without  
> having to understand a complex external data model, then a separate  
> mechanism is required.

Again - use cases?  What we want doesn't matter; what do VO users need?

If enough of the use cases benefit from not "having to understand a  
complex external data model", should effort be directed toward  
developing such things?

For a real-world example, look at the two quotation instances above.   
The data are the quotes from Pat and Doug.  The metadata are the  
authors and the epochs of the utterances.  The epochs are not only  
specified in two different formats, but each includes information that  
the other omits.

Whatever usage the VO specifies should be loose fitting enough to  
allow our users to draw an occasional breath.

Rob