VOEvent session

[Roy Williams]

 >  complexity is brought back in through a few
 > hundred utypes mapping to a few dozen data models - well, then, it
 > isn't clear what value the schema is at all.  Also, this seems to
 > suggest that all IVOA XML will eventually devolve to VOTables or some
 > other common container for utypes.

Here is an example to illustrate "layers of validation". XML can only 
validate at a low level, there are always higher levels of meaning. The 
utypes represent this higher level.

Suppose we start with a statement in English:
     "On 2010-13-15T05:18:38, a supernnnova has been discovered at z=25.5"

And then convert it to XML in the manner of VOEvent:
     <Date>2010-13-15T05:18:38</Date>
     <Concept>Supernnnova</Concept>
     <Param name="z" utype="Target.Redshift" value="25.5"/>

What are the validation problems that we might find here, and who is 
going to report them?

(1) The XML schema, if built properly, will find the problem with the 
date -- there is no 13th month.

(2) There is a spelling mistake in the word Supernova. This would not be 
caught by XML validation, may not be caught at all, unless and until 
"Concept" is a constrained to be a controlled vocabulary.

(3) XML schema would certainly not find the bogus redshift! Astronomers 
know that SN are not found with such large values of z. Validation of 
this part of the message occurs at a higher level than simple schema, 
where the data model is constructed.