RE: [xml-dev] Are people really using Identity constraints specified in XML schema?

["Hunsberger, Peter" <Peter.Hunsberger@STJUDE.ORG>]

w3c@drrw.info writes:

> 
> Peter,
> 
> That functionality is really in the registry interface - and 
> is a designer support functionality - to return potential 
> candidate content fields - as you are creating the CAM 
> template in the first place.
> 
> If I look at something like OAGi BODs and search in their 
> data dictionary on 'date' 'billing' - I probably get about 10 
> or 15 hits, and then I need to decide which one - by looking 
> at the semantics around each - is best suited to my purpose, etc.
> 
> Seems like we are straying now into design-time v run-time.

I knew not explaining that example better would get me into trouble.
The issue is, that in order to know the patient population size you have
to first run the query.  Only post hoc do you know that it falls under
the rule of "too small of sample size can't be constrained by X".  To
compound the issue you need a query history to make sure that the
individual component queries aren't run separately (maybe even by
multiple individuals in close proximity to each other), since in this
case since that would give the same results.

It's got nothing to do with content, rather it's the problem that
privacy rules (business rules) can be violated by negative responses.
To work around that requires that you understand the context of the
query and the results in combination with each other. This is one of
those areas, where if you have to solve the problem, you can likely
justify throwing non-trivial amounts of hardware at it...

> Ooops - it must be Friday afternoon again!
 
:-)