[wg11] sub-supertype constraints on USED entities

[Ed Barkmeyer]

Jochen Haenisch wrote:

> I would think that the crosswise schemas are correct (except for a missing ;
> ). 
> ...
> Concerning P11: I think it allows both and does not give any usage guidance.

I agree.  There is no stated requirement that a chain of USE/REFERENCE 
interfaces be a directed graph.  And we see many examples of IRMs that 
REFERENCE FROM each other.

> The style, however, neglects the capabilities of P11 edition 2. So, I
> agree that your approach is of a much better style (except for a couple of
> syntax errors). The subtype_constraint was introduced for such cases.

Also agree.  I am not sure what the original model was trying to do, but 
the two schemas do not represent a good division of the concept 
Variable.  I think Lothar's model is better.  But I think perhaps in the 
generic_expression_schema, Variable should be declared ABSTRACT (or 
ABSTRACT SUPERTYPE).  The intent seems to be that all actual Variables 
will be typed.

-Ed

-- 
Edward J. Barkmeyer                        Email: edbark at nist.gov
National Institute of Standards & Technology
Manufacturing Systems Integration Division
100 Bureau Drive, Stop 8264                Tel: +1 301-975-3528
Gaithersburg, MD 20899-8264                FAX: +1 301-975-4694

"The opinions expressed above do not reflect consensus of NIST,
  and have not been reviewed by any Government authority."