[sc34wg3] Constraint application

Robert Cerny robert at cerny-online.com
Thu May 27 04:25:20 EDT 2010

On May 27, 2010, at 10:07 AM, Lars Marius Garshol wrote:

> * Robert Cerny
>> Let's say we have a name type NT for which there exists a scope  
>> constraint SC. Does SC apply to all instances of sub types of NT?  
>> As far as i understand the spec, i would say yes: [...]
> Yep. And the two bits of the spec that you quoted are indeed the  
> crucial ones that make it so.

Ok. That's good to know. So if i want my types (topic and statement  
types) to have names which may be scoped by natural language, i have  
basically two options:

A) attach the scope constraint to the default name type which is used  
in the meta schema for naming types
B) create a new name type which can be the thought of the super type  
of all language scopeable names.

Option A does not seem such a good idea with the above fact in mind.  
With option B the question arises if that name type should not be  
considered in the meta schema, since types cannot be named without  
using a natural langauge. If we keep it as it is, there will be two  
name fields on all types (not pretty :).


More information about the sc34wg3 mailing list