[Ur] Repeated datatype declarations

[Adam Chlipala]

Ron de Bruijn wrote:
> Op 12-10-11 13:44, Adam Chlipala schreef:
> > Haskell does not
>> have a solution to the problem, because its "module system" is so
>> inexpressive. It is not obvious to me how to follow your suggestion 
>> in Ur.
> The Scheme48 module system (which is based on Standard ML) is likely 
> also applicable to Ur: 
> http://community.schemewiki.org/?scheme48-module-system.
>
> It provides higher-order modules and there is no needless repetition 
> in sight.

I also don't think any module system for a language with only advisory 
typing is relevant.  Unless you can point to support for abstract types 
that are used in static type checking, then I'll assume Scheme48's 
module system provides no useful example for Ur.