Language change and code breaks

[Guido van Rossum]

> Is it possible that changes to built-in types to allow subclassing could
> be made to also allow redefining operators for built-in types ? 
> 
> Then "/" could be mapped to either behaviour, and gradual introduction
> could be a matter of changing the default behaviour in a future version. 

I'm currently working on subclassing built-in types, but redefining
operations for existing built-in types is not part of my agenda, for
various reasons (most of which have to do with boring implementation
details, but are real showstoppers nevertheless).

So alas, no relief from that angle.  Maybe we'll have to accept a
directive.

--Guido van Rossum (home page: http://www.python.org/~guido/)