PEP on path module for standard library

[Duncan Booth]

Peter Hansen wrote:
> Just a note, though you probably know, that this is intended to be 
> written this way with path:
> 
> >>> p / q
> path(u'a/b/c/d')

I know, but it really doesn't look right to me.

I think that my fundamental problem with all of this is that by making path 
a subclass of str/unicode it inherits inappropriate definitions of some 
common operations, most obviously addition, iteration and subscripting. 

These operations have obvious meaning for paths which is not the same as 
the meaning for string. Therefore (in my opinion) the two ought to be 
distinct.