diferences between 22 and python 23

[Mike C. Fletcher]

Martin v. Löwis wrote:

>"Mike C. Fletcher" <mcfletch at rogers.com> writes:
>  
>
>>AFAIK, that's the plan.  IIRC, rationale was that there would be some
>>other type for 8-bit data, while all "normal" strings would become
>>Unicode strings.  
>>    
>>
>
>No. <type 'str'> will remain a byte string type for any foreseeable
>future. The only change that is likely to happen is this: To denote
>bytes > 128 in source code, you will need to use escape codes. 
>  
>
Sigh, yes.  Bulks up resourcepackage files, but oh well.

>A change that might happen in the future is this: A string literal
>does not create an instance of <type 'str'>, but an instance of <type
>'unicode'>. However, IMO, this should only happen after a syntax for
>byte string literals has been introduced.
>  
>
Sorry, yes, that was my understanding, I should have specified a "string 
literal", rather than just saying "strings" would produce unicode.  And 
yes, I'd definitely suggest not getting rid of the current semantics 
until there's a way to do byte-strings.

Peace all,
Mike

 
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/