PEP 285: Adding a bool type

[Martin v. Loewis]

Peter Hansen <peter at engcorp.com> writes:

> But your point in the message I was replying to was that there
> are 300 functions which have code that says "return 0" or "return 1"
> or "return None" and the change will help users of those functions.  
> Didn't that imply those users are not looking at the documentation,
> but are trying to read the source to figure out how the code works?

No. It implied that users read the source of the code; whether or not
they look at the documentation was not implied.

> If the documentation must be changed, why not just make it clear in
> the documentation what it means when a method returns 1 or 0 or
> None?

The documentation already says that. However "If the section in fact
existed, return 1.  Otherwise return 0." leads to the question "why
aren't you returning a boolean value?", to which the answer is
"because they are not supported in the language".

> True, False = 1, 0
> 
> Or, now that I think of it, will that continue to work as these
> will just be built-ins now hidden by my local definitions?  

Indeed, yes.

> If so, my bad for mentioning code breakage.

There still might be code breakage, but it will be much less frequent
than you feared.

Regards,
Martin