RE: [PyChecker-list] catching string as sequence errors

[Delaney, T. <tde...@av...>]

> From: Skip Montanaro [mailto:sk...@po...]
> 
>     >> Wow, that's even easier than I expected.
> 
>     Tim> Actually, this could for the basis for an optional 
> typed system for
>     Tim> Python ...
> 
> Which is why, easy though it may be to implement, it must be 
> done with some
> care.  If such a system is adopted by a lot of people (and 
> perhaps other
> tools), you'll have a more difficult time making incompatible 
> changes in the
> future should you need them.
> 
> Before going too far in this direction, I think would be 
> worthwhile at least
> skimming
> 
>     http://www.python.org/sigs/types-sig/

Oh yes - I know. I suppose I was just musing aloud. I'll probably work on it
at some stage, but have no fear that I'm going to pre-empt types-sig. A
large part of the attraction is seeing just what I can do with 2.2
metaclasses and properties ...

Tim Delaney