Re: [RFA]: BOOL coding standards (Was: Problem with +numberWithBool: ?)

[Pascal J . Bourguignon]

David Ayers writes:
> I think we are side tracking here a bit.  The issue isn't whether we 
> should use "!= NULL" comparison with pointers or similar comparisons 
> with non-BOOL types.  I think in general non-BOOL (esp. pointers) should 
> use comparisons to attain a BOOL/truth expression to be used in for 
> if/while/for evaluation.  The issue is whether non-YES/NO values for a 
> BOOL are valid.

I'd say that they're not valid.

    - 100% correct and "conforming" code should not generate BOOL
      values other than in {YES,NO}.

    - 100% correct and "conforming" code should not compare a BOOL
      value, it should test it as: if(bool_value) or 
      if([obj isSuchAndSuch]), etc.

    - legacy code could compare a BOOL  value with YES or NO to detect
      "non-conforming" code.

    - the next version should use #define BOOL _Bool 


-- 
__Pascal_Bourguignon__                     http://www.informatimago.com/
There is no worse tyranny than to force a man to pay for what he doesn't
want merely because you think it would be good for him.--Robert Heinlein
http://www.theadvocates.org/