[Freeciv-Dev] Re: PATCH change_specialist

[Adam Olsen <rhamph@xxxxxxxx>]

On Mon, Dec 17, 2001 at 01:24:17PM -0800, Stewart Adcock wrote:
> Why not use the C99 boolean type when available (i.e. in gcc), which can 
> be checked in the configure script, and default to int, or short, or 
> something otherwise?
> 
> Maybe:
> #if _USE_C99_SOURCE
> typedef _Bool boolean;
> #else
> typedef short boolean;
> #endif
> 
> Then the compiler will check it, when the feature is available.

I think it's better to use typedef enum {false, true} boolean; since
then gdb can handle it.

> 
> Stewart.
> 
> Gregor Zeitlinger wrote:
> 
> >On Mon, 17 Dec 2001, Raimar Falke wrote:
> >
> >>On Mon, Dec 17, 2001 at 03:13:48PM +0100, Gregor Zeitlinger wrote:
> >>
> >>>>Whats wrong with TRUE and FALSE?
> >>>>
> >>>I just thought, if some parts of the code will be taken to Java, it makes
> >>>it easier, since Java has no preprocessor.
> >>>
> >>Can sed be used for this?
> >>
> >probably, but thats not a desirable way as it adds an extra unnecessary
> >step and reduces readablity.
> >
> >
> >>>>>+typedef int boolean;
> >>>>>
> >>>>A "can" or "is" in the function name is enough to indicate a boolean
> >>>>return type.
> >>>>
> >>>it's more clear that way. It follows the approach to make code more
> >>>readable. Thus, boolean indicates that only true and false are allowed 
> >>>and
> >>>any assignment statement to a boolean variable should be that way.
> >>>
> >>However in C the compiler doesn't check it. So it IMHO equivalent to
> >>write "return boolean" in a comment.
> >>
> >??? The C compiler treats it as an int, AFAIK.
> >
> >
> 
> 
> -- 
> ____________________________________________________________________
> Stewart Adcock   stewart@xxxxxxxxxxxxxxxx   www.stewart-adcock.co.uk
> Dept. Chemistry & Biochemistry, University of California, San Diego
> 4234 Urey Hall, 9500 Gilman Drive, La Jolla, CA 92093-0365 USA
> lab: +1 858 534 0956 home: +1 858 453 2577
> 
> 
> 
> 

-- 
Adam Olsen, aka Rhamphoryncus