On Thu, Mar 11, 2010 at 12:10 AM, Jonathan S. Shapiro <[email protected]> wrote:
> Thanks for the pointer.
>
> Once again: string representation is not the issue at hand. The
> question is only what the size of the "char" data type should be.
>
I think char should be 32-bit (like C tchar in most of
implementations). Due to field alignment, it will take at least 32 bit
in most positions on stack and in structs anyway. So there is a little
point on saving bits.

AFAIR in Java they regret using 16-bit for char type. In the in
java.lang.Character class there are two methods, one for char other
for codepoint (see
http://java.sun.com/javase/6/docs/api/java/lang/Character.html). So
I'm suggesting going for 32-bit (UCS-4). Otherwise there will be
eventually both char and codepoint types in standard library, and this
make life of programmers unnecessary interesting (like it already is
in Java) ;)
_______________________________________________
bitc-dev mailing list
[email protected]
http://www.coyotos.org/mailman/listinfo/bitc-dev

Reply via email to