Eric Pouech <[EMAIL PROTECTED]> writes:
> in a strict portability consideration, I agree. But, since most code is
> first written on Linux with GCC (thus meaning char == signed char), I
> really think this ia an easier solution to enforce char == signed char
> in the core Wine implementation.
That's the Microsoft approach...
Yes of course it's easier to enforce signed chars; it's also easier to
enforce that everything is little-endian, or that all ints/longs/pointers
are 32 bits. At least, it is easier until you encounter a machine where
these assumptions break down; then you are in for a lot of trouble
(ever wonder why NT on ia64 is still not ready?)
I don't think it's too much to ask that code that relies on the fact
that a char is signed uses 'signed char' instead of char. In the vast
majority of cases you don't care about the signedness of char anyway,
and enforcing it to be signed will just cause a performance loss for
no good reason.
--
Alexandre Julliard
[EMAIL PROTECTED]
