Benson Margulies writes:
Hmm. Why is it compiling that .c function with C++ const rules? We
try adding a cast. I'll look at this tonight.
Recent GCC has been really picky about that stuff, but it issues
warnings rather than errors.
I think it's probably a Microsoft bug that it's an error rather than a
warning, but even demoted to warning it's a pain in the neck, since
for most purposes in XEmacs treating chars as ordered is a bug in
itself. If the standards people really want to save the programmer
from herself, IMHO compilers should divide chars not into "signed" and
"unsigned", but into "ordered" and "unordered", and error on
comparisons of unordered chars.
XEmacs-Beta mailing list