> > I think allowing a way for even C++ compilers to use char for bool (which your proposal would not cover) is a good idea. (I know I’ll use it; I also suggest the smokers use it.) > > When you sent your last message, I had already written the attached patch. What do you think of this approach? > > Sorry, I didnt mean to imply my approach was the whole solution :-). I was suggesting that as a part of your approach, instead of moving the mach include to perl.c, the set caret x function and the mach dependency be moved together. So, instead of moving it from the header to perl.c, it would be moved to its own c file, so perl.c would still get whatever handy.h sets bool to. (I hope thats clearer) --Sent from my phoneThread Previous | Thread Next