Bruno Haible <[EMAIL PROTECTED]> writes: > Wouldn't a simple > > #include <unistd.h> > #undef chown > > have done it as well (assuming 'chown' is defined as a function, not as a > macro)?
Yes, I guess that would work just as well nowadays. I suppose I was reflexively worried about older systems that had "#define chown __something...." in their unistd.h, with no implementation of chown. These had something to do with changing uid_t width or with converting from non-prototypes to prototypes or something like that (sorry, I don't remember the details). You can get a glimpse of these old practices (though it's not exactly what I remember) in: http://groups.google.com/group/comp.sys.next.programmer/msg/7b6952619a4670e0
