> Hi there, > > since we are building a BeOS clone (binary and source compatible), all > header files have to have the same name - so why not just use them=3F > The higher structures also have to be the same. Well, the Be header files are terrible. Not their contents, but the location of the definitions etc is so far from posix... > Instead of redefining ntohl() and friends, why not just including < > ByteOrder.h>=3F Well, I'd rather actually create new headers that follow the posix type standards, but that would mean having new header files that totally oppose Be's offerings, so really I'm not sure. > Is anyone else but mbuf.c using the macros that are defined in mbuf.h=3F > If not, isn't it better to put them in the .c file=3F No, mbuf.h and mbuf.c are OK I think. > > (BTW I really don't want to offend you, David, I am just asking :-) OK, not offended. We need to think about this :) To be honest at present what I've been doing is getting code in place that works, but may not be terribly pretty. Once I add a few more bits it'll be time to stop and consider the overall architecture :) BTW, I assume this means that you've got testread working? david