Martin Spott writes > "Curtis L. Olson" wrote: > > Martin Spott writes: > >> Maybe sometime someone finds the time to 'normalize' this interface to make > >> it platform independent ;-) > > > Could you be more specific? The routines already support network byte > > order, however there are cases (i.e. interfacing to an external perl > > script using pack/unpack, where network byte order is not desirable) > > so I'd like to keep this optional. > > O.k., I understand. But what am I going to do when the 'client' runs on a > different platform with different byte order where an int probably has a > different size (say a PPC750) ? You know that I can't decide to stick to one > plaform ;-) > It's just a thought with platform interoperability in mind. But if I really > start flight trainig this idea probably becomes obsolete because I won't > have the time and money to proceed with my project ....
This issue makes me remember the CORBA's protocol GIOP ( which I use at work ). It doesn't make any byte swapping unless it detects it has too ( the two station have different byte ordering ). This negociation is made at the connexion. So if you have two stations that are not in the network ordering, there is no penalty. Of course, the size of data is standardized so that a CORBA::Short is always 2 bytes and a CORBA::Long is always 4 bites ( there is no CORBA::Int ) -Fred _______________________________________________ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
