Originally Posted by Markbnj
Sounds like you were curing an endian-ness problem by calling hton* on one side and ntoh* on the other. You were probably using a version of Unix that is big-endian. Windows and most versions of Linux on x86 architectures are all little-endian, so it shouldn't be an issue for most cases now.
One reason why I specified the date - demonstrates aging
Very few version of *nx existed at the time on a x86. we had SCO at the timeon a x86.
The files though were on some mainframe that were being used.
Would the Mac line have the same issue?