Dear Think C User:

I have created some THINK C v6.0 projects illustrating the
use of HDF 3.2.  Since I only need scientific data sets (SD),
I only complied the relevant files.

Notes:

1.	I use 68881 compile options with Native Floating Point since this
gives the maximum speed.

2.  You must use FAR CODE and FAR DATA (32bit addressing) with the
HDF Libraries.

3.	I've noticed that with the release of MPW 3.3, that MPW C is 
approximately 10-15% faster in floating-point benchmarks that THINK C
v6.0.  (This didn't use to be the case.)

I have included some compare files indicating the changes needed for
compilation with THINK C.

I have also included some examples.  Note that you should create a 
new ANSI library (by copying the existing one) and rename the new
library ANSI-881.  Then, recompile with the 68881 and Native compile
options on.

Good luck.

Michael E. Mauel
Department of Applied Physics
Columbia University
New York, NY 10027
(212) 854-4455
INTERNET:  MAUEL@CUPLVX.AP.COLUMBIA.EDU