I'm attempting to read a file that is composed of complex 32-bit floating point values on Solaris 10 that came from a 64-bit Red Hat computer.
When I first tried reading the file, it looked like there was a byte-swapping problem and after running the od command on the file Solaris and Red Hat computers, I saw that they were indeed swapped (every other byte). So, I wrote a little Java program that would swap every other byte and the result was the at the byte-level, the files were the same on both computers. The following is the result of running 'od -x' on the file, showing the first 32-bits which represent the real portion of the first complex number:
f8a6 be03 cd76 bc99
What's weird is that when I run 'od -f' on both computers to see the 32-bit floating point value, I get different results! On the Red Hat computer I get: -1.288782e-01 and on the Solaris computer I get: -2.7055474e+34 
. How can I have the same byte level file yet come up with completely different values when I try to print out the bytes as 32-bit floats?
This is really messing me up trying to write a C++ program to read this file on the Solaris computer. I'm getting the same value in my C++ code as what I see when I run 'od -f' on the Solaris computer. I've tried just about everything that I can think of and haven't been able to get past this.
Does anyone have any idea on what's going on with the od command between the Red Hat and Solaris computers? Any ideas on how to get my C++ code to give me the correct value for the 32-bit float?