For the most part, I agree, but there are some exceptions. For example, bit fields in C/C++ are laid out using the byte ordering conventions of the target machine. This can make it nearly impossible to write low-level code to drive, e.g., hardware or communication protocols without conditional compilation. Of course, you can always shift and mask. The code becomes incomprehensible, but identical for big- and little-endian. For my money, using bit fields leads to code that is clear and easy to debug, and the obvious rationale of the conditional code makes it a winner. Steve