Anders Broman <a.broman@...> writes:
> > Should we use ENC_NA here too to prevent confusion?
> My preference is the opposite use ENC_BIG_ENDIAN as that is the
> "natural" encoding for the
> protocol and ENC_BIG_ENDIAN is less confusing in my opinion.
For what it's worth, I agree with Anders here. I tend to think of the endian as
being associated more with the protocol than with each individual field, so I
would prefer a consistent BIG (or LITTLE as the case may be) endian used
throughout.
I think this also has other potential advantages, e.g., if a multi-byte field is
incorrectly coded as a 1-byte field then later fixed, or if a single-byte field
is later expanded to a multi-byte field through a protocol update, the
endianness won't have to be changed from ENC_NA to ENC_[BIG|LITTLE]_ENDIAN. And
of course, the fact that ENC_NA is the same as ENC_BIG_ENDIAN only introduces
more potential problems for little-endian protocols.
- Chris
