Re: Reasonable Assumptions Regarding sizeof(char)
Niklas Matthies wrote:
On 2006-11-27 13:34, Le Chaud Lapin wrote:
I'm wondering: Why would you want to use char (instead of int) for a
big-integer class?
I am using unsigned int, based on feedback in a concurrent thread
discussing what is the most "natural" unsigned int for a class.
Every data-oriented class in my library has built-in serialization
support. I had always assumed that I could circumvent size and
endianess incompatibilities when doing inter-machine serialization
until yesterday when I realized that a machine that does not have
8-bit-byte-addressibility will be a real pain.
That got me thinking beyond my class Integer, about my other classes,
and I realize that, without the 8-bit-byte assumption, my framework
starts to fall apart.
I am a huge fan of portability, but rationality must also come into
play - I am not about to go chasing afer 9-bit bytes for portability's
sake. OTOH, it would nice to know of any major surprises. DSP code, I
think, is exempt. But I would be very interested in knowing if any
general-purpose CPU's did something other than 8-bit.
-Le Chaud Lapin-
--
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]