Re: htonl() htons()

[Hans Aberg]

At 14:00 +0200 2002/10/22, Vincent Caron wrote:
>>>From your description I surmise one can replace them with the identity, and
>> the code will work.
>>
>> Does anybody know what should they be on Motorola RISC CPU's?
>
>The PowerPC can run in little-endian or big-endian mode, but big-endian
>is preferred for performance reasons. That is, PPC actually run in
>network byte order, use identity.

I have a memory of that Intel & Motorola usually use the opposite
endianess, but I did not know Motorola was ACDC (i.e., can do it both ways).

At 14:03 +0200 2002/10/22, Hans-Bernhard Broeker wrote:
>The should be no-ops for the 68K family, AFAIK.
...
>In case of doubt, you might want to check the glibc sources for your
>platform.

I feel sure that my MacOS pre-X is big endian inheriting it from 68k, and
that is most likely what is used on MacOS X as well.

-- I did not have those functions at all, which is why I had to ask. I
figure they would exist on MacOS X, which uses GCC as OS supplied compiler,
so it is not much point writing specialty code for just me.

But it does not really matter, because I will not network export my tables. :-)

Anyway, thanks for the info.

  Hans Aberg