Re: [AD] Allegro 4.2 todos

[Eric Botcazou <ebotcazou@xxxxxxxxxx>]

> Done, see attached patch.

Looks ok, except maybe:

+<b>AL_RAND</b><br>
+   Optional replacement for rand(), this macro actually uses the libc
+   function but does some rework on the result to guarantee more perceived
+   randomness in the lower 16 bits of it.

The last chunk of the sentence sounds weird to me, but I'm not a native 
English speaker...

In any cases, wait for Peter's word on the patch because he might trash 
everything :-)

> The current formula is:
>
> ((rand() >> 8) + (rand() & 0xff00))
>
> As previously suggested on this list. I've found the following formula
> works better under OSX though:
>
> ((rand() >> 16) + rand())

As Ben (IIRC) pointed out, this one is not guaranteed to give a non-negative 
result. We must clamp somewhere.

> I've tested other shift values between 8 and 16, but the best remains 16.

Then

	((rand() >> 16) + (rand() & 0x3fffffff)  ?

> So I wonder: is this due to the machine endianess or just OSX rand()
> implementation? I think the latter, but could someone test the thing under
> another big endian machine and report the results?

I think it is simply the particular implementation of the OS.

-- 
Eric Botcazou