i agree that we should have the old opcode for compatibility.  however,
for the new 31-bit generator, i think it would be better as an argument to
rand, randh, etc.  that way, fewer opcodes to learn!  or at least to
remember.

pete


jpff@maths.bath.ac.uk wrote:

> Message written at 9 Nov 1998 23:23:49 +0000
>
> There has in the past been concern about the quality of the random
> number generator used in the opcodes rand, randh and randi (which is
> totally different from the one used in the other noise opcodes by the
> way).  Clearly to change it would or could break old orchestras which
> is something I am trying to avoid.  I have implemented a rand2, rand2h
> and rand2i set of opcodes which use a full 31 bit generator.
>
> But... is that what is needed or would you prefer to see this as an
> optional argument to the opcode rather than a new opcode?  Or not at
> all?  I would like to see this in the next release, or the idea
> squashed as it has been hanging around for a long time.
>
> ==John