Re: useful additional procedure for random bytes?
Bengt Kleberg 19 Feb 2002 09:57 UTC
> From: David Rush <xxxxxx@bellsouth.net>
> Date: 18 Feb 2002 17:20:14 +0000
> Bengt Kleberg <xxxxxx@cbe.ericsson.se> writes:
> > since most cryptographical
> > algorithms use not integers, but arrays of bytes,
>
> This opens up another whole can of worms, since Scheme doesn't have
> bytes. This is one of my pet peeves. I'm not terribly sure that going
> down this road is a good idea.
An SRFI with bytes might be a good idea.
> > perhaps a list (of specified length), with integers 0-255 as items,
> > would be the most versatile value of such a procedure.
>
> Except that you can already generate range-limited integers, so I
> think there's no real loss here.
I suppose you mean generating several integers in the range 0-255?
That is a sufficient solution, I did not think far enough.
bengt