On 31. Jul 2005, at 05:14, Russell McManus wrote:
> I don't think that sb-sys:read-n-bytes returns early when there is not
> a full buffer to read. Of course, I could be wrong, so I've included
> a self-address, stamped example program.
This is indeed correct, and there's a comment at the top of the
source of read-n-bytes describing this. The stream code is twisted
and dark, and anyone entering there comes back mad (or at least,
needs a holiday).
(I'm back now, btw.)