Re: arithmetic issues William D Clinger 20 Jan 2006 18:12 UTC

> >> You've promoted semantics which are mere accidents of hardware
> >> implementation to the level of requirements. I firmly believe
> >> that any scheme code which relies for its correctness on an
> >> overflow or a roundoff is in error...
>
> >You are wrong about that.  The rest of your argument rests
> >upon that incorrect belief, so I needn't respond to it.
>
> Apparently you needn't actually make your points and
> convince anyone of them.  That's the beauty of the
> SRFI process of course; you can finalize regardless
> of whether you convince anyone you're right, so of
> course you're right about not needing to respond.
>
> 				Bear

We already know that SRFI-77 will be withdrawn before
finalization, so your snipe at the SRFI process was
entirely gratuitous.

As for not actually making your points and convincing
anyone of them, please go back and read the paragraph
to which I was responding.  Its first sentence is a
factually incorrect statement, baldly stated as fact
without any supporting evidence, which I have already
noted in previous messages.  Its second sentence states
one of your firm personal beliefs, also stated without
any supporting evidence, which happens to be factually
incorrect.

I consider it a waste of time to respond to people who
make arguments based entirely upon personal beliefs
that are not supported by any evidence, and whose
incorrectness has already been noted within the archive.

If you truly believe that finite rings and fields are
useless, then I would refer you to the implementation
of bignum arithmetic in Larceny, whose source code is
available to you.

Will