From mboxrd@z Thu Jan  1 00:00:00 1970
Message-ID: <d776c13a0a98c66024205920d0c663dc@quanstro.net>
From: erik quanstrom <quanstro@quanstro.net>
Date: Thu, 22 Jan 2009 16:40:17 -0500
To: 9fans@9fans.net
In-Reply-To: <13426df10901221253le795d82w6eb4bbf5e5e00be8@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit
Subject: Re: [9fans] fun question
Topicbox-Message-UUID: 84cb48da-ead4-11e9-9d60-3106f5b1d025

On Thu Jan 22 15:55:09 EST 2009, rminnich@gmail.com wrote:
> I ask this every time I hit this piece of code, I think. But what the heck:
>
>   if(seed <- 0)
>     seed = seed + m_31;
>
> what's wrong with this code?

supposing the code does what it is supposed
to on some machines, i would guess that this
could be a problem for sign-magnitude or one's-comp.
machines, since 0 and -0 are different.  if that's
the case, one would guess that the object is to clear
sign bit and -0 is the failure case.

then again, that's a very wild guess and one
would assume that sign = -sign would be
better.

-0 is a strange construction and the spacing
is misleading. it's interesting that atoi(2) and
the differ on "- 0".

- erik