From: Kbwms AT aol DOT com
Message-ID: <9754ad9c.365867e3@aol.com>
Date: Sun, 22 Nov 1998 14:37:07 EST
To: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
Cc: ams AT ludd DOT luth DOT se, djgpp-workers AT delorie DOT com
Mime-Version: 1.0
Subject: Re: More rand()
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7bit
X-Mailer: AOL 3.0 16-bit for Windows sub 38
Reply-To: djgpp-workers AT delorie DOT com

Dear Eli Zaretskii,

On 11-22-98 at 13:13:16 EST you wrote:
>
>
> On Sun, 22 Nov 1998 Kbwms AT aol DOT com wrote:
>
> > Who generates this interest?
>
> We do ;-).
>
> Seriously, though: one of DJGPP's main goals is portability, mainly to
> Unix systems.  Every Unix box I have seen has [delmnjs]rand48 family of
> functions in its C library.  Which I believe explains why there is
> interest in having them, and why people ask about them from time to
> time.
>

Whose computer works in 48-bit arithmetic nowadays?  And what
advantages accrue when using 48-bit generators?

> > How does one verify that the *rand48() functions work correctly?
>
> This depends on where will the code come from.  If the sources are in
> public domain, or if a sufficiently precise description of the LCG
> employed by these functions is available, then we don't have to worry
> about performance, since the functions will behave like on other
> platforms.
>

Spoken like a man who has no intention of ever using the output of
a random number generator.  No serious investigator uses the output 
of a random number generator that cannot be identified and verified.

> As far as I can see, the multiplier and the addend of the LCG are
> described by the man page, at least on the nearest Unix box I saw, so it
> seems like functional testing should not be a grave consideration.
>

So, please send me the man page and I can start from there.


K.B. Williams