From: "Perry E. Metzger" <pmetzger@lehman.com>
Date: Fri, 21 May 93 16:41:07 PDT
To: "Kent Hastings" <kent_hastings@qmail2.aero.org>
Subject: Re: PI Compression
In-Reply-To: <199305212233.AA07890@aerospace.aero.org>
Message-ID: <9305212339.AA18403@snark.shearson.com>
MIME-Version: 1.0
Content-Type: text/plain



"Kent Hastings" says:
>                       PI Compression
> It may have been discussed here months ago, but my favorite bogus
> compression scheme is "pi compression". Any number like pi or 
> SQRT(2) can be represented as an infinite sequence of non-repeating
> bits (there are repetitive patterns, but eventually the sequence
> breaks out). A finite bit string can be represented simply as the
> starting location and length in pi.
> 
> Since all possible finite bit strings are, by definition, contained
> in the unending cavalcade of bits in pi, all literary works, radio
> programs, TV, 3D holos, feelies, etc for all sentient species from 
> the remotest past to the distant future, in every possible alternate
> universe is in little old pi.

Bull. You cannot prove that all strings are contained as substrings of
PI. The mere fact that a bit string is infinite and non-repeating does
not mean that it is wholely random. For instance, I can very readily
construct infinite sequences that do not contain arbitrary bit
strings.

See, as an example, this non-repeating bit string

101001000100001000001....

> Who would dare argue against these reasonable assertions?

Me.

Perry