Vi macros can simulate a Turing Machine -- try them, they work.

[Dave Sill]

David Hitz <hitz at mips.COM> writes:
>To win a bet I wrote a set of vi macros that let vi simulate a Turning
>Machine.  Since Turing Machines are universal computational devices,
>this should settle the editor wars debate for once and for all.

Turing machines are universal computational devices, but simulations
of Turing Machines are not.

>Other editors may have better user interfaces than vi, but they are
>certainly no more "powerful".

Ignoring the distinction between simulations and the real thing, this
might be true theoretically.  But "real-world" constraints on
computing resources and their users are such that the "power" imparted
by a program's interface may be overwhelmingly more important than the
capabilities of the program.  Having the power and being able to use
it are two different things.

>These macros have been tested on several systems, including versions of
>both BSD and SYSV, but since they depend on nits/bugs in vi, some
>versions of vi could break them.

Let me get this straight, you've proven that buggy versions of vi may
be more powerful than error-free versions?  :-)

=========
The opinions expressed above are mine.

"The Macintosh is simply a paper simulator."
					-- Ted "Hypertext" Nelson