|
Thu Apr 10 16:09:43 EDT 2008
the difference engine
--
This is awesome;
Some background:
It was used to tabulate polynomial functions, so it wasn't "uni-
versal" the way modern programmable computers are, but it was
still an incredible achievement for its time and glipse of things
to come. The theoretical model for computers as we know them to-
day were developed and proven in 1937 by Alonzo Church and Alan
Turing.
Modern computers represent Universal Turing Machines, and the
hardware model used is called the "von Neumann" architecture.
This architecture is pretty close to the conceptual "parts" of a
TM, except it uses random access memory instead of a read-write
tape. "Universal" means that it can be programmed to emulate any
other "machine". In this case, a "machine" would be better known
as a program today.
The difference engine is basically the physical representation of
a program that can calculate polynomials based on specific input.
In these terms, "programming" the difference engine is analagous
to putting input parameters into the program that the machine
represents. There is a specific TM (actually, infinitely many)
that can be made to function identically. In a universal TM you'd
write a program by which it emulates the appropriate TM, thus ac-
complishing what the difference engine does.
Babbage envisioned a more general machine, known as the analyti-
cal engine.
http://en.wikipedia.org/wiki/Analytical_engine
The original universal computers were mostly mechanical, using
relays and such. Here is a link to a hobby relay computer that
somebody took the time to build. It is not 100% mechanical, but
there's nothing stopping anyone from building one that is.
http://web.cecs.pdx.edu/~harry/Relay/index.html
Enjoy!
ps: if I seem like I enjoy this stuff way too much, it is because
I do; what is more remarkable is that I am actually starting to
appreciate this sort of historical perspective.
--
Powered by vee
Copyright © 2006-2008
|
|