OT: From the man who gave us cowcatchers...

John Francis johnf at panix.com
Sun May 11 13:56:04 EDT 2008

On Sun, May 11, 2008 at 10:11:26AM +0100, Bob W wrote:
> ...the Babbage difference engine number 2:
> http://news.bbc.co.uk/1/hi/technology/7391593.stm
> A quote from the article:
> "... in the 1930s where pioneers of the electronic computer age
> reinvented all the essential principles of computing largely in
> ignorance of Babbage's designs." 
> This sounds wrong to me. The essential property of a Turing machine is
> that it is universal - it can be programmed to become any machine -
> and Turing provided the mathematical basis which led to the later
> physical implementations. The Babbage machines were not universal
> machines, and certainly not Turing machines, nor did Babbage provide
> any such mathematical basis for future computing, and to that extent
> his work was a fascinating dead end.

That's not correct.  Although the Babbage Difference Engine was not
a general-purpose computer, the same can not be said of his grand
design - the Analytical Engine.  That did incorporate just about
all the features of modern computing equipment - programmability
(via punched cards), conditional execution, etc., etc.  It has
more than enough capabilities to implement a Turing machine (not
that that says very much; load/store, increment, decrement, test
and skip are sufficient).  It even had a hardware multiplier!

Even the Difference Engine incorporated parallel execution,
pipelineing, and micro-coding (though Babbage would not have
understood those words).

I've been immersed in Babbage quite a bit, recently:



My favourite quote comes from Doron Swade (curator at the
London Science Museum responsible for building the machine)

    Charles Babbage is equally famous for two things:

      o  Designing the first computer

      o  Failing to build it

> Anyway, it's an interesting article and it will be interesting for
> those of you in the area to see the machine.
> Bob
> -- 
> PDML Pentax-Discuss Mail List
> PDML at pdml.net
> http://pdml.net/mailman/listinfo/pdml_pdml.net
> to UNSUBSCRIBE from the PDML, please visit the link directly above and follow the directions.

More information about the PDML mailing list