On Thu, 7 May 1998, Tony Duell wrote:
> > > I think a computer needs to have some kind of 'program' which is
> > > sequenced in some way, and also conditional branching. That _might_ be
> >
> > I think Tony just defined the distinction between a simple circuit and a
> > computer. A computer can act on logic that allows it to conditionally
>
> Defining 'computer' (either analogue or digital) is well-known to be
> non-trivial. There are far too many border-line cases.
>
> > branch. I'd add that a computer's function can be changed without
> > physically altering the computer (stored program).
>
> Hmm... In one sense I agree with you, and in another, I have problems
> with the fact that a circuit with the program stored in RAM or core is a
> computer, and the same circuit with the program stored in masked ROM (so
> you do have to make physical changes to change the program) is not.
> Marginal case again...
AHA! But I anticipated this objection and already have a response. What
I meant in the sense that the function can be changed without physically
altering the computer (machine) is that there is a definite distinction
between the circuitry to drive the logic unit (CPU) and the circuitry
which defines the stored program (code). So while a masked ROM cannot be
altered, you can swap out the ROM with a new ROM with a completely
different set of instructions, while the rest of the circuitry which
executes the code remains unmodified.
> > There's a fine distinction between a circuit and a computer and when a
> > circuit becomes a computer. I for one would not consider a filter to be a
>
> I quite like Philip's definition that an analogue computer combines 2 or
> more signals in a way more complex than addition. I can still think of
> borderline cases, though - is a 4-quadrant analogue multiplier really
> enough to be called a computer?
I don't think so. But then my arguments are mostly with regards to
digital computers and not analogue. I think the murky area being argued
here is strictly pertaining to analog computers since it is indeed a
series of filters and amplifiers combined to manipulate analog signals. A
digital computer is something quite different, and I think quite easy to
define.
Sam Alternate e-mail: dastar_at_siconic.com
-------------------------------------------------------------------------------
Don't blame me...I voted for Satan.
Coming in September...Vintage Computer Festival 2.0
See
http://www.siconic.com/vcf for details!
[Last web page update: 05/03/98]
Received on Thu May 07 1998 - 02:22:15 BST