27 Mar
2005
27 Mar
'05
5:04 a.m.
Rumor has it that Tony Duell may have mentioned these words:
Ah, but analogue computers don't perform logical functions....
And how much 'aid' does ther human have to provide? Most machines need to
be turned on ;-). Often a human selects what program to run and what
inputs to give it.
OK (and this came up in some private e-mail when
discussing the major
off-topic threads here). What is a computer? Becasue I certainly don't
have an exact definition.
I always heard it defined (and seemed the most logical/sensible to me) as:
"A computer is a device which can perform arithmetic _and_ logical
functions without the aid of a human."
Several of my earlier computing textbooks put it in this manner, explained
why it was so (see below) and it's always worked for me.
That said, a 4-function calculator was not a computer (if that's all the
underlying hardware could handle[1]), but just about any programmable
Yes, but to divide two floating-point numbers requires quite a sequence
of operations (you try working out how to do it on a machine without even
a full adder... [2]). Does that count as 'performing arithmetic and
logical functions without the aid of a human'?
[2] OK, not a 4-function machine, but this does describe the HP9100.
calculator was, if it had any type of control looping
abilities, as that's
the "logical function" aspect of the definition.
I think this illustrates my point. It's easy to find things that are
definitely computers (the PDP11 in the next room, for example, the PC I
am typing this on), and things that are certainly not computers (the
camera sitting on top of said PC). THe problem comes with the boarderline
cases -- like the programmable calculator with no looping constructs
other than an automatic return-to-start at the end of the program.
-tony