On Mar 7, 1:28*am, (Don Klipstein) wrote:
In ,
wrote, in something about microprocessor products
being computers:
SNIP to here
Neither krw nor I ever said that a computer needs to include a
microprocessor. *The reverse is what krw claimed, which is to say that
a system that includes a microprocessor is a computer. * Unless
perhaps the microprocessor is being used as a doorstop. * For the
microprocessor to be of any use, it needs to be executing a program
and capable of some kind of input/output. * At that point it is a
computer. *It could be a very simple program only taking in some
serial data, figuring out what the data is telling it to do, then
activating the appropriate output. * But that is just a simpler
version of what your PC is doing.
SNIP a previously quoted mention of a non-microprocessor computer
I'd also point out that today, virtually all current computers do
contain a microprocessor or microcontroller. *Certainly evey one in
today's cars do. *So, why the trip down memory lane? * For the record,
I did google KURTA and KURTA mechanical computer and came up with
zippo.
SNIP a previously quoted bit on microprocessors doing stuff so simple
as "in my words" "arguably" "not actually doing real computer work"
As I said above, if a system has a microprocessor it's a computer.
It could be a simple one, that takes a few inputs and works a few
switches, but it is a computer. * That microprocessor is executing a
program. *Other than that the program is very simple, how is that any
different than a microprocessor operating in a PC? *And you never
answered this question:
How about I write a simple assembly language program that implements a
switch function,
turning a keyboard light on and off, *put it in a flash memory chip,
and replace the bios on my PC with it? *The light is now flashing.
Is my PC no longer a computer just because it's running a very simple
program?
Here's Collins dictionary definition of the word computer:
computer [k=C9=99m=CB=88pju=CB=90t=C9=99]
n
1. (Electronics & Computer Science / Computer Science)
a. *a device, usually electronic, that processes data according to a
set of instructions. The digital computer stores data in discrete
units and performs arithmetical and logical operations at very high
speed. The analog computer has no memory and is slower than the
digital computer but has a continuous rather than a discrete input.
The hybrid computer combines some of the advantages of digital and
analog computers.
b. *(as modifier) computer technology Related prefix cyber-
2. a person who computes or calculates
Collins English Dictionary
And from Merriam-Webster:
Main Entry: com=C2=B7put=C2=B7er
Pronunciation: \k=C9=99m-=CB=88py=C3=BC-t=C9=99r\
Function: noun
Usage: often attributive
Date: 1646
: one that computes; specifically : a programmable usually electronic
device that can store, retrieve, and process
Tell us what part of those definitions an 8051 running the dashboard
display in a car would not meet.
* Shorten many lines to a few, such as discernment of a computer from a
calculating machine, and saying that if it has "as-best-as-I-recall" I/O,
RAM and a program and a processor, it is a computer.
* What if the microprocessor is controlling a mere battery charger?
With the only program available to it being the one in its associated
ROM that was put there by the factory?
It is in fact a computer, an embedded computer. The fact that it's
program is contained in ROM is irrelevant. In fact, the vast majority
of embedded computers like those in cars, cable boxes, or your
microwave oven have their programs stored in some type of non volatile
solid state memory, ie ROM, Flash, etc. Did you look at the website
link I provided to the embedded computing site that is a technical
website dedicated to this type of computers?
* Yes, I would concede that it meets a definition of "digital stored
program computer" which is the "usual type of computer". *But to what
extent should a battery charger with program being burned into the same
IC package as the processor be considered a "computer" as opposed to
being a fancy sort of battery charger IC?
It doesn't matter a bit where the program is stored. All you're doing
there is moving more stuff on chip because it's cost efficient, less
components, less power, etc, to do so. Microcontrollers are
microprocessors that typically have:
CPU
ROM
RAM
Parallel I/O ports
On board peripherals, eg timers, interrupt controller, A/D, D/A,
UARTS, etc
That doesn't mean they are no longer computers, just that more
functions that used to be done with seperate chips has been brought on
board. If anything, it makes them more capable, not less so.
* How about if the microprocessor has included within its IC package ROM
(especially one-time-programmable "true ROM") a program that makes it
useful or at least advantageous only as part of a ballast circuit for a
single type or a small number of closely related types of metal halide
lamp or other arc lamp? *To what extent would one want to claim that a
metal halide lamp ballast or a fluorescent lamp ballast is a computer due
to having a stored program and a microprocessor of digital nature and I/O?
The ballast is NOT a computer. It does contain an embedded computer
though because it has a microprocessor running a program. Suppose I
take my PC and instead of booting an OS, I replace the BIOS with a ROM
containing a similar simple program to your ballast example. Is that
PC no longer a computer? No, it's no longer a functioning PC, but
it's still a computer.
* But I have digressed... *Getting back on track, to what extent would /
should one consider a lamp ballast or a battery charger, especially if
more-specialized, to be considered to be a computer if it has a stored
program, RAM, a microprocessor and I/O?
Again, it's not the device itself that's a computer. It contains an
embedded one if it has the things on your list.
* If a digital microprocessor controlling throttle in an automobile has
inputs both from any user controls and from any sensors other than user
controls, especially if it controls in addition amount of fuel injection
and/or ignition timing, then I would consider that microprocessor, its
program storage means, its likely-existing "memory" elsewhere (likely
RAM), and its inputs and outputs (short of the sensors delivering the
inputs and the devices responding to the outputs) to be a "computer".
* Any comments from here?
That computer is just another version of what is going on in the
battery charger or ballast. Yes, it has more inputs, more outputs
and it's program is more complex. Consider though what even a simple
microcontroller that is in a car controlling let's say the mirror
position has to do. It starts up and starts executing code. First
thing it has to do is program it's onboard peripherals. So, it
starts loading values into registers for the timers, the interrupt
controller. It loads values to set the speed of things like the
serial port. At some point, it may suspend operating and wait to be
re-awakend by an interrrupt which is caused by the serial port
beginning to receive an incoming command for it to move the mirror.
Now it goes into a loop to read in the packet of info. It has to
check the parity on the packet or other means of determining that it
is valid and not corrupt. Then it must break apart the packet and
figure out what to do, eg move RH mirror to position 7. I could go
on, but the point is that none of that is trivial. It involves taking
in data, analyzing it, acting on it. The fact that it can all be
going on in a single microcontroller that cost $1 is a remarkable
statement of how far technology has advanced. But it in no way
diminishes the fact that it is a computer.
Also consider the total confusion that would result from your approach
or that of CL. You're essentially saying that at some arbitrary
point, you consider an embedded microprocessor to be a computer, but
it depends on factors that no one here has clearly defined. I'm
saying that I've been in the industry and there is no confusion. Any
embedded application that has a CPU, executes a program, has I/O is a
computer. In a car, each of these modules would likely have one:
engine control unit, ABS control unit, airbag control unit, etc.
Now CL has made some remarkable claims here. Among them that the
typical car has but only two computers and many cars now have only
one. He's claiming that everything is being centralized. That is
completely contrary to everything that I have read and know to be
true. I'm still waiting for a single reference for this.
He also claimed that he knows for a fact that:
"Well, I happen to know that the engine and transmission controls
are
"one computer" . The climate control, air bags, radio, compass, etc
are another "computer".
I don't know how any manufacturer divides up the workload among
computers. But I would lay money on the above being false. I told
CL that I can think of some very good reasons why you would not want
the airbag function to be mixed with anything else, let alone the
radio. I never heard back from him. I think you likely know what
I'm getting at. I also know if you google airbag computer, you
sure get lots of hits and it sounds like there are a lot of them that
are seperate modules.
You don't have to believe me. I've given you a link to a very
credible source on embedded computing that says that a modern BMW or
MB could have 100 embedded computers and even a cheap car has a few
dozen.
Here's another good source. It's a book titled "Designing Embedded
Hardware" by John Catsoulis, who holds a Masters in Computer
Engineering.
http://books.google.com/books?id=vcR...mputer&f=false
"Computer systems fall into two seperate categories. The first, and
most obvious, is that of the desktop computer....The second type of
computer is the embedded computer, a computer that is integrated into
another system for the purposes of control and/or monitoring. In
fact, (the average home) may have 30 or more, hidden inside TV's,
VCRs, DVD players, remote controls, washing machines, air
conditioners......"