View Single Post
  #17   Report Post  
Robert Swinney
 
Posts: n/a
Default Thanks for the tip on the HP32 calculator!

Yeah! I guessed it was something like that.

Bob Swinney
"Spehro Pefhany" wrote in message
...
On Sat, 17 Apr 2004 08:42:19 -0700, the renowned "Robert Swinney"
wrote:

Greg sez:
Yeah, but you can no longer buy an RPN calculator that will

comfortabley
fit in your pocket (except used, for big $). I really miss the HP42S
that I lost years ago. It was perfect for my needs and nobody makes one
like it now.


Hey! No problem. I was never "comfortable" with RPN all along. IMO,

RPN
was developed (HP probably)
to accommodate the shortcomings of early microprocessors.


I think it was developed by engineers with programming experience who
thought it was more efficient. It is. Forth and Postscript both are
stack-based computer languages. Languages such as C generally
re-arrange things at compile time so that they can execute similarly
at run time. That re-arrangement is somewhat non-trivial (recursion)
but that's not what AE calculators do- they just store intermediate
results and pending operators.

Eg. 2 * sin(43°) + 0.5

RPN AE
--- --
43 2
sin *
2 43
* sin
.5 +
+ 0.5
=

Eg. 4 * ( 5 + 7)

RPN AE
--- --
5 4
7 *
+ (
4 5
* +
7
)
=


It takes a
helluva lot more processing power to enter equations "the natural way".


Nah, the stack on early HP calculators was 4 levels. If you want to
use "the natural way", the calc chip needs to store one intermediate
result for each level of paren or implied paren, and the operator. So
it is just a tiny bit more RAM (to hold the operators) and probably a
few more intermediate results (depending on how many levels of paren
the calculator can handle).

Again, and IMO, those that became "comfortable" with RPN were deluding
themselves into thinking they were some sort of math geniuses for doing

so.

Entering things from the "inside out" rather than from "left to right"
always seemed easier to me, fewer keystrokes and less chance of making
an error. The results you need for the next step often seem to right
there from the last step. The fact that people would avoid borrowing
your calculator was just a bonus. ;-)

On the other hand -- I assume anyone that could be "comfortable" with

the
awkward back and forth motion of a slide rule might not agree. Now,

before
anyone jumps on me for that statement - I have several slide rules and

know
how to use them. Expertise with a slide rule did not dupe me into

learning
RPN - I chose to wait until AE came along.

Bob Swinney


Too bad no manufacturer (AFAIK) has offered RPN as a simply set
*option*- like radians vs. degrees for trig. It would add negligible
cost, and modern calculators can show several levels of stack at once.

Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
Info for manufacturers:

http://www.trexon.com
Embedded software/hardware/analog Info for designers:

http://www.speff.com