Re: RPN and TI calcs

[Prev][Next][Index][Thread]

Re: RPN and TI calcs

> I remember when I bought my first 'serious' calculator, around twenty
> years ago.  It was a TI SR56, and I liked it a lot.  In the user
> manual, it had a small section extolling the virtues of what they
> called Algebraic Operating System over RPN, and to prove the point,
> the manual gave as example a huge formula for evaluation, full of
> brackets and quotients and products and exponents, all sitting inside
> a giant root sign.  "Just enter it left to right" said the manual.
> After three attempts, I was still getting the wrong answer.  I
> happened to have an HP25 borrowed from a friend, and it produced the
> right answer first time.
>
> Strictly speaking, so did the SR56, once I'd twigged what my keying
> errors were; but the RPN system did mean that you entered problems the
> way you would work them out naturally, as opposed to the way you would
> write them down.
>
> I simply never earned enough at the time to be able to afford a
> Hewlett Packard calculator!

I never understood or cared for RPN until I read a page at the HP Museum
explaining it.  Once I realized that it actually is more natural (it's
rather ironic that this point seems to be something that has to be
explained to be understood), I picked it up in a few minutes and have
been wanting an RPN calculator ever since.

So, of course, I much prefer RPN much more than algebraic notation
(except in some cases, like entering equations to the graph editor or
solver, where I can't fathom any way other than algebraic...).  But I
don't want to spend the money to buy an HP48gx, especially as a few
months ago I bought a TI86 to replace a malfunctioning 85.

Of course, it'd be much cheaper to program some sort of RPN
shell/evaluator/interpreter/whatever.  I also think it'd be extremely
cool.  RPN on a TI!

I've seen one written in TI-Basic for the TI-85, and it unfortunately
seems unbearably slow on my 86 and rather inflexible.  I hope that if
done in Z80 ASM, it would be substantially faster, be able to save the
stack between uses, and even store the item on the top of the stack to
Ans upon exit...

Is anybody out there willing to make such a program?  Or is it safe to
assume that most people who program for the TI own them because they do
-not- use RPN?

- James
References: