Support Books Under the Bridge

Monday, July 28, 2008

R.I.P., old friend

For several weeks, I've been putting off the inevitable: buying a new calculator.

In case you weren't aware, my day job may be goat-eating bridgekeeper, but by night I moonlight as a scientist. (It's tough to make ends meet, what with our little horde of troll-lings.)

It's my trusty TI-85. Actually, it's been dying for a long time; the screen dims within weeks of putting in new batteries. It's not quite dead yet, but I've admitted to myself that it's destined for the trash heap.

I've had this calculator about fifteen years. At the time it came out, the TI-85 and its cousins were really impressive. It could graph basic equations (very useful for pre-calculus) and do symbolic arithmetic. Nothing terribly revolutionary, but to be able to do this on an affordadable handheld tool... very cool. Of course at the time I mostly used it to play Tron in government class.

I used this calculator in university (engineering: advanced bridge design), graduate school (economics: modeling income flow as coupled differential equations describing populations of toll-paying goats), and in my work as a Real Scientist.

I really like this calculator. I'm not ready to give it up.

But being made of stern stuff, I decided to go calculator shopping. I remember it cost more than $100 when I first bought it (or rather, the Bank of Parents bought it). So I figured, Moore's Law, something similarly powerful should only cost a few dollars.

OK, maybe not that low: the price of a calculator isn't all in its transistors. But let's say, a regular scientific calculator might be $35. No way a graphing calculator could be more than $60, unless it has so much horsepower that it's practically sentient.

When I lumbered over to Ye Olde Staples, a wee tear in my eye, and found to surprise that they still sell the same calculator. Actually, I saw the TI-85+ (oooh, "plus") and a few other similar models, along with some crappy knockoff graphing calculators. As near as I can tell, it's the same calculator, with a slight style re-design. (Possibly there are some improvements in memory, speed and screen resolution?)

The asking price was about $150.

Shock! Consternation! It's been 15 years, and the price hasn't dropped! (!!)

There are some differences, but it sure looks like basically the same calculator. The profit margin must be obscene.

Did I buy it?

Well, the interesting thing is: I don't actually need a graphing calculator. A high-school student might still find a graphing calculator to be useful. The rest of us have better tools on our computers. If I want to do graphing, I'll use Kaleidagraph on my computer (Kaleidagraph is the best graphing program out there; sorry, Igor fans). Symbolic mathematics and calculus: hello, Mathematica (although Mathematica is fiendishly difficult to use).

In fact, I haven't used TI-85 to do any graphing in many years. A calculator is just not the right tool for the job. What I did use it for was entering long sums; you can write out many lines of your calculation all at once and double-check. Complicated mathematics is tough to do with a regular scientific calculator; if you make a typo, it's very hard to notice.

That's the real reason I hung onto my TI-85: the multi-line screen. Amazingly, I couldn't find anything usefully similar in the store. There was one four-line-screen calculator, but entering a long calculation on one line just ended up having it scroll to the right (instead of wrapping to the next line). Oh, that's so helpful folks; I just love doing calculations when I can't see all the bloody terms on the screen.

No, I didn't buy a replacement. The price was too high. Offensively high. Texas Instruments should feel free to envision me saying something rude in their direction.


I have other calculators. But they can never take the place of my TI-85.


Anyone who can suggest a replacement (impossible!), please leave a comment. (If your suggestion mentions Reverse Polish Notation, your comment will be ruthlessly deleted; let's not mock the dead.)


Billy Goat said...

Huh, I think I have an old HP calculator lying around, and maybe my old TI-82. They might not be in much better shape than yours, however, and I'd have to think about whether I wanted to give them up. Mrs. Gruff has an old TI-85, which you can use when you visit. :)

What do you think of Matlab for your graphing tasks? That's the only computer tool I have any experience with.

Lastly, why the hate on Reverse Polish Notation? It's just basically postfix notation, which has the benefit over infix notation of not needing parenthesis and operator precedence rules. :)

Keishon E. said...

I've had this calculator about fifteen years. At the time it came out, the TI-85 and its cousins were really impressive. It could graph basic equations (very useful for pre-calculus) ///

My goodness, when I saw TI-85, it brought back memories of chemstry and pre-calculas [shiver] I had one like it ages ago.


Keishon E. said...

spell check there "chemistry."

slinking away now...

Mister Troll said...


I've never seriously used Matlab. I've poked at it a few times and then found other ways to finish whatever job it was. And I haven't even poked in several years, so I can't really compare.

Mathematica (do you like how they've trained me to italicize the name?) is bloody difficult to use, but mind-blowingly powerful.

Now, as for the last part of your comment... I just KNOW you mentioned... that term you aren't supposed to mention... just to yank my chain.

I oughta delete your comment anyways, though -- I'll settle for threatening to ban you for being a troll. :-)

As for why the hate... because it's FREAKING USELESS!! (Duh.)

Here, calculate something like:

(1+(3 cos 0.2)^(0.3-sqrt(4)) * (16 / (ln(1+1/sin(4))+5)-1) ) / 1+(3 cos 0.2)^(0.3+sqrt(4)) * (16 / (ln(1+1/sin(4))-5)+1).

I'd rather type in the equation and spend a few seconds checking for typos. Good luck doing that after you've translated it to postfix.

Five internet bucks says HP calculators were responsible for this and this. (OK, OK, not really. But seriously, one has to check for typos.)

Billy Goat said...

@Mister Troll -

I talked to Mrs. Gruff, and you can have our TI-85 if you want it.

And haha, I was totally yanking your chain.

P. S. -- 1 3 cos(0.2) * 0.3 sqrt(4) - ^ 16 ln(1 1 sin(4) / +) 5 / 1 - * + 1 / 3 cos(0.2) * 0.3 sqrt(4) + ^ 16 ln(1 1 sin(4) / +) / 5 - * 1 + /

The hardest part of that was matching up your stupid parenthesis. :)

Billy Goat said...

@Keishon - Yeah, those old calculators sure bring back memories. I remember when they got banned from tests because people were entering text answers into the programs and looking them up at test time.

"Yes, Mrs. Spencer, performing this calculation is very important for answering who killed Ophelia in Hamlet. Really!"

Mister Troll said...

B.G., the issue is trying to sort out if you've entered the expression correctly. How easy is it to double-check that what you entered matches the given equation?

Ka-boom! Your space probe just blew up :-)

Keishon E. said...

I remember when they got banned from tests because people were entering text answers into the programs and looking them up at test time.///

Yep. [g] Those were the good old days.

Since graduating from college, I haven't had a decent calculator since then and I need one for work but my stuff is usually ratio and proportion type of math, nothing as complicated as graphing. [g]

Anonymous said...

actually, you expected the calculator to cost less than $30 and be as powerful/useful as your calculator from 15 years ago that cost $100. It was very interesting logic - similar to how you often think everything should cost a dollar and are then surprised when it fails to work as well as something you might have paid more for ;)

Mrs. Troll.

Billy Goat said...

@Mrs. Troll - Your post reminds me of the good old days, when I roomed with Mister Troll. Actually, I don't think he was quite as cheap back then. He didn't acquire his love of the dollar store until after he found his own place. However, in his defense, there is at least one thing you can find at dollar store that is well worth the dollar....

@Trolly - Double checking isn't that hard. I would say it's similar to double checking in-fix notation. *shrug* You just have to group things according to their closest operator.

As a side note, post-fix is actually easier to program for than in-fix, and you can even follow the technique manually. All you need is one stack data structure, and it's a cinch. And you can make your own stack on a piece of paper by hand.

As an example for the first portion of your equation:

1 3 cos(0.2) *



Pop the operator off, and then the next two operands. Push back on the result, giving you:

(Whatever cos(0.2) * 3 is)

From here, more gets added on, etc.

Mister Troll said...

How to program RPN?

B.G., you are *such* a troll!

Mister Troll said...

Oh, you mean pregnancy tests! Yes, I agree, those are good to get at the dollar store when you can find them.

(Yet another Helpful Tip from your friends at Books Under the Bridge, please express your thanks in cash.)

I'm not... entirely... sure what that has to do with anything, but all right :-)

I have been thinking about trying to post a bit about babies and kids and science fiction (or fantasy), in the same vein as my posts about religion. It hasn't quite gelled yet, though.