Hacker News new | past | comments | ask | show | jobs | submit login
The unstoppable TI-84 Plus: How an outdated calculator still holds a monopoly (washingtonpost.com)
59 points by jefflinwood on Sept 2, 2014 | hide | past | favorite | 75 comments



I'm the maintainer of KnightOS [1], an open source effort to build an operating system that runs on the TI-84+ (and other TI calculators). I've been tossing around the idea of also building open source calculator hardware for far cheaper than TI calculators (which are way more expensive than reasonable). In this dream I would also spend some time lobbying professors and teachers to break the monopoly and educating them on why open source is great. Is that something I should pursue? Would a kickstarter or something to raise funds for hardware be a worthwhile effort?

[1] http://knightos.org


It's really hard to get a teacher to change platforms (or textbooks or whatever) when they've got years of expertise and teaching materials built up on their current system. This isn't just laziness: you really are a less effective teacher in classes where you're making a big change like this. (You'd think "but calculators are all the same!" would make that moot, but the trouble is that the kids aren't experienced enough to get that yet, and as an instructor you're suddenly rewriting every single worksheet and hint sheet you have so the instructions match the new system.)

Mind you, I'd love to see the TI monopoly get broken! But this is what you're up against. And even if you have a great product and a teacher might be willing to make the change, they'll be really cautious about the risk that your product won't survive and they'll need to switch back to something else two years later.


It's also an issue of which calculators are allowed to be used in standardized testing environments. If you can't use the calculator on the AP exam, why did you learn it's quirks?


Most exams don't specify a brand but rather more tangible requirements like "must not have a QWERTY keyboard".


Must not have IR/wireless communication capability is a big one.


Id rather have a less good but up to date education rather than a really good education on 10 year old obsolete technology.


Why does anyone need an "education" in calculator technology? After class ends, no one will touch the calculator again.


> for far cheaper than TI calculators (which are way more expensive than reasonable).

They are. And TI makes tons of profit on each one. But that also means that if they got any serious competitor, they could drop their prices like a rock, and still make money. Certainly they don't want to do that, but I'm sure they would if they had to.


The hard part isn't the hardware, the hard parts are the software and the politics.

I'm pretty sure that I've used every single feature of the TI-83+ software during my 9+ years of education (grades 7-12 and undergraduate). If your new software is missing just 1 feature that the original software has, it won't fly with educators.

Also, even if it's a massively successful kickstarter, TI can still spend more on lobbying to protect its monopoly ;)

That being said, please try it! I'd love to see an open source alternative!


> Is that something I should pursue?

I can only offer an emphatic YES. I think that there are enough geeks out there that would like to see this monopoly broken that I predict a Kickstarter would be a rousing success.


Even if you build a decent clone, it will be hard to break the monopoly due to one key feature: reliable factory reset. Teachers will require a device that can be reliably reset of all user data to prevent cheating. If the student can hide or mask the factory reset, it will never be approved for testing, and thus will never be approved for classroom use.


https://www.google.com/search?q=ti+calculator+fake+factory+r...

Sounds like the actual requirement is to only be secure against people who never think to try.


I think I would explain that it's a placebo and instead suggest verbally banning students from using their calculators for evil, then discipline them when they do so.


It won't be for a while (KnightOS can't even do math right now, but it is a passable unix system), but I will continue the research and perhaps work towards that goal. The biggest concern I have is that I am only an amateur EE and I don't know if I could take people's money for something I may not succeed at.


It holds a monopoly because mathematics teachers teach things by rote and procedure, not the underlying concepts from experience. Any deviation from the non-TI norm is considered a teaching risk because the staff and students no longer know which buttons to press and in what order.

Perhaps 5% of people may develop an understanding past that.

I myself, as a calculator and math geek, after using just about every damn mid/high end calculator out there (TI82, TI83, TI89, TI nSpire CAS, HP48GX, HP50g TI92, Casio 9750G, Casio Algebra Fx) ended up using a shit Casio through school and my degree and Postgrad study. Conclusion:

Buy the best non-programmable Casio you can afford and leave it at that. My most useful device was £15 (Casio 991 ES PLuS). Its an awesome device. That and a mechanical pencil and some grey matter.

The TI dependency is a symptom of sickness in the system, nothing more.

Edit: found it. This is what I had: http://casio.ledudu.com/images/calculs/casio/machines/fx991V...


> It holds a monopoly because mathematics teachers teach things by rote and procedure, not the underlying concepts from experience.

Really? That was never my experience.

Indeed, algebra got a lot easier once I just started memorizing it. Years later I learned that portions of our system of algebra is largely an agreed upon system for ease of communication and that other algebras also exist for working in other domains.

(Taking formal logic also helped explain things a lot! Having to prove the commutative property from a couple of basic axioms is enlightening.)

All in all, the entirety of my public school math education in the US was highly focused on concepts and theory.


Here in Spain, in the school and high-school is forbidden using programmable calculators in exams. Even on doing the "selectividad", a pre university huge exam, is forbiden these type of calculators. I'm very happy with my old fx-85wa. It have all the advantages of a simple programmable calculator (natural math expressions and edit the formula that you typed), but can't store programs or functions, so isn't a programmable calculator. Plus works on a small lithium battery + solar cell. I only need to change the battery one time in this last 15 years.


Hmm. There could be a market for a "disruptive tech" in this segment: programmable calculators that look exactly like scientific non-programmable calculators, and behave that way to someone who doesn't know some (user-configurable) keystrokes to "unlock" the hidden power. When this is invoked, it changes the meaning of certain buttons, like say stats functions or whatever).

There could be some connector on the circuit board for easy programming via your PC (say over USB). This would not be exposed unless the plastic is taken off. With that, you could conveniently cram the calculator full of useful code prior to an exam, then access it in the special mode.

Now imagine this was done as a mod kit for several popular mass-produced calculators ...


I imagine that poorly funded schools in poor neighborhoods have trouble getting students to buy these things (or buying them for students who can't). I feel like plenty of teachers would tell their students to buy a cheap graphing calculator, just so that they could teach a class where everyone actually had the device. There's definitely a market for this.


I spent a lot of my middle-school math classes programming on my TI-86. This was in the mid-90's. My algebra books had a lot of BASIC programs that I translated into TI-BASIC. I had no background in programming at the time, and didn't even own my own PC for a few more years, but that early brush with programming (plus using Apple II's at my elementary school) helped set me up for doing computer science later.


2000s here, junior high to college on an 86 and then an 89 titanium - I remember when the 84 came out! It wasn't much of an improvement on the 83. But I learned a decent amount in my more basic math classes just by writing apps for my calc when I wasn't writing games. My husband and I rediscover our old calculators once in a while (83 silver, blasphemy!) and while they are pretty outdated, they were well built for what they were meant for.

I assume one of the big reasons they're still around is because they're not as useful as phones are, which is a big "benefit" for test taking and for minimizing distractions.


They're more useful because they have buttons too. It's virtually impossible to accurately enter numbers and expressions on a glass display with no tactile feedback.


It's not that bad; I mostly use an app to do quick calculations if I'm not at a computer. But I agree that for any real number crunching, it's worth using a physical calculator.


Yeah, I upgraded to an 89 in college and then had a math professor who did all sorts of programming with them. I still have the 89 and my wife has her 83. I can't bear to get rid of them :)

I think another reason they're still around is that there's just a lot of material written for them, decades and decades worth of programs and instruction notes.


I think I had the Commodore Amiga of that era... The HP 48g. RPN ftw.


Back in the day when I was still a land surveyor, the HP48GX was hugely popular in the field among staff surveyors because it was so easy to programmatically extend them or use a COGO card: http://www.hpcalc.org/hp48/science/civil/


Apart from the fact that it was sloooow, impossible to repair as it was glued together, the display was abysmal and you needed to drag the manuals around to be productive.

Not a big fan of HP nostalgia to be honest, despite rather liking RPL.


I dunno about slow. Once I learned its quirks, I became one with the stack, blasting in complex equations in RPN while my peers fiddled with brackets and mode keys.

Plus I'll admit to feeling no small amount of nerdy badassery at typing in what looked like alien gibberish to those around me while they slowly WYSIWYGed their way to an answer.


The UI is incredibly slow to draw. Many users may not have realized that keystrokes were buffered (up to something like 32 key presses) so if you knew your way around the menus and dialogs you could keep typing and the UI would eventually catch up. Unlike when a desktop computer's GUI lags, the 48 series would always behave as though the GUI had drawn instantly, so you could direct input to a dialog box or submenu that wasn't yet visible on screen.


My trusty 50G got me through High School and parts of College. Programming w/ RPL was something of an enlightening experience. RPN and RPL are what made the concept of 'the stack' finally click for me.

RPL was also fantastically simple, too. While other kids were struggling to copy the TI-basic instructions from photocopied sheets: I would quickly have a working program _plus a nifty little GUI._

The equation library and CAS were also incredibly useful when I was given the opportunity to use them.

---

The only downside is I couldn't swap any games w/ my classmates. That's OK, though, as I had 3D pinball.


My TI-83. First gaming system. First programming experience. Maybe first true love?


+1 my first program was a hilariously poorly coded pong game on the ti-83. A ridiculous mess of labels and goto's, but it worked! I may not have gotten into programming if not for the early introduction through graphing calculators.


I made a Mine Sweeper with it! Half the class downloaded it and played it in class!!!


I've been surprised that there hasn't been a good, programmable calculator app gain wide traction.


I left my HP42S in the drawer at home and now use the Free42 simulator on iOS. It's insanely handy.

http://thomasokken.com/free42/

I know some here have mentioned the 48G, the 42S was always a bit more useful and compact for my day-to-day use.


I used an HP41C for many years (replacing an HP55). I found it had a better key feel than later models. I bought an HP28S at some point but I never much cared for it and, to tell the truth, didn't need a graphing calculator at that point anyway.

I only used a TI for one year in college. This was during the period when calculator prices were dropping very quickly and between entering college and becoming a sophomore HP prices dropped to something relatively affordable and I never looked back.

It doesn't have the key feel obviously but I use an HP41 emulator on my iPhone.


As a denizen of HN, maybe you're the kind of person to have been exposed to "synthetic programming" on the HP41?


I think I actually bought a book on the topic at one point but never ended up doing a lot with it. Unfortunately my HP41 is pretty much dead at this point (battery leak and a flaky key) and it's one of those things I can't really justify replacing. (And, like assembly language, I think I'll probably keep HP calculator programming in the "it was cool at the time" bucket :-))


The phrase "synthetic programming" certainly brings back fond memories for me. Something about entering a weird but valid program instruction, then removing a memory module and power cycling to forcibly corrupt the RAM and build the initial wedge to then enter the synthetic codes?


One of the first things I install on all my phones is Andie Graph. It provides very good emulation of all the Z80 based TI calcs.

https://play.google.com/store/apps/details?id=net.supware.ti...


http://calca.io/ doesn't have graphing, but it's pretty wonderful otherwise.

(But FWIW, I'd rather see general-purpose programming become easier to use than calculators become more complex.)


TI afficionados should check out GraphNCalc83 for iOS.

Behaves like a modern calculator should (touch interactive, colorized graphs), no ROM shenanigans necessary, and a very familiar interface. All the TI-83 programs I've thrown at it have worked so far.


Graph89 is a port of a TI-89 emulator to Android. Works quite well. You need a ROM image from a calculator, but TI actually has them for download on the website, which saves effort.


You can't use phones to take tests.


This is a great example of a company building an ecosystem and catering over decades to that ecosystem to build an appealing computing platform - not for its technical merit, but rather for its community. It's analogous to the hold of Apple IIs on education in the 1980s.


Owned an 86 and a 92+. The 86 got me through all the courses that needed a non-qwerty TI-83-like calculator, but it had more than enough memory for me to type in all my formula, etc. for various classes. It was just weird enough that teachers didn't bother with it, but was close enough to a TI-83 that they let it go.

I spent hours programming it. I made an animation tool that would assemble a sequence of pictures (painstakingly drawn dot by dot) into a flipbook-like animation and a couple simple RPG games. It's a great first programming tool since the language is pretty easy and you have to be aware of resource usage since the memory is so tight.

My 92+ got me through my undergrad. Teachers didn't care since all of them required us to show our work and didn't allow any calculators on exams. But it was a great study tool, especially in Calculus. The symbolic system is fantastic and the pretty print algorithm they use is quite good.

Being a Motorola 68k-based calculator, it also ran some great games courtesy ticalc.org. There used to be a fantastic version of SolarStriker for it that I played relentlessly. It was a little like having an ultra portable Atari ST that just happened to have a very good CAS built in.

There's some fantastic work being done there in Lua for the TI-nspire calculator. like a SimCity 2k clone http://www.ticalc.org/archives/files/fileinfo/460/46069.html


I can't even begin to calculate how many hours I wasted playing snake or drug wars on my tI-84/TI-86 in high school in the 90s.


As I recall, even among TI graphing calculators, the TI-89 et al were vastly better, with their functionality for symbolic computation. Of course, nowadays, we all carry around vastly more powerful computers than any of these in our pockets, yet still standards limp on inertly and TI gets to charge a ridiculous amount for a suboptimal product.


The TI-89 destroyed my ability to do calculus. It was just too easy to plug in a problem and get an immediate answer. Great calculator, but be careful not to depend on it too much.


It sounds like the TI-89 vastly improved your ability to do calculus, to the point where it would be a shock to give up that improvement in ability. I would also say sine tables vastly improved people's ability to do trigonometry, back in the day...

People will often express concern about growing dependent upon these devices, but I am in just the same way dependent upon pen and paper, dependent upon dictionaries, dependent upon Google, and a million other things. In a world where access to automated computation is so ubiquitous, what is the harm in, well, growing comfortable using it?


A million times this. Not only that, the Derive CAS was buggy as hell so a good tutor can set some traps to see if it was you or the calculator doing the work.


The calculators with symbolic manipulation are often banned from standardized tests.

(I don't intend that statement as an endorsement of either calculators or standardized tests)


I, for one, endorse calculators wholeheartedly. I loved my TI-89, and I truly believe the opportunity to play around with it played a large role in my becoming a mathematician, of the sort which the "Calculators ruin math education!" brigade never acknowledges.


This was one more benefit to using an HP-48G. No proctors ever knew what one looked like, so they mostly just cared if it had a QWERTY keyboard.

Not that I ever used the symbolic manipulation, but the equation library was insanely handy from time to time.


When I was in high school and college, they typically only banned calculators that had full keyboards (e.g. TI-92). I believe I was allowed to use a TI-89 on both the SATs and the GREs.


It's a good product that will last a long time yet. You grab it out of a drawer and use it and put it back. It doesn't require any maintenance past some batteries changing occasionally. And it has buttons, real ones.

It's rarely about power either. Very little in mathematics requires much computational power other than a few joules between your ears. Then you're going to use a computer or CAS anyway.


Apart from having real buttons, I don't see how that doesn't apply to people's phones just as well. Nonetheless, I agree that the TI-89 is a great product. It's also a 16 years old piece of electronics; there's no good reason for it to be as pricy as it is. And the value proposition for the TI graphing calculators which lack CAS abilities is even far worse. The pricing does not seem to reflect a truly competitive market.


TI-83 is really sufficient for most engineering classes I've taken (Computer Engineering and EE courses) and many of my professors simply did not allow calculators. They pick problems that can be solved by hands, maybe with a few complex number tricks expecting students to know from doing homework or something they'd been picking up from lower-division courses.

But TI-84 is, honestly, the only calculator one will ever need. It has the extra functions that 83 doesn't provide and yet it's really enough do almost any practical calculations one have to be done by hands. The rest are just hypes.

The last reason is simply brand history. Once everyone owns 83 and 84 and everyone has good review about them, they go for the good review.


No person would ever willingly choose the TI calculator to learn or apply math, but they need them for school. It's a similar issue for textbooks, where the publishers can charge very high amounts, since the buyers (students) don't have a choice.

Really everyone should be learning to use a tool like Mathematica, not a 1996 calculator, but the education system advances slowly...

(Also: http://xkcd.com/768/ )


I used Derive on an HP100LX. I still keep two HPX00LXes running in case I ever need to do a bunch of symbolic math again. Best PDA/calculator ever for physics.

Funny how old calculators never go away. The HP12C is still used everywhere in finance and accounting.


What is preventing competitors from releasing a TI-84 Plus clone, where the buttons are named and placed nearly identically, with a very similar OS?



I think people in finances know of even more extreme example of calculator monopoly.

My dad still uses HP-12C, a financial calculator made by HP in the 80s (http://en.wikipedia.org/wiki/HP-12C). He's been using the same calculator for over 20 years. Extremely reliable and practical. It's no wonder people still get them.


I went to high school in Canada where we had to take a standardized exam for each course at the end of the semester. For our math course we were allowed to use an "approved" graphing calculator. Our teachers told us not to buy anything but the TI-83+, because they wanted to guarantee a "factory reset" before exams. Anything else would waste too much time for them.


Because schools require them. After that every sane person uses a computer with Mathematica/Maple/Matlab/Octave/etc.


In some districts it is because they require students to purchase them even if they are not used or are only used for simple arithmetic [1].

[1] http://2020science.org/2010/07/11/texas-instruments-graphing...


I never understood TI people; I used the HP 48GX from middle school through high school; by college it was usually permissible to use Matlab.

TI seemed to have good cheap multipacks for grade schools, thus getting them out there, but they were objectively inferior.


Article says TI-84 released 2004. My TI-85 (on my desk right here) is from 1996. I bought it for my Maths and Further Maths A-level courses (England). It's probably TI's fault that I'm a software developer now.


These things always remind me of [1]. Why are these things still that popular and quite expensive?

1: http://www.smbc-comics.com/?id=2582


And why Casion not makes a clone of TI-84 ?? Same layout of keys, same programming language. In any case. I never had a graphing calculator. I was happy with the typical cheap Casio calculator. The last calculator that I bought, when I was on school, is an old fx-85WA and it keeps working perfectly. For doing some quick calcs works perfectly, and if I need something more powerful, I simple open a python console or gnuplot on my computer.

I never saw that a teacher enforce us to use a particular kind of calculator, except the prohibition of using a programmable calculator for exams.


It's against the law?


Maybe. Some parts of the design may be copyrighted. I think the programming language cannot be, and patents, if any, have probably expired by now.

The problem then is, once you have the clone, you need to get the professors to know about it, and not requiring any specific calculator.

In any case, I'm not sure it would change much in practice. When I was at school, teachers didn't require any particular calculator. Half my class chose Casio, the other Ti. Two competitors are apparently not enough to drive prices down.


Copyright takes a lot longer to expire than a patent.


Copyright never expires: it gets extended faster than time itself.


I've always wondered what the cost of manufacturing would be for the the TI-84 would be today? Is there a BOM or estimates out there?




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: