This article doesn't actually directly address the question in the title. It observes that Texas Instruments calculators are still in wide use (due to network effects and some specific advantages over competing technologies), and that the calculators are a very high profit-margin product for TI. What I find somewhat more interesting is that TI is evidently in a monopoly position and has no real incentive to seriously lower prices or to produce a new, cheaper, and improved version of their baseline model.
I also find it interesting how relatively little attention is given to the advantage of tactile buttons:
> Plus, she says, her kids love the fact that the calculators are clunky and tactile, an opinion anyone who misses the tactile, QWERTY keyboards of earlier smartphones shares.
> "The learning curve is higher, but the buttons... they love pushing 'em," Yenca said.
Frankly, I've never seen a touchscreen interface for anything that could effectively compete with a well-designed physical interface for experienced users. Touchscreens are excellent for their flexibility but have serious drawbacks when used over a long period of time, when precision is paramount, or when there is a large space of inputs.
What's really needed is more or less the same physical input layout but with a larger, high-resolution screen. I would think the market in the US is sufficiently large that even if TI doesn't see the need to produce newer, cheaper and improved models, there's space for another player, say from China, to produce a better product.
The market is large, but TI maintains their monopoly because their calculators are entrenched in education. They will fight tooth and nail against the college board approving calculators for any new competitor, Chinese or not.
> I also find it interesting how relatively little attention is given to the advantage of tactile buttons
If I remember correctly, while in programming mode, almost all TI-BASIC keywords could be input by pressing a menu button followed by a number. You could be very efficient once you learned the combinations. By the time I started high school, I had learned the menus well enough that I was writing games in class while concealing the calculator under the desk. (Of course, as I found out later, my teachers weren't fooled for a second.)
Exactly. Everything in TI basic, least on the 82 I had way back in the day, is a keyword that you can pull in from a menu. Because of this, there was no need to remember what that function was named. Because of that, it was trivial to write all those little functions that just might be on the test, plus all those games and other diversions. I mean, what's a calculator without a game of craps on it?
While I agree that it's ridiculous that this is the case, I side with the students who were given a choice and chose the physical calculator. The thought of doing Numerical Analysis problems and long calculations on an app sounds excruciating, and I know that Accounting classes would have taken an order of magnitude longer without tactile feedback.
The replacement for a $100 problem is to buy students iPads and use an app and touchscreen? That's not a viable solution whatsoever for the vast majority of schools. The real solution is for schools to simply advocate using a cheaper and similarly functioned calculator. I used a Casio EL531 all through school and college and while it didn't graph, it covered me for everything in an Engineering degree.
Maybe TI should upgrade the processor in it a bit? Put a better screen? Maybe one based on e-ink? Imagine a LISP machine with dedicated calculator buttons, the size of a small kindle.
An iPad with an app is not the solution, but how come graphing calculators haven't advanced at all.
They DO have vastly improved and modernized calculators. The Nspire line has fast processors, lots of storage, bright color screens, touch capability, greatly improved software, etc. They don't really even cost more than the "old school" TI graphing calculators. TI can really only be blamed for not dropping the price on their old models. It would be misleading to suggest that TI hasn't worked on more advanced calculators.
Aside from the color screen and faster processor, these new Nspire calculators have no new features except gimmicky stuff (No one is going to buy a calculator because they need a 2"x2" spreadsheet). Even my TI-89, which I bought in 1999, has the same basic functionality. (Maybe that's why the TI-89's price hasn't changed -- it costs as much as the Nspire CAS). The only truly useful improvement TI has made is a rechargeable battery pack.
The NSpire OS is significantly more usable than the 83+ and 89 that I went through school with. In one case, I was able to put a dataset acquired from measurements into the spreadsheet editor, run a linear regression, get the r^2 for the regression, then draw the graph of points and line and save it in just a few seconds of menu choices - without ever having done that on an NSpire before. This would have taken me significantly longer on the older models if I hadn't already memorized the procedure.
So I wouldn't say that the NSpire has any radically new features, but I think that it's a lot more approachable then the older models. It feels less like something that you have to specifically learn to use and more like an intuitive software package - which is perhaps what our author is really getting at when they talk about apps.
> but how come graphing calculators haven't advanced at all.
They don't really need to so there is little commercial reason to develop better models. The old TI-81/TI-83 designs are still perfectly sufficient for the courses to a certainly level, and above that you need something much more than just a bit improved so you can run chunky software like matlab. There isn't really much of an exploitable niche in between.
> The replacement for a $100 problem is to buy students iPads and use an app and touchscreen?
When put exactly that way, yes that sounds silly.
But what about an app and a stack of much cheaper tablets? Not one per child but a pool for the school which are available to use in lesson time, exam time, and for the less well-off students to be borrowed for homework. Those that have their own Android tablet can just use their own and run the app on that, just needing a school one for exams (where they would not be allowed their own because they could bring notes into a closed-book exam).
For exam use they'd need to be locked down but that should be easy enough. Require that all students use a school provided one (so they can't hide notes and other such on their own devices) and re-image each before the exams start (so yes, some admin time to factor in here) with a base OS plus the desired app and all wireless locked off.
I mentioned Android above for cheapness mainly, if you are going to target one platform then for this use the one you can get the least expensive units for should win. Having said that, you don't need to go the native app route: build it in JS and you can run it on any device. For what the students will be needing even those cheap god-awful-slow tablets will cope quite nicely unless your JS and DOM manipulations are grossly inefficient. For the locked-down situation in an exam room "run" the app from local files and use the local storage options in the browser if you need to maintain state.
There would be an up-front cost (buying a job lot of tablets initially) and ongoing maintenance (a rolling program of replacing them as they break and some admin time updating/resetting them as needed) but you can get easily get 7" tablets for less than $50 each as single units and that price should drop when buying in bulk. For a smaller device something like https://www.kickstarter.com/projects/1598272670/chip-the-wor... might be more suitable (smaller screen but easier to pocket and has physical keyboard) when they are available.
And of course the schools don't have to provide them, if they already expect the students to have a TI calculator how is requiring them to have a cheap tablet or similar any different?
The key problems with this idea as I see it:
* Getting the new app-and-devices idea accepted as a viable alternative (the current graphing calculators are pretty entrenched).
* Writing the app in the first place, and then maintaining it. I doubt there would be a commercial impetus for anyone to do it so it would need to be some form of F/OSS arrangement or funded from education budgets. Maybe if I win the lottery and need a project to fill my spare time with I'll give it a go...
They are entrenched but there's also a lot of reasons that they've stuck around too beyond the inertia of the education system. Any alternative is fighting against a lot of we established training and knowledge. Teachers have been using them for years and the functions are well known and pretty easy to figure out.
TI graphing calculators just work in a way that will be really hard to match with more complex or newer devices like an Android Tablet running an app. The software has been ironed out over years of use and I can't remember ever causing a crash on my 83s or my 89. Plus there's no admin or reimaging required because anyone can go and reset the device to factory in a couple seconds before a test.
Don't discount their low power draw and instantly replaceable batteries either. TIs get weeks of use out of a pair of batteries and you know that they're getting low long before they run out so there's plenty of time to get replacement batteries. Compare that to any cheap Android tablet where 5 hours of active use kills it dead and it takes too much time to charge it back up before a test or class is over so you have to either a) replace it with a charged spare or b) get power to the kid's desk. That's a pretty big knock against having a pool of devices as an important component of testing or in class work.
Can anyone suggest a viable alternative for the TI graphing calculator? A mobile app would be great, but the ones I tried from the app store were pretty awful.
So far the best affordable alternative I've tried is... a TI calculator emulator. Which requires a ROM dumped from a TI calculator.
I quickly tried out the Desmos webapp suggested in the article, but it's rather basic and while it might work for a high school student, it's not a serious tool for scientific and engineering work that the TI can be. It does pretty plots but that's about it (I generally use gnuplot for that, and it has more useful tools than Desmos has, albeit a bit difficult to use and discover).
Back in my high school days, I was pretty bad ass with the TI-85 graphing calculator, to the point that I felt like cheating because I could easily use it to solve and verify most exercises and exams. I could have learned a whole lot more maths had I not had such a powerful calculator. And it wasn't even one of the fancy symbolic calculators that high school kids these days have.
The last time I put my TI through its paces was in university orbital mechanics class where I used it to do initial orbit determination of an asteroid based on three observations. That wouldn't have been possible without a powerful calculator, with variables in memory and such.
These days my calculator has fallen into disuse because I need it so irregularly that I don't have a good reason to carry it around. But it would be great to have something as powerful in my pocket at all times.
> Can anyone suggest a viable alternative for the TI graphing calculator?
It's not cheaper, but while taking some nightschool math classes a couple of years ago I bought an HP Prime[0], which was new on the market then. It's a truly excellent calculator which puts the TIs to shame. It's a calculator with a modern multitouch interface so you can use pinch-to-zoom on graphs, etc.
It uses the xCas engine[1] as a backend and has a spiffy calculator frontend, so you can write formulas as they would appear in a textbook.
EDIT: It also has an ARM processor, IIRC around 300MHz. It really shows when graphing complex equations, where the TIs take forever. I hear the newest TI nSpire ones have a slightly more modern CPU, but I really hate the interface there (it's no longer really a calculator for me, but a shitty tablet).
If you want an actual physical calculator, I strongly recommend this one.
I don't want another physical calculator but thanks for pointing me to Xcas. I haven't heard of it before, I must give it a go.
I've been using Mathematica for most of my computer algebra tasks, but it's not free. I've dabbled with Maxima and SymPy but neither are as easy to use or as powerful as Mathematica.
I think what is needed is something that works for high school students. If you need to do serious scientific and engineering work, programs like Matlab and gnuplot, which you mentioned, work, and even learning Python will probably give better results than an old graphing calculator that isn't high-precision. As was suggested in the article, a cheap Chromebook is very well capable of running Python, and cheap linux laptops could run Matlab.
Funny enough, that's what you get when you purchase an HP-49 (not even sure if it's still available, but it's what I bought for engineering school 13 years ago). The story goes that HP couldn't get a faster version of the (12MHz Saturn?) CPU in the '48, but wanted to release a faster version.
Since they couldn't get a faster compatible CPU but wanted to continue using their battle-hardened codebase, they decided to replace the CPU with a 200MHz ARM and write an emulator. I suppose that's maybe easier than trying to port a heavily optimized codebase to a new architecture. Plus, moving forward, you could write C code and compile it directly for ARM.
If I'm remembering right, I did manage to get a C "hello world" building with gcc before losing interest.
Wikipedia reckons that the fab they were using could no longer make the Saturn CPU, and I guess it was a lot cheaper to write an emulator at that point than to redesign it for a more modern process node.
Not at all what I'm looking for. There's no nice mobile spreadsheet software and if I have a laptop with me, I'd much rather use something like Mathematica or Maxima or Octave or Python+SciPy or Gnuplot depending on the task at hand.
Excel doesn't do symbolic computation and it's rather clumsy for dealing with mathematical expressions rather than arrays of numbers. Although I must admit that I'm not very good with Excel, perhaps it would be useful to spend some time with it (or the LibreOffice equivalent).
For some basic tasks, I can get along with Wolfram Alpha mobile app (it does plots and symbolic integration/differentiation), but it's not very good for long calculations where I need to store results from previous calculations.
If I ever find the time to mature KnightOS [1] to a state that's usable for the classroom, I hope to make my next project a calculator whose hardware is open source that can be sold for much cheaper, running KnightOS.
So, say you accomplish that. You create a calculator that you can sell for (let's say) $40, that blows the TI-84 out of the water. I believe in you, I think you could do that.
And say you're successful, and get some traction against the network effects discussed in the article. You start to threaten TI's money printer.
What do you suppose that TI will do? They charge $140 for the calculators, but because they can, not because they have to. So what will you do when they drop their price to $30? Or $20? Because they can, and that's why nobody bothers making any big moves in the market. It looks like TI would be easy to undercut on cost, but trust me—no matter how cheaply you can make your calculator, they can make their cheaper.
I know. I've heard this speech before. I can't compete on price at that point, but then I've won anyway - TI is no longer gouging the market. I don't think that they'd immediately shift, either, which gives me time to establish a foothold and keep the market competitive.
I mean, I'm all for that. If you can get them to drop their prices, you'd be a hero to a huge swath of the country. People could all donate 1% of the money you save them, and we could easily build a giant gold statue of you :-) So go for it, and I'll cheer you on :-) As I've said elsewhere, I really hate TI as a company.
(I did the math after posting that... a life sized solid gold statue of a human would cost about $47 million, so if you save $100 per person, for 47 million people—that seems doable).
> i had to get one too for high school in Denmark.
Was it a requirement, or was it just that the teacher recommended it because it's what he knew?
In my experience in a European school, the teachers generally recommend a specific model they're familiar with (usually casio, with some preferring TI) so they can answer questions easier. But I had no problems buying and using a different one (HP Prime) for the classes.
WFIW, in Luxembourg a TI V200 is required as part of the final secondary school exam for maths and economics. The V200 being one of the most expensive models.
Why that specific model was chosen, I am not sure, but I guess that for exams they want to have something standardised in order to avoid people cheating (all calculators are reset at the start of exam) or being at an advantage/disadvantage due to different feature-sets on different models.
You're not actually allowed to use graphing calculators in most exams here in the UK. Pretty much everyone's standardised on Casio's scientific calculators here for examination purposes, for whatever reason, and they're a lot cheaper.
In Poland, you're supposed to only use 4-function calculators both at local and national exams - to the point that I was accused of cheating by using my graphing calculator (and, in hindsight, it was cheating).
In 1993 I bought an HP48 for high school trig and calc. The TI-81 was pretty pathetic, and I think it didn't do calculus. The other HP user in my school didn't know the difference and probably didn't care - her engineer dad bought her a 48. Neither of us had any problems keeping up. While our textbooks didn't feature the TI, the instructors did, with a special version that was usable with an overhead projector.
I still have my HP48 on my desk and use it every day. I wonder how many of the other kids even care about the devices they had in high school.
I used a Casio FX250 scientific calculator, which I purchased for about $5 on sale in 1998, all the way from middle school through college, including not only high school calculus class, but also a large number of college math, science, computer science, economics, etc. courses.
For anything it couldn’t handle, a PC running Maple or Python was a much better tool than any pocket graphing calculator could ever hope to be.
Insisting students (or their parents) shell our $100 for graphing calculators is a scam, and spending class time teaching students to punch formulae into TI calculators is an immense distraction from learning mathematics.
A lot of the options kicked around are tablets running some app. Most suffer from some combination of these problems:
1) Unproven and unfamiliar software: the TI calculator OS just works, no crashes or bugs to really speak of especially in the simpler 83/84 models. (Though there are some in the 89 where you start to get into symbolic solvers and dealing with infinities.) Teachers and resource makers have also had a lot of time to create and learn materials for the older calculators.
2) Power: Calculators really sip power compared to any modern touch screen devices and with AAA batteries as the power your just a few seconds to go from a dead calculator (assuming you didn't bother to replace the batteries in the multiple days that the warning kept popping up) to a working device again. They also last much longer than any tablet out there and the batteries don't degrade in performance over time.
3) Physical interfaces vs touchscreen: There's a lot to be said for physical interfaces in speed and accuracy and the feedback they give. I've found them faster & more accurate in a lot of cases.
4) Admin overhead: This falls a bit under the unproven part of (1) but is a bit broader. Locking the tablets down and making sure that they're running just the base software and cannot communicate during tests is massively harder with a tablet alternative. Graphic calculators can be reset to factory default in well under a minute and it's difficult to put much in there beyond simple plug and solve programs to begin with. Plus free isolation because (generally, newer models like the NSpire are an exception) they can't communicate wirelessly to begin with.
None of these are necessarily show stoppers in and of themselves but they're a pretty high ramp for any competitor to top.
I've never found much use for calculators. The first time through my college science and engineering classes, useful calculators didn't exist (we all used slide rules). Now, I've almost always got a laptop with me when I have work to do so the python REPL is just a few keystrokes away, and for tallying a sum of a few numbers my phone apps work great.
Worse, the calculator is starting to show its age. "Perpendicular lines don't look perpendicular because the window is a rectangle," one Texas-based math teacher told Mic.
...what? If this isn't a misquote, I'm wondering whether it was really a math teacher they talked to.
It's not a completely unreasonable thing to say, at most it's phrased a bit sloppily. As far as I remember, when you set up the graph there's nothing guiding the user to use limits for the graph that result in a proper 1:1 aspect ratio. Instead if you e.g. want a graph from -3:3 on the X axis, you probably also set up the Y axis as -3:3 too. Since the real screen isn't 1:1, this will result in the graph being squashed down.
And except for the special case of lines that are parallel with the screen edges, a stretched or squashed aspect ratio would indeed distort the angles such that you can't tell which angles are perpendicular and which ones aren't.
In the TI-83, the graph does have several shortcuts you can use to fit the graph, one of them giving you a 1:1 aspect ratio and showing perpendicular lines as actually perpendicular.
The default range for graphs is (-10,10) in both dimensions, while the screen has a 3:2 aspect ratio. The default makes perpendicular lines that aren't on the axes stretched horizontally and appear to be at non right angles.
I think by default it plots from -10 to 10 on both axis. So if you plot y=x and y=-x together they will not look perpendicular on the rectangular screen. You can of course adjust the axes, but if I remember right it's not obvious how to make them match the screen. Graphing tools on a computer or smartphone probably have better default settings.
Actually now that I think about it, two straight lines, one at 45° and one at 135° would display this problem assuming that the display is not displaying the lines at the correct aspect ratio.
Yeah, that is certainly incorrect. Not only are perpendicular lines perpendicular regardless of the display aspect ratio (edit: as long as you set the projection axes correctly), you can adjust the displayed range in the TI.
And the pixels in the display are square shaped (unlike e.g. Commodore 64 in some modes), so there's no distortion from that either.
I'm not sure what you think the display aspect ratio means, but a 2:1 aspect ratio means that the horizontal axis of a graph will get scaled by a factor of 2 if the window is displaying a square section of space.
The lines y = x and y = -x are perpendicular. 2:1 scaling means the line y = x will appear on screen as y = 0.5x . A line perpendicular to that would have slope -2. But the line y = -x, which is actually perpendicular to y = x, will instead appear on screen as y = -0.5x .
> that the horizontal axis of a graph will get scaled by a factor of 2 if the window is displaying a square section of space.
Yes, exactly. But you can adjust the axes of the graph to get an 1:1 display (ie. x is -10..10 and y is -5..5, not a square space). And that is the default in the TI graphs IIRC.
So as long as you set up the projection to match the screen resolution aspect ratio, you'll get perpendicular lines correctly.
the texas instrument is also a great tool to start programming (TI Basic and then assembly), start understanding basic concepts of what a computer can do, it is a gateway to a whole new fascinating world and i'm pretty sure that it is due to the limitation of the machine that many will want to program. sometimes old is better.
I wrote my first "game" on a TI-83. Tictactoe. With an unbeatable "AI" (hardcoded if/else lookups for every possible move. the only "intelligent" thing about it was that it could "rotate" the board to minimize the lookups).
Then I downloaded some games from the internet and simply gave up on my TI-83 game dev dreams because I saw no way to compete with these "3d" games that are out there... Still a fun little experience.
I don't get that. The TI nspire is way improved from older calculators. It has a color higher resolution screen and a ton of functions and apps.
Unfortunately they removed the BASIC scripting. And there are a lot of features removed from it to keep it approved on standardized tests, like the ability to automatically solve equations and a QWERTY keyboard. But it's really good at graphing things and I loved it when I was in school.
At the time that comic was written, the TI nspire was only a few years old, didn't have a color screen yet, and was way more than $110 (still is, btw). It had almost no presence in the market because it wasn't yet being adopted wholesale the way the TI-83 family was.
As a comment on the site, I accidentally scrolled "past" the end of the story (e.g. hit some arbitrary marker whilst still reading and scrolling) and was presented by a full screen popup advert saying "Since you enjoyed reading this....".
I'm running NoScript, and I scrolled all the way to the bottom, read the whole article, and was not interrupted once with any popup adverts. Maybe install NoScript?
I tried a global script killer a few years ago and it made the web unusable, with having to "turn on" scripts on almost every site I used in order to do the most basic things.
It's also worth mentioning that I have both Ghostery and Adblock installed, and running, and this still came up.
I also find it interesting how relatively little attention is given to the advantage of tactile buttons:
> Plus, she says, her kids love the fact that the calculators are clunky and tactile, an opinion anyone who misses the tactile, QWERTY keyboards of earlier smartphones shares. > "The learning curve is higher, but the buttons... they love pushing 'em," Yenca said.
Frankly, I've never seen a touchscreen interface for anything that could effectively compete with a well-designed physical interface for experienced users. Touchscreens are excellent for their flexibility but have serious drawbacks when used over a long period of time, when precision is paramount, or when there is a large space of inputs.
What's really needed is more or less the same physical input layout but with a larger, high-resolution screen. I would think the market in the US is sufficiently large that even if TI doesn't see the need to produce newer, cheaper and improved models, there's space for another player, say from China, to produce a better product.