Hacker News new | past | comments | ask | show | jobs | submit login
Calculator hackers crack OS signing key, opening a closed platform (ticalc.org)
111 points by modeless on Aug 17, 2009 | hide | past | favorite | 41 comments

A man has just become immortal. TI scene will never be the same again; the TI-83 calculator turns to an open platform today (The 83 is mandatory in U.S. education and sometimes distributed for free, all American High Schoolers use it, and did so for at least the last 10 years.) Bonafide 3rd party OSes can be loaded on it now, no more shells.

Help crack the rest of the keys (Win32 and Linux clients; linux version requires X I just found out):


Some fun statistics:

- The factorization took, in total, about 1745 hours, or a bit less than 73 days, of computation. (I've actually been working on this since early March; I had a couple of false starts and haven't been able to run the software continously.)

- My CPU, for reference, is a dual-core Athlon64 at 1900 MHz.

- The sieving database was 4.9 gigabytes and contained just over 51 million relations.

- During the "filtering" phase, Msieve was using about 2.5 gigabytes of RAM.

- The final processing involved finding the null space of a 5.4 million x 5.4 million matrix.

> (The 83 is mandatory in U.S. education and sometimes distributed for free, all American High Schoolers use it, and did so for at least the last 10 years.)

It's a bit off the subject, but this has always struck me as a terrible scam. For the material it is used for (introductory calculus) and the way it is used, one of these TI calculators hurts student learning and understanding more than it helps, and there is no excuse for schools to force students to purchase a $100 piece of electronics without a damn good reason. That these devices are absurdly overpriced compared to the state of the art (~$100 multipurpose netbooks, for instance), and that the models required are made by a single company with monopolistic coordination with standardized tests and textbooks, only make things worse.

Learning how to use a calculator is perhaps a useful skill, but it's one that can be learned quickly if needed and should not be pervasively required: for most math done for math's sake (that is, as opposed to computations for some engineering problem), students would be better off if teachers instead made problems that could easily be worked with pen and paper.

(Personally, I made it through high school and then up through upper-level undergraduate mathematics, physics, econometrics, etc. courses without ever buying a graphing calculator, so it’s clearly not strictly mandatory for current curricula, but students and parents get the impression that it is essential, and the vast majority of students do buy them, so the effect is about the same.)

And now a cracked OS allows cheating since they have an approved 'test mode" to prevent access to stored files.


So blinking the led without any access blocking will be an early task for the subversive TI hacker.

Crazy what they have these days. Back when I was in high school (first year they required the 83) our teachers simply required we showed every step. So we could get the answer with a few key presses, but unless we showed how to get the answer we wouldn't get credit. Of course I wrote programs to solve the problems step by step and print out the answer, but I guess the system works because in the process of being clever and writing the programs I of course learned how to solve the problems step by step.

I had a nice and crazy little Chinese professor for numbers theory in college. I never knew what he was saying, but he would tell us the questions for the exam the day before. I asked him if I could use my ti-92+ and he said yes, so I just programmed in brute force algorithms, entered in the data, then sat back while the 10MHz processor crunched.

Cue the next round of profits for TI, after all, now everybody will have to buy a new and more secure model.

Or, at least, it would if anyone used the Nspire. I've never heard of a school using them - everyone has standardized on the 83 / 84.

However, it looks like that test mode doesn't apply to the model that was cracked.

>(Personally, I made it through high school and then up through upper-level undergraduate mathematics, physics, econometrics, etc. courses without ever buying a graphing calculator, so it’s clearly not strictly mandatory for current curricula, but students and parents get the impression that it is essential, and the vast majority of students do buy them, so the effect is about the same.)

It's actually much harder to make it through high school math without a calculator than the upper-level stuff. At the end of my senior year in college, I tried to fire up my TI-89, only to discover the batteries had run out, and I had no clue when that happened.

I'm not so sure about this. In high school, I spent the first two years doing everything using a TI-83+ (Algebra II and Pre-Calc). However, when my TI-83 stopped working (and got a TI-89), I stopped using a calculator almost all together. For both years of Calculus, I did all the work by hand, turning to the calculator only at the end when I needed a final result (and usually tried to do that in my head/with pen and paper anyway).

I would say the that helped me immensely with understanding not only understanding the material, but being able to work it out faster. I'm now able to roughly estimate (within a decimal point or two) most basic math problems in only slightly longer than it does for most students to do it on a calculator. In fact, I tend to work problems out significantly faster than most students who simply punch everything into a calculator and get the results from there, if only because I understand the problem better from working it out by hand and know better how to approach it. Being able to estimate square roots in my head has been vastly beneficial, and would be a skill well worth knowing.

Yeah, I never saw the point of the TI-83.

The TI-89 is nice, but there is no point in having your CAS running on a handheld device anymore. Might as well give everyone a computer with Mathematica.

I have the exact opposite reaction to graphing calculators. I used them to ace my SAT IIs back in the day simply because I understood all the functions of my calculator than every other non-hacker in the room. I even used to program small BASIC games on it when I was bored in class. There still isn't a good device to program on that is inconspicuous in the classroom.

Holy Cthulhu, you can crack 512-bit RSA in 73 days on a single computer these days?

On a single ridiculously slow computer. Even the slowest Core 2 Quad would be able to factor 8x faster.

Could you do this for the ipod nano, 2nd generation? I've been wanting to get RockBox on it for years, but the firmware is encrypted!

Now I'm curious. First of all, any random Core Duo could probably put a significant dent in that timeframe if all he had was an Athlon 64. Second, what would the effect of CUDA be on this? It was extremely useful in that MD5 contest earlier.

The engineyard contest? They had CUDA participants iirc.

Do you have any refs for the TI83 being mandatory in US high schools? A quick google search puts your post at the top and nothing too relevant anywhere else. I graduated from a public HS less than ten years ago, and I used the far superior HP-48G throughout the entire thing :) I never took math classes from my HS, but I do remember my friends who were taking HS courses having an assortment of TI81 - TI89 calcs.

I don't have any references for you, but my middle school and highschool required graphing calculators. And they very strongly pushed the ti-83+, claiming that students would be horribly lost if they couldn't follow along with the teacher's instructions. In reality the teachers were all incompetent with the calculators and the school wanted everyone to have an 83+ because that was the only platform on which the teachers knew the process for erasing the memory.

I can correlate the "strongly recommended TI part," having graduated HS in 2001. I used an HP-48GX and yes, at times, it was hard to adapt the teacher's instructions. My friend and I both had 48's, and we helped each other work it out. When we got into symbolic manipulation, statistics, and inspecting 3D graphs, the HP left the basic TI models in the dust.

But yeah, no problem using a HP graphing calculator when I was in high school; took trigonometry, AP calculus, AP physics, and the SAT all using that calculator. Even had a full year's trig notes stashed in there, though I honestly didn't use them on any exams; just handy for calculus homework.

It doesn't take much to become immortal does it?

don't you dare piss off a generation of calcers, jacquesm ;-) this is our holly grail, there are many others like it, but this is ours.

[Edit: inserted smiley so not to appear "touchy" :-P ]

Not trying to piss of anybody (wow, you're touchy).

I just think that you're putting the bar for immortality a little low here. Brute forcing a key is not exactly rocket science.

Congratulations though :)

edit: smiley duly noted.

edit2: now if only the downmods would grow a sense of humor too :)

Running the general number field sieve is not exactly brute forcing. IMHO, understanding and tuning an algorithm as complicated as the GNFS is pretty close to rocket science. Not as close as actually inventing it or writing the code yourself, but this is more than just writing a short program to try all combinations.

Understanding and tuning an algorithm once somebody has devised it may be complicated, but coming up with the method in the first place is where the real work lies.

Anybody can do cool stuff with the jpeg library for instance, but the first guy to take DCTs and apply them to image compression was the real hero.

Unfortunately it is very rare to see the people that come up with an algorithm (if it isn't named after them) achieve 'hero' status (Hoare ? Knuth ?) other than in the circle of hackers.

Implementing and applying those algorithms and getting publicity around it will do wonders though.

Anyway, I seem to be unable to get my point across so I'll leave it at that.

Edit: To put it in one line: Hurley and Chen are 'heros', Tom Lane is a footnote.

But without Tom Lane Hurley and Chen would be a lot less rich than they are today.

Hurley and Chen are YouTube founders. Tom Lane organized the Independent JPEG Group, which isn't the group that designed JPEG (that group is called "JPEG") but rather an author of some important open-source JPEG software. What's the relationship?

He didn't implement the GNFS; he used GGNFS and MSieve.

Might be fun to try making a javascript workers version. While it is not as powerful I am curious how many more users you can get.

Unfortunately it's a ridiculously shitty hardware platform to be developing for in 2009. It's a motherfucking Z80, for $100 -- a vintage Game Boy minus any decent sprite graphics hardware, with enormous profit margins and a captive market (enforced by regulation) of a million a year.

Yes, the OS on the TI-83 is dogshit, but I have a feeling that it's pushing the limits of what's reasonable on the hardware. The TI-89 has a terrific input language and OS (almost mac-like compared the the apple-II-ness of the 83), but running on a m68k.

Who knows, maybe someone can at least make a firmware that not only forces you to close your parentheses, but gives you some benefit for doing so.

But... it's a calculator!

And it has a headphone jack! And a bitmap screen! And a full keyboard!

The post on TI forum reminds me of when decss was cracked and an email was sent to the list saying something like: "Here are some random numbers that are good for initializing your card game shufflers and stuff like that" Followed by several dozen keys. It got printed out and put on a dorm room door :)

Oh those good memories of the TI-83, my first and only foray into the world of assembler programming, trying to cram a complex algorithm in 6 registers, the good times!

You mean five registers, all of them accessible through the sixth! I had a nice hand written note for register allocation for the Z-80; I used an stack-based IR and mapped the TOS to #A. I hope to god my parents haven't trashed my stuff, would love to see those perversions again :-)

What about my oldskool TI-86? No love for that? :(

I guess I have a little taste of what the guys who went thorough school using slide rules felt like when calculators that could do trig functions came out.

I went through high school trig, calculus, and undergraduate college math classes with nothing more than a standard 10-digit "scientific" calculator. Graphs? You plotted those with a pencil.

What's the legality of this, I wonder? (Toying with the idea of joining the compute cluster or not)

Isn't this at least technically illegal? (DCMA)

You mean DMCA, and IANAL but probably not. The calculator's OS may be a copyrighted work, but the calculator itself is not.

hehe, I remember trying to convince my teacher for letting me use the nintendo DS on exams. You have to try, it's funny to see their faces.

Now you can use: 1.Nintendo DS. You can make or port(from linux) whatever you want in c, import Lapack libraries, use eigenmath homebrew for everything else. 2.Iphone.Good support for c too. 3. netbooks...

So, we can get a native MirageOS? Nice.

Applications are open for YC Winter 2021

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