This encryption is theoretically impossible to crack (IRL, it's just really really hard).

One-time pads are impossible to break if you have a true source of randomness. (in practice: a sufficiently non-deterministic source)

Basically it does this: (Alice & Bob are the participants, Alice transmits to Bob)

[this is what I remember from reading Bruce Schneier's book 'Applied Cryptography']

Alice and Bob are both in possession of the random string R, of length n. (or both have a cryptographically secure RNG with identical seeds)
Alice wants to transmit message M (with length n) to Bob. She XORs every bit of M with the corresponding bit of R, and transmits the resulting ciphertext C to Bob. Bob, having received C, XORs every bit of C with the corresponding bit of R, thus recreating M. Then, both strings are discarded. (it's insecure to continue using them)

Eve, the eavesdropper, stands nowhere nearer to the solution, since R is completely (at least, in theory) random, and no part of R recurs systematically throughout the ciphertext. Impossible to break this scheme, but hard to implement, since the strings R must be distributed in advance to guarantee security.


EDIT: oops, I missed Galandros' post earlier... Sorry.

I can't read French SMS-like style... This is why I don't frequent TI-Bank so often. Somehow, my Wernicke-center is more tolerant to bad spelling in English than in French

August, and beginning of september.

Yup, indeed, tonight I start the coding run.

This afternoon, I'll have to lend a hand at my little brother's birthday party... Don't really like that, but someone has to volunteer to run around the house with a bunch of 10-year olds.
There's no fencing training tonight (last friday before Easter), so nothing impedes the Sacred activity of Coding.

My 'awesomeness by calc84maniac'-sensor is going nuts again.

Kudos! Let's get rid of this crappy emulator by TI!

The problem with strings is that they are tokenized, and not every token is the same size, so the calc has to go through ALL OF THEM to access a random point in the string. This is not much of an issue in map displaying, but with collision detection, the delays add up real quick.

Mmmh... I actually think the pronounciation-spelling relation is far more complicated in English than in French. (in French, when I hear a word, I can almost always immediately spell it, and vice versa, while in English, this is way harder for me.)

There is _WriteFlashUnsafe...


There is _WriteFlash and WriteFlashUnsafe, and, funnily enough, _WriteAByte and _WriteAByteSafe. Both use fallthrough to prevent writing to the cert. page. (and maybe also the 7x pages in OS space? Iambian?)