Reputation
Top tag
Next privilege 250 Rep.
View close votes
Badges
6
Newest
 Teacher
Impact
~268 people reached

Apr
4
awarded  Teacher
Apr
4
answered What is this “finite field cryptography”?
Apr
4
comment What is this “finite field cryptography”?
@kasperd: Correct. In an elliptic curve, you can only add two points together or take the negation of a point. You can also multiply a point by a scalar, which is merely shorthand for adding a point to itself repeatedly. You can't multiply two points, thus it's not a ring.
Apr
3
comment What is this “finite field cryptography”?
Note that the term "finite field" excludes elliptic curves because these curves are only groups, not fields.
Mar
27
comment Why is CRC said to be linear?
I have a longer explanation posted previously: stackoverflow.com/a/7005801/839689
Feb
18
revised Why are recomended RSA key lengths so high?
Improved spelling and syntax
Feb
18
suggested approved edit on Why are recomended RSA key lengths so high?
Feb
18
suggested rejected edit on Why are recomended RSA key lengths so high?
Sep
11
awarded  Commentator
Sep
11
comment How to propagate error in variable-sized message?
What about this? ciphertext = (AONT(message XOR keystream) XOR more keystream).
Sep
11
comment How to propagate error in variable-sized message?
Hmm, you're right about XORing the AONT of another plaintext, because short plaintexts might be guessable. So I guess I need to make the AONT a keyed secret function, huh? But I feel like this solution is getting too contrived, and there should be a more elegant way to solve the problem...
Sep
11
comment How to propagate error in variable-sized message?
@RickyDemer: Not a problem - the keystream is different each time because there is a different secret key per message. I didn't mention it, but this crypto problem is embedded in a larger system which has supplies more data and verification capabilities.
Sep
11
comment How to propagate error in variable-sized message?
It renders the same for me. I guess it's the extra spaces and the lack of list bullet points that made me think something was weird.
Sep
11
comment How to propagate error in variable-sized message?
Well, what I mean by "security by obscurity" is a warning that any custom scheme you or I might propose will be subject to public scrutiny. I'm not criticizing any existing published schemes out there. It looks like PCBC is vulnerable to block swapping. Now I'm leaning towards an algorithm of AONT preprocessing + stream cipher.
Sep
11
comment How to propagate error in variable-sized message?
Sorry, I have trouble understanding what you're trying to say... The formatting of your text isn't helping.
Sep
11
revised How to propagate error in variable-sized message?
added 195 characters in body
Sep
11
revised How to propagate error in variable-sized message?
added 722 characters in body
Sep
11
comment How to propagate error in variable-sized message?
I have a few comments as well - the source code for the to-be-designed algorithm will become public, so no "security by obscurity" is permitted. This also means that EME and patents are a no-go either. And finally, IIRC PCBC is still malleable because if you flip the same bit position on two ciphertext blocks, the decryption error will be healed thereafter, right? (Thus prefering something like IGE instead.)
Sep
11
comment How to propagate error in variable-sized message?
Thanks for your informative answer. I have one main concern - it doesn't address the fact that many messages will be less than a block long.
Sep
11
awarded  Editor