Reputation
33,837
Next tag badge:
95/100 score
20/20 answers
Badges
2 50 137
Impact
~563k people reached

Jan
28
comment Compressing AES
@Shockwave: you wrote " I've tried that already, and the result is the same". I find it unlikely that lzo (with default parameters) fails to compress text files, even more the concatenation of text files. I guess that you did not correctly "Compress before encrypting, not after".
Jan
28
comment Compressing AES
@Shockwave: if your "I've tried that already" refers to "Compress before encrypting, not after", then you need to better understand the nature of the content of your "multiple files" and find or devise a compression algorithm suitable for that. If your "multiple files" are already compressed (including many common sound, image, video formats), that will be very hard.
Jan
27
comment Finding a secret cipher given the key and known plaintext?
The function that I propose is a partially unknown function, since you do not know $c$. Hence if one could solve your problem in general, (s)he could find the constant $c$ in my function.
Jan
27
comment Are there reasonably secure online voting implementations e.g. for student council elections?
I like the first two paragraphs, except for not mentioning the possibility of vote selling, or vote under duress. $\;$ But how is any of the proposed "ways to solve your problem" keeping individual votes convincingly secret including to the organizers of the vote, and avoiding gross fraud count by the organizers (or allegations of that)? What if the college email imap/pop service is hacked, unread emails with the unique ID collected shortly before vote closing, and rogue votes casts using that? What about mere allegations of that? And so on.
Jan
27
comment Are there reasonably secure online voting implementations e.g. for student council elections?
@cygnusv: short answer, no. The website accmv.org is entirely in French, dated, and mostly empty; even the address is wrong. My apologies but our activity has been legal (fighting and loosing a battle in the constitutional court) and direct lobying of the authorities in charge (with some sucess, perhaps thanks to the former).
Jan
26
comment Finding a secret cipher given the key and known plaintext?
Hint: assume $\operatorname{Enc}_k(x)$ is defined as $\operatorname{AES-128}_{k\oplus c}(x)$ for some unknown 128-bit constant $c$.
Jan
26
comment Is there any more information on this RSA backdoor?
In the context of the question, $P=A(P^\prime+i)+P^\prime$ for $A$ an unknown 384-bit prime, $P^\prime$ an unknown 128-bit prime, and $i$ small (likely $0\le i<500$). I doubt that makes $P$ weak in the sense of the paper.
Jan
26
comment Is there any more information on this RSA backdoor?
Can you expand on how "we can easily factor this number with lattice based attack (even if we dont know $A$)" ? That is exactly what the question is about, and I find that non-trivial.
Jan
24
comment Is there any more information on this RSA backdoor?
For one not knowing $A$, how would the key space be reduced to 256 bits? I count closer to 384+128+128 (minus comparatively little to account for the fact $A$, $P$, and $P'$ are prime).
Jan
24
comment Role of Fermat's little theorem in the proof of correctness of ElGamal signature
Hint: apply the definition of $H(m)\equiv xr+sk\pmod{p-1}$ so that you remove the modulo. You should then see how establishing $g^{H(m)}\equiv g^{xr}g^{ks}\pmod p$ involves Fermat's little theorem.
Jan
19
comment Status of Algebraic Eraser key exchange?
@mikeazo: reading it was indeed a memorable experience ; see updated answer.
Jan
19
comment Example of Projective Coordinates
+1 for the AI translation of the comment: "If z = null or z = 0, then the value Z becomes 1", which is better than the code is.
Jan
19
comment Cycling hashing in PBKDF's and their limitations in strength?
Why this complex (and evolving) generation of Output (currently Output := E(Hash(PassHash),Salt) | PassHash ) rather than just Output := PassHash ?
Jan
19
comment Convert projective to affine coordinates in ECC?
$(x_3,y_3,z_3)$ is not three points, it's a triplet for the coordinates of one point in protective coordinates. $\;$ Then the answer to the question seems to be clearly given in the paragraph with the two bullets. Are you having a problem understanding what is meant by $X/Z$ in the question? That's $X$ times $1/Z$; where $1/Z$ is the inverse of $Z$ in the base group (e.g. $Z^{-1}\bmod p$ as obtained using the extended Euclidian algorithm when working with base group $\mathbb Z_p$), of the same nature as $X$, $Y$ or $Z$ are; and times is multiplication in the base group.
Jan
17
comment Proof that this is not a secure pseudorandom function
What you have tried does not work because $y_1=y_2$ does not hold in general when $x_2=y_1+p$ (further, adding $p$ is identity in $Z_p$ ). $\;$ You want to build a distinguisher for $F_k$. Arguably, $F_k(1)=1$ is enough for that, but you can build a more general distinguisher from $F_k(x_1\cdot x_2\bmod p)=F_k(x_1)\cdot F_k(x_2)\bmod p$.
Jan
16
comment Proof that this is not a secure pseudorandom function
Hint: multiplicative property
Jan
14
comment Public key exponent coprime with totient proof
@fkraiem: you are right, the question as it stands now (and at the time I made my comment) asks a proof of:$$\gcd(e,\varphi(N))=1\implies\forall(x,y)\in\mathbb N^2,\;\big(x^e\equiv y^e\pmod N\implies\;x\equiv y\pmod N\big)$$ On the other hand, that was not originally clear, and I then got positive confirmation from the author that the question asks a proofs of:$$\gcd(e,\varphi(N))\ne1\implies\exists(x,y)\in\mathbb N^2,\;\;x\not\equiv y\pmod N,\;\;x^e\equiv y^e\pmod N$$Both are true, the first is well known, the second is less common.
Jan
12
comment Simplied DES why 10-bit key?
Some data: William Stalling's Simplified DES has 8-bit plaintext and 10-bit key. The overview of appendix G of his Supplement to Cryptography and Network Security (Fifth Edition) states: "The algorithm could have been designed to work with a 16-bit key, consisting of two 8-bit subkeys, one used for each occurrence of $f_K$. Alternatively, a single 8-bit key could have been used, with the same key used twice in the algorithm. A compromise is to use a 10-bit key from which two 8-bit subkeys are generated...".
Jan
12
comment Can I shorten the large ECDSA public key output file from OpenSSL?
Yes. This gives constant size because the encoding uses BIT STRING rather than INTEGER.
Jan
11
comment Is finding collisions in a part-hash not often enough a bad problem?
Applying the formula given, I get ~82137