1,534 reputation
1416
bio website
location Madrid, Spain
age
visits member for 1 year, 8 months
seen 4 hours ago

Dec
10
awarded  Notable Question
Nov
29
comment What would make it impossible to deny that decryption of a package has taken place?
I don't think this is strictly possible with crypto alone... it's more of a software/protocol problem. If you can assume that the server only delivers the data to be decrypted once the user has been authenticated, and that decrypting the data is trivial (ie, that particular user has the key to decrypt it), then it's probably safe to say that the data has been decrypted by that user, meaning that the server can make a note of it's retrieval/decryption. Vaguely related are time-sensitive crypto questions here and here.
Nov
28
comment What is 'key agility' in relation to symmetric-key encryption?
Poor key agility can be an advantage in some situations. The slow/expensive key initialization (ie, key agility) in Blowfish means that it's more immune to brute-force attacks.
Nov
26
answered scrypt and bcrypt for benefits of both?
Nov
24
answered applying HKDF at both the client and the server
Nov
21
answered Encoding multiple fields with same AES key and vector?
Nov
7
comment Is encrypting credit card numbers one by one with rsautl secure?
@owlstead - yup, key size is only a small part of key management - but one that often gets overlooked with RSA - I'm amazed at the number of people still using 1024-bit keys in the wild. Thought it was worthwhile making a PSA.
Nov
7
comment Is encrypting credit card numbers one by one with rsautl secure?
Just making this comment because it wasn't specified in the question - a 2048-bit key should be used as a minimum (but I'd go even higher if I were encrypting something as sensitive as credit card numbers).
Nov
5
comment Splitting a password for dual roles
Have you considered using a slow KDF like PBKDF2 or Scrypt to compensate for the low entropy of the password, and then simply split the output into two keys?
Nov
2
comment Are there any Javascript CSPRNGs?
@e-sushi - the whole chicken/egg problem with javascript is real - but still, the world could be forgiven for having trust-issues with TLS/the-internet/NSA. This question is strongly related to this one. It's a contentious issue - I err very much on the side of experimentation and creativity - not qualities that usually lend themselves well to crypto.
Nov
2
comment Are there any Javascript CSPRNGs?
@CharlesHoskinson - C'mon - I'm a fan of JS, really, but 'King'? In my opinion, a language that can't handle 64-bit integers natively in crypto is going to struggle. I also think that weakly-typed languages like JS are asking for trouble.
Nov
2
comment Are there any Javascript CSPRNGs?
@e-sushi: "Am I really dealing with a random number generator, or with something provided by an attacker?" - what's your line of reasoning for doubting this assertion? Say for example, a sound javascript implementation of CTR_DRBG as defined in SP800-90A, delivered via TLS, using decent quality entropy (or even something like the service offered here). Given the NSA's recent indiscretions (amongst others), I think JS crypto is something worth considering.
Nov
1
comment HMAC vs Encrypted Hash
@RickyDemer - on second thought - I see your point - I'm removing my initial comments.
Nov
1
comment HMAC vs Encrypted Hash
@user2908183 Good discussion here...
Nov
1
comment Does ssl_rsa_with_rc4_128_md5 have known weaknesses?
@e-sushi - good for you for not being a See-you-next-Tuesday. I think this forum could use more of your inclusive attitude.
Oct
31
comment How to use HMAC-SHA1, HMAC-SHA2 in conjunction with AES
@user907810 - I can't speculate as to what's happening internally in your library. Regarding a 'master key', well assuming you begin with a 128-bit (minimum) cryptographically strong key (the master key), you could use something like HKDF to derive encryption and authentication keys.
Oct
31
answered How to use HMAC-SHA1, HMAC-SHA2 in conjunction with AES
Oct
27
comment The weak link is the password?
@MarkLitovsky - directly from wikipedia: "When the standard was written in 2000, the recommended minimum number of iterations was 1000, but the parameter is intended to be increased over time as CPU speeds increase." I've heard it said that the number of iterations should (roughly) be doubled every year. 100000 or a million iterations these days is common. Of course, nowadays, things like scrypt offer better protection.
Oct
26
comment The weak link is the password?
@MarkLitovsky - PBKDF2 is used in Truecrypt. The user-selected hash algorithms you refer to are used as the underlying hash in PBKDF2, and also in Truecrypt's PRNG. It's worth mentioning, however, that Truecrypt only uses 1000 PBKDF2 iterations, which is far too few.
Oct
25
comment Do any non-US ciphers exist?
Salsa20's author is DJB - an American. Don't hold that against him though - he's a freedom fighter.