Reputation
3,506
Top tag
Next privilege 5,000 Rep.
Approve tag wiki edits
Badges
2 15 40
Newest
 Nice Answer
Impact
~85k people reached

Mar
22
comment Seed / reseed DRBG too often?
@runeks The most common use of a DRBG is to generate unpredictable data (i.e. data that isn't correlated with anything else). You can do that with a DRBG with an unpredictable seed. In my answer I give a few reasons to use a DRBG seeded with /dev/urandom to generate unpredictable data: performance, portability, demonstrable security. It's rare to actually need the determinism aspect; an example use case is when you want to keep logs of everything and you want to be able to repeat a run of the program, e.g. to debug it.
Mar
21
comment Signing MD5 hashes with RSA?
(cont.) Not to say that MD5 shouldn't be replaced, but it's not systematically broken. The points about unpadded RSA are a bigger concern. But the recommended padding mode is PSS (PKCS#1 v1.5 does have weaknesses, not necessarily critical but it shouldn't be used in new protocols).
Mar
21
comment Signing MD5 hashes with RSA?
“it may be possible to create a forged message with the same hash as the message you signed”: no, this is not the case. It is possible to create two messages with the a given prefix and a given suffix. There are situations where this is a dealbreaker, because the attacker can cause the signing party to include a payload in the middle of the message that the attacker controls, with P1 being an inoccuous payload that the signer willingly signs and P2 being a nefarious payload such that the resulting MD5 are the same. But if the message is under full control of the signer then MD5 isn't broken.
Mar
20
comment What's the best way to pad a message?
Are you signing or encrypting? Why aren't you using the standard methods (PSS or OAEP)?
Mar
9
comment Fast algorithm for reduction modulo a prime
@ArtjomB. This is math that's directly relevant to the design of cryptographic algorithms, and it's about an algorithm which makes it more computer science than math, so I really can't see this question being off-topic here.
Mar
8
comment Why haven't we proven many things computationally secure yet?
NP-completeness wouldn't help since it's unproven that this is distinct from P. By the way, there are cryptographic algorithms that are based on NP-complete problems, but they're too slow in practice and there's no reason to think they'd be more robust than other constructions.
Mar
4
comment Confidentiality then Integrity with different keys
This is a migrated repost of crypto.stackexchange.com/questions/33380/… . Please do not post the same question on multiple sites.
Feb
28
comment Does TLS have to provide both confidentiality and authenticity?
“You've really nothing to lose when using encryption” That's not true: you lose caching (except at the endpoints). Caching is only useful for large volumes of data though, not for authentication protocols.
Feb
18
comment What's the issue with Apple's backdoor?
@ChrisPeikert Oh, I wasn't familiar with the details of that case, thank you. In this case I wonder why FBI doesn't do offline enumeration. FBI might have trouble with a secure enclave (they could probably break it but it might cost a lot), but with ordinary flash memory, they can just make copies or use independent software.
Jan
19
comment Understanding the cryptography used in a file
I'm sorry but this question is neither on-topic nor suitable for Stack Exchange. What you have here is a reverse engineering problem, not a cryptography problem. Reverse Engineering would be a more appropriate forum, but don't just re-ask there — this question is too broad, essentially you're asking us to tell you all we know about reverse engineering. You need to do some of the job of breaking this up into manageable parts. Do note that attacking such problems by looking at one data dump is rarely useful, usually you need to confront multiple datasets or observe the code in a debugger.
Jan
13
comment Should I use RSA encryption since RSA is said to be broken by NSA?
If you're typing the letters RSA into your code, you're doing it wrong. Cryptography is not a magic bullet and is hard to get right. Use protocols and libraries that experts have written and let them pick the algorithm.
Jan
1
comment TPM's monotonic counters
@MaartenBodewes On the contrary, this is about the implementation of a cryptography-related protocol (for storage integrity). It would be off-topic on Information Security since this isn't about the security of the protocol but about how it works.
Dec
27
comment Prevent double-spending with decentralized digital currencies without all transactions being public?
Mentioning Zerocash is interesting, but unfortunately your answer doesn't explain the crucial part: why is it that before spending a coin, I can prove that I own it, but after spending it that proof no longer works?
Dec
15
comment Cyclic group collision resistant keyed cryptographic hash function
Please post text rather than an image. An image is hard to read at the wrong resolution, impossible to read if you're blind, impossible to search, impossible to copy from… You can use mathematical notations with MathJax's LaTeX-like syntax.
Dec
14
comment Hash functions vs Stream Ciphers in terms of Speed
I don't understand your revised question better than the original. If you want variable-sized output, what does this have to do with a hash? You seem to be asking which is faster, an apple or an orange.
Dec
12
comment Help with RSA-2048 crypto ransom virus
I'm voting to close this question as off-topic because it's about recovering from malware, not about cryptography.
Dec
1
comment Checking the integrity of RSA public key
How about the attacker sending chosen ciphertexts or plaintexts with varying values of $e$ (assuming a scenario where the attacker can force the cryptographic module to reload $e$ from the breached storage at will)? Could that reveal information that fixed-$e$ attacks wouldn't?
Nov
22
comment Probability of guessing random 128-bit AES key
I'm voting to close this question as off-topic because it is about counting, with only a superficial application to cryptography.
Nov
21
comment How does generating random numbers “remove entropy from your system”?
“once you've used a seed you should really throw it out and collect another one”: that's grossly misleading. Once you've used a seed, you throw it out, but you can keep using the PRNG seeded from that seed for a very long time.
Nov
19
comment Keyless integrity checking with SHA-256
Why don't you store the size as well as the hash? It's usually useful for other reasons (allocating sizes, detecting obvious transmission errors, etc.).