Hacker News Reader

Comment by eternalny1 | Hacker News Reader

So, the NSA, who intentionally compromised NIST with bad crypto proofs, is now warning NIST not to make better crypto, in the name of "security".

View on HN

The headline is clickbait. The only reference to not rushing into quantum-proof encryption happens near the end of the article, and the warning is that folks adopt them gradually so that any vulnerabilities are caught.

This is standard practice blown out of proportion.

I don't get that sentiment from the article at all. The article is talking about the difficulty of targeting a problem space 'too early' vs 'too late'. Too early and you solve the wrong problem, too late and, well, the obvious.

Breaking the old RSA requires more quantum computing horsepower than breaking newer ECC schemes. So although ECC is far more resistant to current attacks, it is less resistant to quantum attacks. The NSA is just saying that you're better off staying with strong RSA security than you are investing tons into migrating to Suite B since you'll just have to migrate to whatever NIST has coming that is quantum-proof.

Where did you read that? The only thing she says is to wait the competition to end (by 2022):

> “It's very important that people wait for NIST to do its due diligence,” Frincke said.

This is the standard process.

The NSA, who have a history of feeding NIST intentionally compromised crypto algorithms, suggest we need to wait to hear what NIST recommends?

How very fucking convenient.

It's a totally different ball game though: this isn't about NIST recommending a shady algorithm with mysterious parameters, this is about a well-known standardization process that accept submissions from cryptographers all around the world where anyone can review the proposals and make comments.

I don't care what they end up selecting as the winner (and to be honest, I'm so ridiculously paranoid that I don't trust Keccak, for instance), I just think that having a competition where everyone is spending all their energy into looking for flaws on the others candidates is a great thing.

Next year China will announce a similar standardization process. Do I trust China? of course not, but I really welcome this initiative anyways.

Quantum computers that can break traditional crypto do not exist. Whether post-quantum crypto exists today or not doesn't matter.

> Quantum computers ... do not exist

How can you be so sure? What if a nation state invents one and treats it as a Manhattan Project style secret?

> Whether post-quantum crypto exists today or not doesn't matter

Unless the first "Quantum computers that can break traditional crypto" are invented and announced tonight... then it matters about the current state of post-quantum crypto.

That's exactly what they're saying.

Reply on news.ycombinator.com