@signalapp

And you still require a fucking phone number to make an account.

"Whoop doop we're post quantum encryption but you still have your phone number and give away ALL your metadata and who you call!"

@signalapp
Amazing work you guys are doing!

Though I'd like to see some community oriented features like spaces and/or moderation tools

Still my most favorite messenger
Keep up the good work!

@signalapp @Em0nM4stodon I mean yay, but you say that like quantum computing isn’t a complete waste of time and money

@signalapp wow 😲

@signalapp Fascinating! Thanks for the great work, you make the world a safer place.

@signalapp > ML-KEM 768
https://blog.cr.yp.to/20231125-kyber.html this one, right?

@benroyce

Thank you for this!

(I read the original post, and all I could think was: "I like cake.")

@erikcats @signalapp

@benroyce @erikcats @signalapp Michael Chrichton wrote about this (Prey? maybe). One of the (fictional) points he made was that brute force had been pushed to high levels using quantum computing. But the Gov't continued to push the idea that they couldn't decode high level encryption so that 'bad players' would continue to use encryption that had been compromised.

@vnikolov @jwcph @erikcats @signalapp

the great weakness of all fascists is their arrogance and hubris. when they surround themselves with yes men and push lies over truth, they fall for their own bullshit about "superiority"

Possible, but maybe far from certain.

Breaking the Enigma wasn't really smooth sailing.
It had its ups and downs during the most critical years,
but those can't be summarized in 100 words.
Also the Germans helped by overestimating the strength of the Enigma and thus neglecting some measures that would have been in their interest.

The Venona project is also instructive.
It achieved a lot and yet decrypted a small part of all intercepted messages.

@benroyce @jwcph @erikcats @signalapp

@gbargoud @erikcats @signalapp 🤮

@benroyce @erikcats @signalapp

I really really do not want to see a picture of his dick. I'm certain that's in like half of them.

@gbargoud @erikcats @signalapp

OOOOH

all those trump land encrypted messages will be seen in the future

that's a great bit of news

@benroyce @erikcats @signalapp

As mentioned elsewhere, we don't have quantum computing but one big issue is that some people are banking on it existing in the future and are hoovering up encrypted data that they will later be able to crack if they are correct.

@mkj @benroyce @erikcats @signalapp is there an actual example of a current quantum computer breaking even a simple rot13 encryption?

@mkj @erikcats @signalapp

yeah that's my understanding as well

"you can break this but you need a bank of 1 billion computers operating for 1 billion years, so..."

{quantum computing enters the chat}

"oops"

so you just change the method to something that is not so vulnerable to quantum computing

Bottom line, I guess, is that there are plenty of nuances here which I am happy to let people who know the stuff much better than I do handle; but it's not quite as clear-cut as "the whole world will break" as it is sometimes presented.

There's a number of detail assumptions which may or may not turn out to be true which impact the actual result. But taking a cautionary stance, we do know that the risk is non-trivial and thus taking mitigative steps is good.

@benroyce @erikcats @signalapp

@benroyce I'm not an expert either. My understanding is that for example modular multiplication math (which is used for almost all classical public key cryptographic algorithms, both encryption and signing) is potentially highly impacted by QC; but much math used for symmetric-key encryption and for hashing is significantly less affected. E.g., the effective security of AES-256 is reduced to ~ AES-128, BUT that also assumes QC operations are similar to classical operations.

@erikcats @signalapp

@mkj @erikcats @signalapp

is it Winternitz One-Time Signatures/ Lamport signatures?

i'm not a cryptographer but this stuff fascinates me

https://en.wikipedia.org/wiki/Lamport_signature

@benroyce @jwcph @erikcats @signalapp well, one of the problems is making sure the errors stay within certain limits of acceptability (stay "bounded", as we say). There are also other issues, such as hardware scalability, but these will become less important as technology progresses. Just think how much computers have advanced in the last 40 years!