Re: Rewriting SSA. Is This A Chance For GNU/Linux?

Le 06-04-2025, Richard Kettlewell <invalid@invalid.invalid> a écrit :

Stéphane CARPENTIER <sc@fiat-linux.fr> writes:

Le 06-04-2025, Farley Flud <ff@linux.rocks> a écrit :

On 06 Apr 2025 08:59:33 GMT, Stéphane CARPENTIER wrote:

Le 05-04-2025, Farley Flud <ff@linux.rocks> a écrit :

>

No, but we can move to quantum computing, which may become
a reality before too long.

 
I heard about that before I was born.

>
In the US, the NIST is already researching algorithms for "post-quantum
cryptography:"
>
https://csrc.nist.gov/projects/post-quantum-cryptography

>
Yes, the algorithms are farther away from the computers. Doesn't that
ring a bell?

>
Not quite sure what the argument is here,

I mean the algorithms are very well advanced. They just need a computer
to switch from ready to usable.

but “already researching” is
severely behind the times. Multiple PQC algorithms are well past the
research stage, with finalized standards published in August and a
couple more on the way. Adaptation of higher-level standards (APIs, PKI,
etc) and adoption leading to deployment is well underway.

Yep. The algorithms. For the big computers, it's another story.

Quantum computing is definitely going to happen.

>
Yes, I know. Soon. Very soon. It's almost there. I heard that before I
was born.

>
I’m not sure anyone thinks quantum computing is “almost there” in the
sense of e.g. a quantum computer big enough to break RSA existing this
year.

The one I was responding to does.

However the risk is real enough to be worth actively mitigating,
both because we may not know when the first one is deployed and due to
the “harvest now, decrypt later” strategy.

It depends on what you want to protect. Sometimes the information does
need to be secret only for a short period. You never want anyone to know
something, so yes, it's better to consider the quantum computer is
already there, like that there won't be a surprise later.

But the quantum computers aren't designed only to break RSA. And on
other ways to use them, I'm far from sure waiting for them is always the
best solution.

--
Si vous avez du temps à perdre :
https://scarpet42.gitlab.io