On: The Matasano crypto challenges (2014)

[eru]

We have some amazing theory around cryptography, but we have no theory that's good enough for your '100 year crypto' goal.

[baobabKoodaa → eru]

Well, depends what you want to do. If you can get away with using a one-time pad, that's 100 year crypto for you.

[tptacek → baobabKoodaa]

If you're storing a message for 100 years, and you're considering a one-time pad, just put the plaintext message wherever it is you plan to hide the equally long key for 100 years. "You can use a one-time pad" is a little like saying "you can just not encrypt".

[eru → tptacek]

You are a bit too harsh on the one-time pad.

What a one-time pad gives you is the ability to time-shift: You can meet in person now, exchange your one-time pad, and then securely communicate later. (And destroy the pad after you communicated.)

[baobabKoodaa → tptacek]

I don't really get the hostility in your response. Yes, one-time pads are not a practical solution for most use cases, but I never claimed they would be. I was just making a point about their "theoretically unbreakable" property, which I thought is interesting in this context.

[tptacek → baobabKoodaa]

There's no hostility here, just realism about what one-time pads represent in cryptography.