A scifi novel I read, "A Deepness in the Sky" described how the pads themselves were a valuable item of trade. I don't think it's farfetched to imagine purchasing OTP data to use with internet browsing, the way we buy yubikeys to use with passwords. It would be a far simpler encryption scheme than those we currently use, and that simplicity would make it easier to catch bugs like heartbleed in the future.
How hard would it be to extend the ssl protocol to allow the use of OTP pads, where available?
That sounds like a beyond ridiculous system. We can do better.
So your mobile phone for example would need to have local storage the same size as all the data you ever expected it to send/receive.
OTP is not a very complicated scheme. All you need is a good source of entropy, a place to store a big fat array of it all, some XOR operations, and a safe way to hand the codebook to your trusted parties (e.g. phone-to-phone transfer options).
Elaboration: https://news.ycombinator.com/item?id=6008695
Once we're past the point of users doing something beyond downloading an app from a corporate app store, all secure encryption would be back on the table.
Of course we can foresee a true ban on encryption some years out, but at that point XOR has the exact same signature as AES. Steganography is the corresponding approach for that attack.
Is there some hypothetical reason we can't just append the SHA256 of the message to the message before encrypting it? It should be impossible for an attacker to alter any message bits undetected with this scheme.
First off citing yourself isn't an elaboration. Second if you are arguing that one-time-pads don't work / won't work in modern times that goes against our entire understanding of certain bits of cryptography.
You'll need to yield some real sources first.
> A one-time pad XORed with a message doesn't provide integrity, and therefore can't reliably secure an arbitrary protocol
The point is that sure, one time pads "work" the way they're described. But their properties don't actually fulfill what we require from modern cryptosystems. Note the sibling comments talking about augmenting them with half baked authentication schemes. Which are all into the realm of computational-complexity cryptography, and no longer "mathematically unbreakable!!1!"
A children's book would be a great demonstration of this, and should be protected speech (let's hope). Teach kids how to use OTPs in an illustrated book. Maybe the story ends with 5 year old Jimmy getting hauled away by the feds for doing some illegal math.