tbmbob

6 ms·

1. 1.▲
   by tbmbob 1y ago

   > Now if I actually brought these questions to an interview the interviewee could ruin my day by asking "what's the runtime complexity?" This completely undermines the author's main point. Constraint solvers don&#x27
2. 2.▲
   by tbmbob 13y ago

   This is not quite true: the mathematical problems upon which they base their security, though related, have some important differences. Most significantly, the problem upon which this new implementation is based (Ring-Learning with errors)
3. 3.▲
   by tbmbob 13y ago

   To build off of the comments below, this attack is foiled by the fact that one message can be encrypted to one of a large number of ciphertexts. As mentioned below, this is true for ElGamal. It is also true for all other styles of FHE schem
4. 4.▲
   by tbmbob 14y ago

   "If you need extra data to prove this "yes" in polynomial time, then the answer isn't a boolean anymore, but a boolean plus the extra data, so how can you still call that a yes/no problem?" You are combining two distinct concepts: the answe