I heard Charles saying that Ouroboros “solved [the problems encountered in] proof-of-stake”. Also, researchers claim that Ouroboros is elegant because the papers are finished, and each protocol (Ouroboros, O. Praos, O. Genesis) is self-contained.
But as far as I’ve found, none of the papers talk about how malicious/evil/byzantine/misbehaving or faulty validators are punished/reprimanded. There’s no talk about validators getting slashed (=they lose their deposit) in certain conditions.
How could such a proof-of-stake protocol even work? What is it that I’m missing here?
I’m not qualified to answer this, but if I had to guess they don’t have to punish malicious behavior in order for the protocol to be secure and I this this is what they proved in the papers.
The whole point of the Ouroboros is that it does not require slashing to guarantee security. For example, Casper directly depends on the slashing mechanism to argue its properties, e.g. if validators disagree - some of them are slowly slashed out of their stake until their stake is irrelevant.
Slashing is not a central part of a PoS protocol in any way or form. It may be optionally added as a feature at any time for some provable malice, but it’s not necessary for the protocol to work.
I would also point out that this is the beauty of the protocol. It reduces the need for a punishing mechanism which eliminates the complexity of designing and maintaining it.
Incentives can be a strange way of viewing Ouroboros, like your tile points out, not punishing malicious validators? I stumbled over this as well and only became clear once I realized that not being able to participate is indeed incentivizing monetary and moral actions.
