> There are two key equivocation rules a validator must follow, taken from the Gasper paper: > > Double vote - no validator makes two distinct attestations for the same target block > > Surround vote - no validator makes an attestation that surrounds or is surrounded by a previous attestation

Which are both punished by having the saked ETHs being "slashed" (32 ETHs staked, you end up losing 16 of the 32).

I never understood that: it opens a door for a huge attack in that it's now impossible to safely "stake". Ethereum devs says you can safely stake in that staked ETHs aren't at risk of being stolen (the key to stake and to move the ETHs aren't the same, so it's impossible for an attacker to steal the 32 ETHs) but...

An attacker accessing a staking node can, on purpose, double vote or surround vote to get the ETH slashed. The difference between losing 32 ETHs or 16 ETHs is not that big (sure, in the case of slashing at least the attacker doesn't get them, but for the person losing the ETHs, it doesn't make a huge difference).

So it's impossible to stake without risking your ETHs getting slashed, even if the staker is honest.

Here's my prediction: we'll see exactly that attack on slashing nodes, people are going to lose ETHs and we'll see excuses to the tune of "you didn't securize your node well enough".

I do think it should be possible to stake without that risk (just as it is possible to hold the private keys to cryptocurrencies on hardware wallets because, well, computers are all too often insecure).

Here's another prediction: some are going to come up with HSM to stake ETHs (maybe they already exist?). The risk of getting people's ETHs slashed is just too big.