He’s not wrong users >>> validators. I’m not going to say I know all the technicals about what they’re talking about. But at the user level if a transaction fails, they first assume it’s the app, then they assume the chain sucks. And it happens fast.
For example a stripe payment, the use doesn’t care or know the infra but if something fails, they blame their card, check their balance see it’s fine, then they try again, fail, they try again new card, and it fails, they blame the website and probably don’t bother anymore.
This is why Shopify checkout has been one hosted page forever. Because checkout UX is Shopify that the user sees and they know it works based on brand.
When a eth transaction fails the users blame the chain. Because it’s nearly always the chains issue.
Solana must work practically always so the user thinks it’s the app and not the chain.
Core UX.
It’s the opposite.
A user who sends a tx to a leader that gets included but then the block gets reverted sees bad reliability. The cost of dropping a block has to be astronomically greater than the fees from the block because it is a 100% failure to the user in that moment, but a small reduction in earnings for the validator.
That user then goes to use a different network. If you can make a deterministic slashing that yanks that validators stake as soon as they drop below 99.995% slot reliability, then they can do wtf they want.
Users >>> validators on the priority list, i dont gaf how painful or constrained or miserable the validators are if their efforts result in improved user reliability.