While I may agree with certain other things you’ve wrote in that comment, Polkadot indeed has made many sub-optimal technical choices for being the world’s next “ubiquitous computer”.
- We should have used binary trie instead of hex trie. We’ve known for a long time (actually, since the day when I was still doing in Ethereum hard forks in Parity Ethereum) that hex trie has certain performance problems. Our state size is rather small at this moment so nothing’s apparent. But when Polkadot state grows big, we’ll be rediscovering them as it was on Ethereum.
- EVM is certainly not the optimal solution for smart contracts, but WebAssembly is also not the silver bullet. It does have its problems. The optimal VM for a smart contract is probably closer to RISC-V or eBPF (but we may soon discover issues there as well).
I would argue that for many decisions we made in Polkadot, it’s “marginally” better than Ethereum, but not “magnificently” better. If you compare it with the actual “optimal” solutions, it’s still sub-optimal. The question is then, for anyone thinking about using Polkadot vs Ethereum, whether the technical trade-off is worth it compared with the lose of ecosystem.
In Polkadot, I’ve seen many times that the debate, especially on things that Polkadot thought to be “superior”, ended up being an ideology battle, rather than what it should be – a technical battle. This happens in the debate of on-chain governance system, of WebAssembly vs EVM, and of parachains vs Ethereum roll-ups. I don’t think this helps us at all. It just makes us blind of the development on the other side.