the versioned bytecode experiment will result in immense tech debt and an even deeper moat for the perpetrators to stick around and keep introducing bugs into the protocol for a long time. we should fork bytecode exactly once to support upstream ebpf then never again.