The HVM Slash Hack Description of Async Operations Is Pretty Solid

The main challenges we had were network petitions and general cluster interruptions. We could recover mostly by kicking everyone from web sockets and forcing them to reconnect and catch up. Instead of using node though, they did move to go lane. But again, I think that was motivated by a desire to have kind of a more baked, more baked in story as to have a handle concurrency. Totally.