How Cesar Performs Better Than Other Network Protocols?

Being able to provide services like this as a service is a huge benefit to any application developer, if they don't know what's going on under the hood. One of the components in the paper was some actual empirical results that I don't know we see in every network protocol paper. So I was glad to learn a little bit about how the theory works out in practice. Could you share some details on those experiments? The protocol we developed is very sensitive to contention. That is because different replicas can issue conflicting commands.