In this lecture, we start to develop notions of replicated data based on multicast.  Here we look at some simple ordering properties on updates, using Lamport's logical clocks and an extension called vector clocks as the mechanism for measuring the advance of time in our distributed system.  In fact the connection won't be completely clear until next week when we use LT and VT representations in ordered multicast protocols.  In this first exposure to them we develop the ideas, point out that a consistent cut algorithm can be understood using them, and then start to look at how order is visible in multicast systems.  Implementing ordered multicast delivery will wait until the upcoming two lectures.