As a second example, consider the communication paradigm referred to as queued RPC, as introduced in Rover [Joseph et al. 1997]. Rover is a toolkit to support distributed systems programming in mobile environments where participants in communication may become disconnected for periods of time. The system offers the RPC paradigm and hence calls are directed towards a given server (clearly space-coupled). The calls, though, are routed through an intermediary, a queue at the sending side, and are maintained in the queue until the receiver is available. To what extent is this time-uncoupled? Hint: consider the almost philosophical question of whether a recipient that is temporarily unavailable exists at that point in time.

What will be an ideal response?

This relates precisely to the point made above of how we interpret independent lifetimes. If we interpret this as the receiver (in this case the server) not yet existing then this is not time-uncoupled. If however we interpret it as not being available at a given time, that is the server exists but is disconnected, then this is time-uncoupled. Again we see this is rather a subtle issue and one that is not fully understood in the literature.