We gave a solution to consensus from a solution to reliable and totally ordered multicast, which
involved selecting the first value to be delivered. Explain from first principles why, in an
asynchronous system, we could not instead derive a solution by using a reliable but not totally
ordered multicast service and the ‘majority’ function. (Note that, if we could, then this would
contradict the impossibility result of Fischer et al.!) Hint: consider slow/failed processes.

What will be an ideal response?

If we used a reliable but not totally ordered multicast, the majority function can only be used meaningfully if
it is applied to the same set of values. But, in an asynchronous system, we cannot know how long to wait for
the set of all values – the source of a missing message might be slow or it might have crashed. Waiting for the
first message delivered by a reliable totally ordered multicast finesses that problem.