How about this:
Most interface design choices are ultimately driven by concerns about
rendering load and computational load. Design seems to be very often
officially about "making it look good" e.g. intuitive lay-out, whereas
effectively these design choices are more heavily influenced by the concern
of "making it feel good" e.g. faster.
However, when designing very futuristic CVE software, one can also assume
that, on those machines of the future, rendering power and network traffic
speed will be less and less of a limiting factor.
Does it then still make sense to search for those design solutions with the
best reduction of network traffic and computational load?
For instance, transporting CVE participants who have decided to be a member
of a group in the CVE, as a group from world to world, means not only that
the group members do not lose each other, it also means that multi-cast can
be used for the state-transfer.
Is CVE design actually always about finding the optimal balance between the
lowest possible computational load and the highest possible usability
score, and - no matter how fast your future machine?
//Jola.
http://www.crg.cs.nott.ac.uk/~jgt/