How is the conference managed? In section 4 we propose a
general floor control system, where the users may specify the
style of floor control that they wish for the conference in as
abstract a way as possible. This entails choosing whether the
conference is ``chaired'' in the most general sense. If the
conference is not chaired, is there a queuing system to gain the floor,
and if so, what is the queuing discipline? Are users
time-sliced in a manner similar to operating system process
scheduling?
-
Can a user pass on their place in a queue to another, and if so, can
they regain part of their time?
-
How may sub-groups be split off from the main thread of a
conference; how may they rejoin; how conferences might be merged;
mechanisms for mumbled and muttered asides?
-
What are the possibilities for displaying these floor control
policies on the screen? We look at a novel idea (due
to John Taylor of UCL) for graphically showing the precedence of speakers
and how recently they may have spoken.