what i got at the moment:
1. voronoi working superbly in autocad environment.
the home work is to fix boundary problem, which in same way should also solve "the void" problem.
2. basic brownian swarming agent
this is still so basic, i haven't got to code the agglomeration process. the interpretation is that socially functioning space in the widest space syntax sense is voids' occupation by at least two agents. each agent at the moment has some definition of visual field, but i haven't found a "traffic" in any of paul's swarming code. so this should be another agenda to do.
the home work are to get agglomeration
to manifest idea of traffic in swarming system

thus, yesterday paul reminded me to "get back to what we're doing"
this means, he wants
1. we got voronoi, so what?
my reply to this should be: we know voronoi have been widely used for many kinds of purposes, and it has many useful definition attach to it. voronoi used within the project because it generates divisioning of space, - so useful since each division (or we call it cell) is actually a convex space - of free form. it is actually so similar to space syntax convex map. what is missing in voronoi is how to show connection between these conves spaces.
what to do then is to rule the agent so its now permeable from all sides of voronoi cell.