i hand drew this one in mustard on a veggie burger.

This browser does not have a Java Plug-in.
Get the latest Java Plug-in here.

but i lost the picture, my bad. all i have is the synthetic version, i guess.

the algorithm used here for connections is based on a stochastic power law, making it a scale free network. the idea of hand-typing each connection in xml made my fingers cringe, so instead I wrote two python scripts and one superCollider function to help me synthesize this network.

it started with a desire to synthesize networks in terms of their adjacency matrices, to be expanded into more complex and redundant xml.

python script

i'm new at this whole python thing, as you can see. it's a really beautiful language though, visually. much easier on the eyes than superCollider, for this kind of stuff, anyway. so matrixToXML.py will take a file containing an adjacency matrix, line break delimited. so something like this:


this is a lot faster to write than xml, by the way. i wrote a superCollider function to write them for me, based on a simple algorithm.


the algorithm, while it may be somewhat strange, shows off the smalltalk-y beauty that is superCollider. it's way shorter than the following:



and finally, since this leaves you with a somewhat transparent naming scheme, i decided to encode a kind of orthographic heirarchy using a base-ten system.


i also used the same two python scripts to generate the xml for the other two networks. also i didn't eat them, because they were on paper.


a distributed network of surfaces in three-dimensional space.



a decentered network representing the lists of enrolled students in two ITP classes.