Now it's 2003. tteoac has turned into an epic.
I started out by setting it to music. Unfortunately (?), the piece i chose is a 16 minute 43 second duet i recorded in 1990.
Also, there are a lot of physics. First, there was the Volcano spew (lava) bouncing off of each other.
Parsing this in the interpreted text parser of POV-Ray proved way too tedious. Well over 90 minutes
to parse each frame. Thus was born a Java servlet, where the parsing was done once and stored in a database.
Now each frame asks for just the parts it wants.
Other people partially solve this problem by writing intermediate values out to disk (output of the POV parser). This solution doesn't work for me because in my 8 machine renderfarm frames must be rendered independently and in an arbitrary order.