Golly: how an open source project took over the study of cellular automata.

Update: Presentation details are now available.

Golly is an open source, cross-platform application for exploring Conway's Game of Life and other cellular automata (CA). The primary authors are Melbourne-based Andrew Trevorrow who has been largely responsible for Golly's powerful user interface, building on his earlier work as developer of the leading Mac CA app LifeLab and California-based Tomas Rokicki who has largely been responsible for the underlying CA algorithms, building on his hlife adaptation of legendary MIT Life hacker Bill Gosper's "super fast" hashlife algorithm.

Golly's already extensive CA rule collection is extremely extensible via the RuleTable and RuleTree algorithms which let you plug in new rules while the operation, input and output of Golly can be customised through Python or Perl scripts which are run within the Golly environment with APIs giving scriptable access to Golly's many functions.

The author has developed a set of Perl scripts, some internal and some external to Golly, which enable him to capture many of these animations with pan and zoom of particularly interesting patterns evolving; to provide compressed renderings of a vast diversity of large patterns generated from small viable seeds after 100,000, 150,000, (rotated 45°) 200,000 and 250,000 iterations; and to otherwise manage data collection and analysis across three computers.

On 18th May 2010, an out-of-left-field announcement on the CA message board of reference proclaimed the production of a holy grail of heavy engineering of Life patterns; a "Universal Constructor Based Spaceship", developed and demonstrated using Golly.