|15 minutes of fame came by way of the September 1986 edition of Scientific American which feature Pattern Breeder as one of three kinds of "Wallpaper for the Mind" in that month's Computer Recreations section.|
Once upon a time, pattern reproduction was considered a holy grail of cellular automata (CA) until Ed Fredkin demonstrated that reproduction was trivially easy in a CA which determined the next state of a cell by exclusive oring the states of it and its eight nearest neighbours.
Reading about Fredkin's CA in an earlier Computer Recreations column and receiving a computer with GW-Basic to review, both on the same weekend in 1983, inspired a first shot at graphical programming. With the advent of the Mac and especially its support for XORing directly into screen memory, it was easy to generalise Fredkin's CA to arbitrary neighbourhoods and add a few interactive bells and whistles.
For many years this all looked a bit trivial and a bit cute, something to have happily moved on from. But the wheel has turned and the lessons of Pattern Breeder are starting to look a bit more relevant for a couple of reasons:
While a full listing of Pattern Breeder is available, it relies on an environment that is very different to today's. It has been hard to try to think about reviving it in the absence of an obvious way of XORing at least buffer bitmaps, but if a way to do that was available it might be a useful contribution to the public domain.
- The diversity of "Life in a Tube" is very much a product of the known fact that a single line (circumference) of active cells in torroidal Life spreads at c as a single live cell does under Wolfram's (1D) Rule 22, and, as often happens, when most of the live cells are at intervals of 2n the behaviour of every 2nth cell at every 2nth generation is governed by Rule 18 which is a nearest neighbour XOR rule. Life in a Tube reveals that Rules 22 and 18 show some more interesting behaviour than might be expected from their classical nested forms.
- Given the emerging understanding that CAs are an accessible model but that space time energy matter are ultimately a realisation of a Planck scale network rather than any kind of CA, there is likely room enough in the 23 orders of magnitude between the Planck and atomic scales for the network topology to underpin one or three things that are not being picked up by mainstream physics. Beyond hypothesised dark matter and dark energy, Pattern Breeder suggests that something like Rupert Sheldrake's Morphic Resonance might be easier for the Planck network than say conservation of linear momentum.