Aaron Parsons, Aug 2005 The point of this project is to allow little scripts (critters) to replicate themselves and mutate, hopefully with interesting results. To keep this system stable, a second (static) script handles file and process management, and generally imposes rules upon the critters. Requirements: -------------------------------------------------------------- Pyvolve is made to run on a bash shell. Currently, I've only tested the scripts on Cygwin on Windows 2000 and on Fedora Core Linux, but in principle any standard Unix/Linux system should work. You can download Cygwin from www.cygwin.org -- I highly recommend it. Evolution uses programs that come with most standard UNIX distributions. These are the programs you need to have on your computer: make, bash, python, cut, date, grep, zip/unzip, gzip/gunzip, echo, expr, zcat, and gmake You also need dparser, with the python swig interfaces. Just download it from http://staff.washington.edu/sabbey/py_dparser/d-1.13-src.tar.gz unpack it (tar zxf d-1.13-src.tar.gz), and follow the instructions in the readme. Make sure to follow the steps for building the python part. Additionally, the scripts in ./scripts and the file 2.py need to point to the correct location of python and bash. You can do this automatically by typing: $ make conf This should be enough to get you going. Basic Usage: -------------------------------------------------------------- $ make conf configures life for your unix-like system $ make life starts the life program $ make halt brings life down gracefully. Next time you run life, it will continue where you left off. Can take a sec, so be patient. $ make clean gets rid of extra, remakeable files floating around. $ make real_clean throws out all current critters, forcing you to download a new set of critters. $ make update syncs your version of pyvolve with the latest release. Does not mess up any critters you are evolving. $ make post submits latest release of code to repository. For my use only. :) $ make demograph groups existing critters by ancestry and gives percentage of population each subgroup represents. $ make gen_num prints how many generations have elapsed. $ make viewlog shows the last several entries in the system log, describing what's going on. $ make visuals prints an abstracted version of the script currently running, highlighting in color ways in which it is different than the original script. Red letters represent items in the original which have been deleted, blue letters represent items which have been added, and green letters represent items which have changed in place. The top pictoral is of the running script, and the bottom pictoral is of the children it produces, compared against the parent (i.e., how the offspring are different than the parent. If life is running (in the background, or in another session) the display will refresh as new scripts run. Not all mutations produce a difference visible in this visualization system. $ nohup make life & useful command for running life in the background, and keeping it active even after you log out of a machine. Known Bugs: -------------------------------------------------------------- There are plenty, but I don't have them documented yet. Good Luck!