The randomly-generated seeds for the 20 runs reported in my ALife VIII paper are:

Run  1: 12522 (The run reported in detail)
Run  2: 29726
Run  3: 22746
Run  4: 22776
Run  5: 13571
Run  6:  7645
Run  7: 14064
Run  8: 21744
Run  9: 14106
Run 10: 31360
Run 11: 13428
Run 12: 22767
Run 13: 22784
Run 14: 31234
Run 15: 12399
Run 16: 10980 (Run reaches pseudo-extinction - see 'Results and Discussion' section of the paper)
Run 17: 12635 (Run reaches pseudo-extinction - see 'Results and Discussion' section of the paper)
Run 18: 31031 (Run reaches pseudo-extinction - see 'Results and Discussion' section of the paper)
Run 19:  1329 (Run stagnates with no more reproduction or killing - see 'Results and Discussion' section of the paper)
Run 20: 12844 (Run stagnates with no more reproduction or killing - see 'Results and Discussion' section of the paper)

You can replicate a run by using the following command line (replacing XXX with the relevant seed):
./geb_v07c*.resets.Linux.i386.bin -display -full -autofile -rulestat -seed XXX

Geb is available for download from http://www.channon.net/alastair/

Take care not to introduce excess precision by compiling without the -ffloat-store flag (see makefile).

Please see makefile_rs for instructions on processing the raw data to produce activity wave diagrams and statistics.

It is also possible to replicate the 20 runs reported in my ECAL 2001 paper, but this is somewhat more involved as
these runs predate my implementation of a random-number-generator snapshot mechanism, so each run requires several
seeds as well as the use of retired code.  Please contact me for details: I can burn you a CD containing full
details of all runs (both ECAL 2001 and ALife VIII), including all seeds and retired code.  This CD would also save
you from rerunning these experiments, which would take approximately three years of 1GHz single-cpu time.  

Best wishes,
Alastair.