The B0 + S8 rules look okay so far also:
Code: Select all
x = 64, y = 64, rule = B02ci3anyecr4yikrqtjzec5yji6nkec7e8_S2an3njyeqr4aynwkze5krcejna6ikca8
64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$
64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$32ob31o$33ob30o$31o3b
30o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$
64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o$64o!
runs the same in Golly (via isotropic-rule-gen.py again) as
Code: Select all
x = 3, y = 3, rule = B02ci3anyecr4yikrqtjzec5yji6nkec7e8_S2an3njyeqr4aynwkze5krcejna6ikca8
bo$2bo$3o!
runs in Molly. Here they are after 10 ticks, from Golly and from Molly:
Code: Select all
x = 63, y = 23, rule = B02ci3anyecr4yikrqtjzec5yji6nkec7e8/S2an3njyeqr4aynwkze5krcejna6ikca8
obobob3obobob3obobobo17bobobob3obobob3obobobo$2o3bo2bobobobob2o3bo18b
2o3bo2bobobobob2o3bo$bo2b2o2bob3ob4o2b2o19bo2b2o2bob3ob4o2b2o$2obo2b2o
3bo3b2ob4o18b2obo2b2o3bo3b2ob4o$2b3obob3o4b2o3b2o20b3obob3o4b2o3b2o$2o
b5o2bo4b2o3bo19b2ob5o2bo4b2o3bo$b3o3b2o7bo4bo19b3o3b2o7bo4bo$2obob5o2b
o2b2o2bo2bo17b2obob5o2bo2b2o2bo2bo$3o3b3o3b3ob2obo2bo17b3o3b3o3b3ob2ob
o2bo$3o7b6o5bo18b3o7b6o5bo$2b2o3b2ob2o2b3o3bo21b2o3b2ob2o2b3o3bo$2ob6o
b5o25b2ob6ob5o$5bo4bo2bobobo27bo4bo2bobobo$2ob2o3bo2bo2b2ob4o19b2ob2o
3bo2bo2b2ob4o$2b2o3b3ob2ob2o2bo23b2o3b3ob2ob2o2bo$2obobob6ob4ob2obo17b
2obobob6ob4ob2obo$2obobob2ob2o2bo3bob3o17b2obobob2ob2o2bo3bob3o$2o2b2o
4b2obo2bo23b2o2b2o4b2obo2bo$2b4o4b2o2bobo3bo21b4o4b2o2bobo3bo$3ob8obob
o24b3ob8obobo$4b3ob3o2b2obo2b3o22b3ob3o2b2obo2b3o$2obob5obob2ob6o18b2o
bob5obob2ob6o$2bobo2bo2bobo2bo2bobo21bobo2bo2bobo2bo2bobo!
Notice that this is
not the same as what you get when you run the un-reversed glider in Golly.
The difference is in the first tick, when Golly turns all cells ON using the S0.../B...8 rule, and then keeps using that same rule.
Molly is basically assuming that at T=0, the pattern is already a rule-reversed pattern -- i.e., that the OFF cells are the interesting part, so the inverted+reversed rule should be applied immediately. This appears to be what Golly does also, for non-totalistic rules -- for example, try setting the rule to "AntiLife".
Seems to me that Golly would have to apply a different rule for the first tick, to produce ON cells to match the OFF cells that would ordinarily (using the isotropic-rule-gen.py rule table) be generated during the Big-Bang universal inversion at T=1.
I'm not quite sure what that different rule would be -- the simple bit-inversion of the original rule, I think, without the [8-n] reversal. But given Golly's current behavior with the "AntiLife" rule, maybe we shouldn't do that, anyway...?