duel sim eco

[Shadowgod2]

so i had an idea about evolving complex ecosystems. have 2 separately evolving sims starting from the same or close to the same settings and bots(this would be similar to my duel evo sims both plant and animal evolving). when they both evolve a good bit to be considered different species, put in a temporary teleport in both to allow interactions between them. after a short time then delete the teleports and allow them to evolve separately for a while once more and repeat. what do you guys think about it and if anyone has done anything like this what was their results.. maybe this will get the forum a bit more active for a little bit...

[Botsareus]

I am working on native predator prey systems aka project green light when I have time. Should help you out with what you are trying to simulate.

[Shadowgod2]

cool and hows that going? i know they can be tough to get stable enough to evolve... kinda why i thought of this, maybe to evolve it in some way

[Botsareus]

I am thinking of a good time to publish a video but I need to be sure I have time to finish coding it. The video part was easy but it changes stuff enough that I may as well evolve the robots I am currently evolving from scratch. Therefore, for now

[Botsareus]

off topic: One of the features I will need beta testing on is my fix to have both start and else in one gene. The code is rather messy there and very easy to break.

Code: [Select]

Cond
...
Start
...
Else
...
Stop


[Shadowgod2]

Yea i noticed that too.. it always seems to act more as another start or nothing at all now then else.

I would definitely like to see a new vid out. Other than the swarms the videos that are out dont really have much interest in them. Non seem to be about evolving a good deal..

[spike43884]

Seems an interesting idea. Not sure if I'll have time to go running an instance of it though, sorry. 

[Botsareus]

That is why I have more then one comp. 

[Shadowgod2]

lol i just run it in background all day especially if im not using it...(which for me is most of the time though)
edit: also ive finally got some settings that i like for the 2 sims and so far only played for today. they already are taking different approaches in evolving but mostly by the alga... one has opted for bigger body sizes and fewer numbers while the other with about the same mutations has went for more numbers with about 2 mutations. the animal unfortunately has not yet gained even 1 avg mutation.. well lets see how it goes over night. last one i tried crashed with no bots on either screen and graphs cleared out. on the base simulation i had it to where they never really reached the nrg cap even at their highest. not so much anymore... might have to up the cap a little to keep things interesting and allow more room for development

[Shadowgod2]

day 2: 1 had crashed and was restarted. the other only 1 new average alga mutation and still zip for animal.... come on animal you're falling behind, i might have to up the mutation rates

[Botsareus]

Just out of curiosity, what error did it crash with?

[spike43884]

Im wondering if you might be able to over-complicate the animal_minimalis code so it does the same but in lots more code to make it more prone to mutation but less fragile?

[Shadowgod2]

actually the program its self didnt crash, there were no bots in the simulation so im gessing that the alga put itself in a fatal position..

Spike i might have to do that... thanks for the suggestion

[spike43884]

actually the program its self didnt crash, there were no bots in the simulation so im gessing that the alga put itself in a fatal position..

Spike i might have to do that... thanks for the suggestion

No problem, I think in the past I've tried throwing in redundant code and such into bots, but I've never had particularly good evolution sims occur. Any overcomplication of code and addition of redundant code should make deletions less likely to kill off the species while allowing the same amount of mutations as a lower % of code is lost.

[Shadowgod2]

well ive started another go but went a different rout.. instead of using animal min... i used a much simpler bot and had it in a populations stable, high mutations environment for a while till it had around 30 mutations. i saved the bots and started a new sim split.. the mutations rate is still high but not as high as the starter sim at around 1 stable or permanent mutation every 10 generations. although behavior hasn't changed much yet they are very different bots now in their dna currently at around 30 mutations. the color verifies this but one being yellow and the other almost black when they both started out red. i'm thinking of either letting them evolve till they have 100 mutations or just different behaviors. oddly the alga haven't changed much in color.. still green at about 25 mutation.