Bots and Simulations > DNA - General

Top bots who are still build truely modular?

<< < (2/3) > >>

Peter:
Well, as multiply is my bot, I am going to post too.

Believe it or not animal-minimalis was a starting point, it became better.

You are very true in your statements. When I posted multiply I thought it was corrected, but it wasn't. Beliefe it or not it did even with that awful mistake a pretty good run in the leagues.

Further the metabolism of multiply is pretty bad. Don't use it. I have done multiple inprovements on multiply, where I also found, becouse I simply didn't test multiply really through. Multiply hasn't got a real power. With testing I just found that it is pretty normal in every region and the combination did couse a third place in F2(tested on 2.43). A new version is coming.

I have documentated my mult a little. I will make some inprovements you will see it in the beastery soon.

Peter:
Okay, sorry Welwordian it took so long, was busy with some other more importent stuff.

Here you got Multiply4, right now not heavily documentated. But with a base knowledge of DB-dna you should be fine. Ask anything you do not understand. I should know it, (I hope  )

Edit: Is there a reason everybody is posting SG-bots instead of modular, I find the Sg-bots hard to read. Maca, why are you for example building SG-bots, I can't read them.

And a fact is that it beats F1 and F2, Jippie.

Multiply4

Numsgil:
Aside from just being a challenge, SG bots are easier to defend against viruses: there's only two places the virus can insert itself, so it's real easy to delete anything that's not placed in the proper position.

EricL:
Two small points:

The new in-line conditional paradym (and new stack operators) make writing SG bots much eaiser (and if done right, much less breakable by point changes).

Traditional SG bots are brittle in the face of mutations.  Gene compartmentalization provides some resilance to changes.

Peter:

--- Quote from: Numsgil ---Aside from just being a challenge, SG bots are easier to defend against viruses: there's only two places the virus can insert itself, so it's real easy to delete anything that's not placed in the proper position.
--- End quote ---
Well, I am not sure if the effort is worth it for the extra virus-protection, Multiply has pretty fast defeated  Etch, Reaper, Aura and even Guardian pretty fast. I am not sure if guardian is using a real virus, it was just loading up a virus the whole time.  


--- Quote from: EricL ---The new in-line conditional paradym (and new stack operators) make writing SG bots much eaiser (and if done right, much less breakable by point changes).
--- End quote ---
I have seen it, I haven't used it for the bot becouse it is most times using multiple conditions and then I thought a simple cond-start-stop gene works better. But are bot created with it even seen as SG-bots, to me it seems like they have multiple build in cond-start blocks.


--- Quote ---Traditional SG bots are brittle in the face of mutations.  Gene compartmentalization provides some resilance to changes.
--- End quote ---
I there are brittle, then just make modular bots

Navigation

[0] Message Index

[#] Next page

[*] Previous page