Bots and Simulations > The Gene depository

Messaging Bots

(1/2) > >>

Endy:
I think it'd be possible for the bots to cheaply send complex messages to each other.

The basic method could use just out1/2 changed to either 1, 0, -1(via inc/dec) to yield some 9 different messages.

The response could be seen as a single number with this code:

*.in1 10 mult *.in2 add

A more advanced method could use memval to yield some 27 different messages (or oddly enough, 26 discounting 0,0). Bots tied together could in addition use tmemval to yield up to a maximum of 81 messages.

*.tmemval 1000 mult *.memval 100 mult add *.in1 10 mult add *.in2 add

Now the only question remaining is what should these guys be talking about? :)

Endy B)

P.S.
I haven't actually checked up to 81, 27 took long enough, the equation for it looks like the number of numbers to the power equal to the number of available spaces.
x^y

Numsgil:
Yep, you're working in a base 3 number system, so all the values possible will be powers of 3.

It's an interesting idea.  It would be most useful with some kind of built in Finite State Machine.

Like the bot has 7 different modes.  Each mode has its own genes.  It switches modes based on input from other bots.

I'm probably going to add section conditions that turn on/off large sections of DNA for this purpose.  I discussed it a long time ago, but just haven't gotten around to it yet.

Light:
Excalibur talks to its self

Numsgil:
So do I.

Endy:
So both Excalibur and Nums talk to themselves :pengysmiley:

Made a deliberate semi-Canni as an Evobot(kind of like the clan bots but has a chance of agression with children). So far it still became a true Canni, although that might be because the turning genes are woefully incomplete(right now it just uses the  refxpos/ypos setaim gene).

Thanks Nums for confirming the pattern. I was almost to the point of actually doing it by hand, to figure out the numbers possible.

Endy B)

Navigation

[0] Message Index

[#] Next page