Messages

61

Bot Tavern / Shots

« on: October 04, 2008, 10:09:28 PM »

Shot flavor can see the specific memory location modified. You could see if the enemy is firing things into .shootval and react accordingly.

I actually made a bot that uses this technique, does pretty well right now against most of the league bots except for fruit fly and Etch. Its actually effective against both insta-kill and normal shot bots. For shot-bots just fire an insta-kill shot into shootval, for the Insta-kills it forces the bot to fire a feeding shot.

Could do something like *.shflav * to find how the location itself was affected, would have to do those calculations first though, otherwise later code could mess it up.

Weird how Info-shots got to be the big weapon, seems like they've been hardly used before.

The real issue with shootval is that it only resets if fired. Otherwise the bot has to reset shootval itself, and none of the bots are doing that.

Note to Self: Reset shootval...

62

Simulation Emporium / Congo Lines

« on: October 04, 2008, 05:27:50 PM »

Really oddball behavior that turned up in a zerobot sim. Basically the two congo lines of bots along the edges will switch sides in the sim every so often moving in unison with a group bots on the opposite side. It "seems" to be tied to a low population level, but could also be using timer somehow to co-ordinate the movement.

Presumably the behavior did something useful for the bots when the sim was fully torroidal, but it wasn't as noticeable. The bots got part of their dna from Ikke's zerobots, so may look familiar.

63

Short bots / Fruit Flies 0.1 (F1) (Moonfisher) 23-09-08

« on: September 26, 2008, 10:58:50 AM »

That pretty cool.   I've got a bot that uses shflav to swap shot attacks, I'll have to try something similar for ties and see how it does.

64

DNA - General / A condition in one gene is having effect on another?

« on: September 08, 2008, 03:44:29 PM »

Does look weird if you don't know about the robage zero eye thing. Had me going there for a second, below is the rewritten code. The first gene is fine, the fourth gene was rewritten.

Code: [Select]

cond
*.refeye *.myeye =
*.eye5 0 >
start
stop

'multi control center
cond
*.multi 1 = or
start
50 .sharenrg store
150 .fixlen store
100 .stifftie store
628 .fixang store
stop

'if alone create a second part
cond
*.numties 0 =
*.robage 0 !=
start
50 .repro store
stop

'second part is recombining with it creator.
cond
*.robage 0 =
'removed line below, invalid at birth unless bot has zero eyes
'*.refeye *.myeye =
start
'made tie inc to save energy
.tie inc
stop

'if enough energie split.
cond
*.nrg 30000 >
start
.deltie dec
stop
The MB tutorial I believe I wrote, I honestly only vaguely remember it now... PY's tutorial written for a single bot is on the wiki too and is still pretty useful to look at. Both are pretty old. If I can ever figure out the new eye stuff I might write a new one.

65

DNA - General / A condition in one gene is having effect on another?

« on: September 07, 2008, 01:28:46 PM »

Need to rewrite the birth tie gene, refeye isn't updated on robage zero so the values aren't equal. The inline comment isn't really a good idea. It can be done but it might mess up the stack some.

Code: [Select]

//second part is recombining with it creator.
cond
*.robage 0 =
start
.tie inc
stop

66

Interesting behaviour bots / MicroNet2

« on: September 05, 2008, 03:13:31 PM »

Odd micro net forming bot, heads towards the direction of the veggies and ties to them then pushes them along, while carrying with them the rest of the bots in the mb. Works best in torroidal mode.

Code: [Select]

'MicroNet2
cond
start
'31997 rnd 2 add .setboy store
.delgene .memloc *.robage sgn 1 sub - mult store
.delgene .tmemloc *.robage sgn 1 sub - mult store
-715 .delgene *.delgene -715 sub sgn abs mult store

'firing
-6 -7 abs -505 abs * sgn mult *.memval -715 sub sgn abs mult store

'tie
-330 abs *.refshoot sgn sub inc

'primary food gathering
100 50 *.tmemval -715 sub sgn mult sub .sharenrg *.multi mult store

'repro
50 .repro *.nrg 2000 sub sgn 0 floor mult store
 clearbool
200 *.shell sub 0 floor 200 ceil 200 *.shell sub sgn abs .mkshell mult store


'acceleration
-505 abs * -1 abs -505 abs * sgn mult store

'start of tie section
*.tiepres .tienum *.tiepres sgn mult store
stop


cond
*.body 3 >
start
*.body 1 sub .fdbody store
stop

cond
*.body 2 <
start
1 .strbody store
stop

cond
*.repro 0 >
start
0 -1 abs store
628 .setaim store
stop

cond
*.tienum 0 >
*.numties 0 >
*.tmemval -715 !=
start
1000 .tieval store
.tieval .tieloc store
stop

'turning
cond
*.memval -715 =
-505 abs * 0 >
or
start
*.refxpos *.refypos angle .setaim store
stop

'backup for above
cond
start
-715 .delgene *.delgene -715 sub sgn abs mult store
stop

end

67

Veggies / Plankton Kinetic (MB/Veggie)

« on: July 23, 2008, 07:55:18 PM »

Code: [Select]

cond
*.readtie 0 >
*.robage *.trefage <

*.in5 .out5 > and

*.tin5 *.tout5 > xor
start

*.tin5 *.tout5 store

*.in5 .out5 store
stop
I'm guessing that you meant for tin5 to be stored in tout5 itself and not the reference. Whenever *.tout5 is 1-4 it'll cause the bot to move. Did you mean to compare *.in5 with the location of .out5 ? Not sure what you wanted here...

Might want to add a waste shooting or a share waste method, in the rare case there's more than the veggie can convert, not too likely to occur though...

68

Simulation Emporium / Q: shortest zerobot reproduction time

« on: July 23, 2008, 01:15:17 AM »

Just wanted to say that you can get alot of the same effects from waste build-up. Reproduction and share-feeding(w/ties) are the easiest to get this way. Probably similar to what happened in nature, a chemical process promoting replication prior to a self-replicator appearing.

The bots do seem to favor either continous or periodic actions over normal conditional logic. I think most of has to do with the problems associated with conditions breaking down from mutations, might be interesting to run a test to see.

69

Bot Tavern / The amazing Antbot

« on: July 23, 2008, 12:21:15 AM »

Why not just fix the queens position at her birth. Then just unfix all the non-queens.

70

Suggestions / Color Again

« on: July 20, 2008, 08:41:19 PM »

Well, if it was something like the actual "species/mutant color" I could understand, but in this case it would just be something like in/out but with the added bonus that we could see it ourselves. The outer color still wouldn't be visible to the bots(yes, I know seeing the inside but not the outside is a bit odd) just like it is now.

Could be used for some pretty impressive things like showing gene activations across an entire species, or just to give us a better idea how their conspec ID's are changing over time.

71

Suggestions / Color Again

« on: July 20, 2008, 03:51:22 AM »

Had an idea for the bots being able to use signalling colors. Yes, I know it's been suggested before, but the problem of its effecting the mutation colors was a major stumbling block.

What I was thinking was instead of changing the outer color of the bots, we could set it to change the inner color of the bots. It's already redrawn each cycle anyways(add a bkground pic to see), so it shouldn't be any additional processing time required than for the additional dna code itself. It wouldn't be too different than how in/out currently work, but it could be useful for us as well by providing a direct indication of something occuring in the dna.

Not sure how the sysvars would be set up there's more cominations than we could fit into a single sysvar so multiple ones would be likely needed.

72

Simulation Emporium / my first evo sim

« on: July 17, 2008, 12:42:46 AM »

I normally just use evolution to look for "improved" dna I can make use of in a standard bot. You can't take this too far though, since alot of the improvements evolution comes up with won't be useful in a normal F1 match.

73

Tips and Tricks / Random Stuff

« on: July 14, 2008, 10:12:53 PM »

It could be thought of as a form of quorum sensing. Though I don't think that nature has such an exact mechanism for determining an exact number of cells. I do like being able to use it for IM(whenever my net computer works  again  ), you can tell when your bot has just arrived or when it's achieved a level of dominance in another sim.

74

Tips and Tricks / Random Stuff

« on: July 14, 2008, 03:53:06 AM »

Not much happening in the forum right now...

Figured I'd take the opportunity to post on a few things I've been working on.

Getting more use out of delgene and vshoot:

Started using both vshoot and delgene for conspec ID via memloc/memval. At the start of their life both sysvars are set to different large negative values, this helps prevent viruses that only use inc I addition to allowing the bot the ability to safely use these for recognition. Since these values are always required to be activly checked for changes anyways(normally just being reset to zero) this combines resetting the recognition system with protecting the bot from viruses. Memloc is set to one of the two values at the start of their life as well, the actual conditions use the refrence of memloc to know which value to check against.

My virus, Myself:

In the same vein as delgene and vshoot, its actually better not to use zero for virus prevention. For 1G bots it's best to simply make your own gene at the start and whenever mkvirus is not equal to your own gene with vtimer  equal to zero. In this way your bot can prevent regular viruses from utilizing mkvirus. While foreign viruses can still activate vshoot, the worst that'll happen is your bots will launch their own dna out with a decent chance of "infecting" other bots or alga.

Turning without turning:

Been working on a bot that doesn't actually rotate itself but instead moves it's eye5dir for sighting other bots, it's more of a beta-bot than anything else but it's interesting to see. Seems pretty decent they're able to move at a constant velocity, while still remaining on the lookout for food. Once it actually spots something it uses the standard setaim method for aiming. Still a work in progress as it continues to have problems with edges and the turning code could be better. Could likely also be useful for a MB like SeaSnake since the back bots could still be on the lookout for food to inform the head about.

Checker Veggie

Made a better veggie, it uses *.totalmyspecies to reproduce when it's total population is below it's old highest value. Not completly effective due to the slight surge that'll occur but I've found it works better than the standard energy check or simply trying to constantly repro.

I still think something should be done to prevent .totalmyspecies from being stored into memloc for leagues. Seems like it should still be able to be used, just not as a unique species number.

Alrighty, all I have for now. Hopefully this'll break the July 4th spell that seems to have settled over the place.

75

Bug reports / Pond mode doesn't turn on buoyancy?

« on: July 01, 2008, 01:22:17 AM »

Got the same bug in mine, pissed me off.   I was hoping to use rdbuoy for sneaky conspec recognition(via memloc/memval), and sadly couldn't.