Author Topic: Strange Grouping  (Read 2724 times)

Offline Jez

  • Bot Overlord
  • ****
  • Posts: 788
    • View Profile
Strange Grouping
« on: August 30, 2006, 10:31:19 AM »
I've had strange behaviour from bots before, my coding is haphazard at the best of times, but this time I was impressed. I took a 1g bot, Blue on Blue, Bubbles and a version of Zebedee and pitted them all against each other. The 1g bot and Zebedee got slaughtered but the funny thing was that Blue on Blue often managed to get to 30 genes, same as Bubbles. (F1 conditions)

That's when things get a bit strange, everytime I've watched this Bubbles ends up forming a mass (as tighty packed as bots can get (wish ties allowed same:))) in the lower left of the screen, all the other Bubbles drift towards this point and Blue on Blue, if it's not surrounded by Bubbles, slowly increases in numbers until Bubbles is destroyed.

Doesn't sound so strange in writing I guess!

Anyway if anyone wants to replicate this the save game is posted here:
« Last Edit: August 30, 2006, 10:40:18 AM by Jez »
If you try and take a cat apart to see how it works, the first thing you have in your hands is a non-working cat.
Douglas Adams

Offline PurpleYouko

  • Bot God
  • *****
  • Posts: 2554
    • View Profile
    • http://kuro-tejina.com
Strange Grouping
« Reply #1 on: August 31, 2006, 08:49:10 AM »
Quote
(as tighty packed as bots can get (wish ties allowed same:)))
How do you mean?
Tie lengths can be set extremely short.
There are 10 kinds of people in the world

Those who understand binary.

and those who don't

:D PY :D

Offline Jez

  • Bot Overlord
  • ****
  • Posts: 788
    • View Profile
Strange Grouping
« Reply #2 on: August 31, 2006, 10:30:35 AM »
Ah, my bad then. Guessing that perhaps I needed to use negative lengths?
Was trying before to get ties as short as possible with no flex, as in bots touching each other.
I'll have to try ties again.  
If you try and take a cat apart to see how it works, the first thing you have in your hands is a non-working cat.
Douglas Adams

Offline PurpleYouko

  • Bot God
  • *****
  • Posts: 2554
    • View Profile
    • http://kuro-tejina.com
Strange Grouping
« Reply #3 on: August 31, 2006, 11:34:35 AM »
It's a bit tricky, granted.

use the fixlen commands and remember that the ties have to harden before they work.
You might have more luck in 2.4x as we know of that bug that Helios uses in 2.3x to make ties grow very long, very fast without waiting for hardening.
There are 10 kinds of people in the world

Those who understand binary.

and those who don't

:D PY :D

Offline Testlund

  • Bot God
  • *****
  • Posts: 1574
    • View Profile
    • Games And Electronica
Strange Grouping
« Reply #4 on: September 24, 2006, 06:33:53 PM »
You mean like this, PY? I thought it was fixed so ties couldn't grow this long. It's strange how different bots seem to affect the program differently.

And in this sim it slowed down to a freeze allready in the beginning on my new fast computer. :-(
« Last Edit: September 24, 2006, 06:35:31 PM by Testlund »
"God is an ever receding pocket of scientific ignorence." - Neil DeGrasse Tyson

"God is a kid with an ant farm" - Constantine

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7718
    • View Profile
Strange Grouping
« Reply #5 on: September 24, 2006, 07:19:20 PM »
5000 bots do have a way of slowing the computer down

And you're right that ties shouldn't grow that long.  It's a bug.

Offline Testlund

  • Bot God
  • *****
  • Posts: 1574
    • View Profile
    • Games And Electronica
Strange Grouping
« Reply #6 on: September 25, 2006, 08:14:11 AM »
It only happend the first couple of times, after that when I started a new sim all bots just died within seconds. I was hoping to run some interesting ecosim with these bots but it didn't work. Some of those bots are probably made for the 2.37 version.
"God is an ever receding pocket of scientific ignorence." - Neil DeGrasse Tyson

"God is a kid with an ant farm" - Constantine

Offline PurpleYouko

  • Bot God
  • *****
  • Posts: 2554
    • View Profile
    • http://kuro-tejina.com
Strange Grouping
« Reply #7 on: September 29, 2006, 11:54:21 AM »
Ties should break if they get longer than 1000. for a while they were forming at any distance then immediately breaking again. In 2.37 this doesn't happen. Don't know about other versions. Haven't had much of a chance to play lately  
There are 10 kinds of people in the world

Those who understand binary.

and those who don't

:D PY :D

Offline Elite

  • Bot Overlord
  • ****
  • Posts: 532
    • View Profile
Strange Grouping
« Reply #8 on: October 01, 2006, 12:40:29 PM »
I can't seem to get ties any shorter than a few bot lengths in 2.4

In fact, I can't seem to change the tie lengths at all

Ties are very buggy generally, in most versions

Offline Numsgil

  • Administrator
  • Bot God
  • *****
  • Posts: 7718
    • View Profile
Strange Grouping
« Reply #9 on: October 01, 2006, 05:36:22 PM »
2.4 shouldn't be buggy as far as ties.  Not that ties are 100% logical.

Offline EricL

  • Administrator
  • Bot God
  • *****
  • Posts: 2266
    • View Profile
Strange Grouping
« Reply #10 on: October 02, 2006, 12:04:13 PM »
Quote from: Elite
I can't seem to get ties any shorter than a few bot lengths in 2.4

In fact, I can't seem to change the tie lengths at all
I would be happy to investigate if you could provide a bot code segment that does not behave as you expect...
Many beers....

Offline Elite

  • Bot Overlord
  • ****
  • Posts: 532
    • View Profile
Strange Grouping
« Reply #11 on: October 03, 2006, 11:13:47 AM »
Negative values of .fixlen seem to be absoluted

But otherwise, a quick test seems to indicate .fixlen is working properly, and it's my bot coding that's causing the errors  

One thing I did notice that was a problem, however, is that when two bots are tied together in a line, it is impossible to turn.

cond
*.numties 0 =
start
50 .repro store
1 .out1 store
stop

cond
*.robage 0 =
start
7 .tie store
stop

cond
*.multi 0 >
start
7 .readtie store
7 .tienum store
200 .fixlen store
628 .fixang store
stop

cond
*.multi 0 >
start
20 .dx store
stop

^ It should turn, but it doesn't