Darwinbots Forum
Code center => Bugs and fixes => Topic started by: Testlund on April 21, 2006, 12:57:24 PM
-
Everyonce in awhile some bot appears fixed on the screen while the rest are affected by brownian motion normally. It appears to me to be a bug, or maybe the bot has evolved to be fixed, wich I think is odd. Here's a gene for a fixed bot:
sub inc
cond
xor
dist * store
xor
add mult ~ 1 sgn or
ceil
-
If I understand right it shouldn't be doing anything because it uses a cond without a start or else.
Could be a bug. Try "save organism" on the fixed bot and upload it, and we'll see if it's fixed or not.
-
Here it is.
-
Save organisms seems to be corrupted somehow, (info like waste, body, etc. was lost) but I did manage to find out that yes, it is fixed some how.
-
Maybe that's something for Eric to fix? In the mean time I just delete them.
-
They might be adapting.
What's your waste threshold set at? Alzheimers seems the most obvious candidate to me. Try disabling it (I think -1 disables it).
-
I guess you mean changing treshold number to -1? I'll try that and see what happens. Later on I whould like to be able to use waste tredshold though. In this sim I'm running I guess I can disable it as a workaround. The bots are too primitive to do anything anyway.
-
Think of it primarily as a bug check. If you disable it and bots are still fixing from time to time it might mean we have a major bug somewhere that's going stealth
-
I've got this bug too .. in 2.37.6 ... but the other way round. I found that sometimes my vegs in long evosims somehow got unfixed, even though the other bots didn't + shoot them or unfix them in any other way. It was weird. I just checked back one time and all the vegs were free floating.
-
Maybe we should have a system level fixed/unfixed that bots can't change if you don't want them to. As it is bots are only fixed/unfixed depending on what they store in .fixpos.
-
That sounds like a good idea. There may be times when you want stuff to stay fixed from the control pannel. It would be simple enough to have a toggle that locks the locked position.
-
We can make the toggle on the species tab in options work that way.
-
We'd have to change veggies then to fix themselves if we overload the existing toggle. Otherwise, battery bots couldn't unfix them.
I'll noddle on it and maybe do something like this in the next drop, add a new toggle or something...
-
We'd have to change veggies then to fix themselves if we overload the existing toggle. Otherwise, battery bots couldn't unfix them.
I'll noddle on it and maybe do something like this in the next drop, add a new toggle or something...
This would be pretty easy to do.
Find the place where information shots are added to memory locations and add a teensy little bypass routine so that a shot designed to unfix the veggie, actually reverses the boolean state of the toggle too. Internal memory location changes from DNA (ie. Altzheimer) could be overridden by the toggle.
-
Think of it primarily as a bug check. If you disable it and bots are still fixing from time to time it might mean we have a major bug somewhere that's going stealth
I'm afraid it's a serious bug then. Just found another fixed bot, despite that I have waste treshold set to -1. Usually it is rare, sometimes a little more often that the bots get fixed. Maybe you have allready find a solution to it.
-
What's the offender's genome?
If the genome doesn't look like it should be able to do anything and it's not alzheimers, we have a problem.
-
Oh, yeah, I forgot! Here's the gene:
& >
* sub *.dx or
*.shoot inc
and
cond
++ << 3 *.up <
I have the saved bot too if you need it.
-
Hm, I don't know then. It's going to be hard to track it. Maybe Eric should run a sim where bot DNA doesn't execute and waste threshold is disabled. If it's still occassionally fixing bots...
The problem is that there are so many potential sources that bots could get fixed through. DNA, alzheimers, shots, ties, divine intervention...