Darwinbots Forum
Code center => Suggestions => Topic started by: Botsareus on January 13, 2013, 12:10:31 PM
-
The next release DB2.45.03Beta will focus on making a sexually reproducing Eco-system in Pond Mode.
There was an idea to add shadows to Pond Mode. I kinda liked that idea and am planning to implement something similar.
My idea is to have eye strength diminish the deeper in the pond a robot is based on light intensity at that level.
Following the same thought I have decided to reduce vision by, say 20%, for robots during night time (day and night cycles) quote:
"i don't think that any animals can fully see in the dark but some have improved night vision like cats"
If you have specific comments please post 'em.
Thank you for your participation in advance.
-
[Bump]
Working on a buoyancy fix now. Buoyancy will cost energy. If all conditions for buoyancy are met the formula is:
Robot's distance from top (calculated from it's setboy level) / Top down distance * the energy loss to fight gravity with voluntary moment ( Y-gravity calculated in ) at a given mass ( chloroplast effect on mass not factored in)
-
Is the amount set in buoyancy(.setboy) correspond to a level the bot floats at, or does it offset the mass of the it.
Should nrg be charged constantly or just when bots change their buoyancy. I think the latter.
-
No, .setboy only sets the level at witch the robot floats at. It has nothing to do with the mass since mass effects only to z-axis gravity (not y-axis gravity) but I still want mass to be calculated into the formula.
Should nrg be charged constantly or just when bots change their buoyancy. I think the latter.
Think of a hot air balloon, you constantly have to keep the air worm or the balloon will sink.
The thing is I don't want buoyancy to end up working just like a '90degree .setaim and N .up' commands , but I do want to charge it with voluntary movement costs.
Robots will start at zero (button of screen), a max value of absolute 2000 means float to the top of the screen.
Hmm, I am still considering it though. Can you give me a formula you had in mind? I'll think of something myself.
I am thinking we will take voluntary movement cost, multiply that by mass, multiply that by Y-gravity, multiply that by a value. But what should the value be if a robot can remain buoyant indefensibly?
I don't want robots to slowly lose buoyancy because this yet again makes it look like the '90degree .setaim and N .up' commands.
I am going to use Y gravity itself to float the robot, I have not tested this yet but the idea is:
Let's say the robot is set to half way up the screen.
If the robot is above that point the Y gravity is down.
If the robot is below that point the Y gravity is up. (reversed)
The idea is, the robot should start bouncing at half way up the screen, and if there is enough friction (this one will be fun to test) the robot should stabilize at half way up the screen.
-
Ok, since buoyancy auto-magically keeps them at a certain level, then it should be charged constantly.
("or does it offset the mass of the it," I meant weight, like a balloon like you present; although, if in a liquid it would be just air and not leak that much.
Then the bot would spend nrg keeping itself at the desired level with condition statements, extra dna, and the cost of adding buoyancy)
-
[Bump]
Just added the new 'debugint' advanced command to make the DNA (including boyancy and tie ang tie len 1 ... 5 ) easier to debug. The new system will be accessible trough the console window, see picky. Don't worry, the mutator will be blocked from evolving new 'debugint' commands.
-
[Bump]
Few, fully implemented buoyancy
check out this little crazy formula:
.nrg = .nrg - (SimOpts.Ygravity / (SimOpts.PhysMoving) * ((.body / 1000) + (.shell / 200)) * SimOpts.Costs(MOVECOST) * SimOpts.Costs(COSTMULTIPLIER)) * rob(n).Bouyancy
working on tielen tieang 1...4 fix next, this will be the only 10 commands that differentiate between input and output, should make cool MBs easy to develop
-
[Bump]
Had a bad few weeks, resuming work soon.
-
[Bump]
There was a bug where a multibot loaded back into a simulation breaks up. -fixed
-
[Bump]
Right now I am working on a problem where the caterpillar MB does not save or load correctly. I may need some help on this later because there is a good chance I will not be able to figure this one out.
This is a very old bug , I tested all the way back to version 2.44
This bug is very easy to replicate:
Load a single caterpillar with no other robots into a simulation. Exit DB, Run DB and select "continue simulation" The caterpillar either breaks itself apart or stops moving.
You can also duplicate the same bug by saving and loading a .dbo file.
-
[BUMP]
Fixed, it turns out the ties where being created 2 cycles after the multi-bots where loaded. This creates errors in tin tout commands which caused the problem.
-
[Bump]
Just submitted a bunch of fixes and improvements, including fixes for slime and venom.
-
i think that modifying eye strength based on light levels is a good idea because then there can be nocturnal bots and stuff
-
[BUMP] check this little robot out, I created this MB to check my tie code:
-
[Bump]
Just submitted a fix for tielen...tieang 1...4, all sysvars and commands should work now, except for -8 shots, that's what I am fixing next. I will also fix genetic distance graphs in the process.
-
I used Darwin2.45.02F.
1a. The changing eye strength looks nice; there is just one thing. Can you make it so the eye strength is unaffected with less light intensity. I want to lower the intensity so plants don't make as much nrg but still have bots be able to see.
With less than 100 (i.e. 10) now the eyes don't see as far, even with sediment set to 1.
1b. I guess making eye strength based on the fraction of intensity/sediment and depth?
2. On buoyancy, should the bot be charged nrg if it is above the .setboy level? In F, it is charged.
Bugs:
1. Setting sediment level equal or greater than intensity produces run time error 5. (Invalid Procedure call or argument)
Besides 0 intensity and 0 sediment.
2. The bot gets 1 nrg when intensity is set to 0 at any depth (when set to repopulate).
Thanks your hard work on implementing awesome features in DB. I may try to help in the future, after I figure out vb and the source code :Uhhhhh:.
p.s. Are chloroplasts implemented in F, I see the settings but didn't get .chlr to work?
p.p.s. Muahah, I'm a bot builder. 8)
-
2. On buoyancy, should the bot be charged nrg if it is above the .setboy level? In F, it is charged.
I think they should, think back to the air balloon experiment and add rocket engines, You have a certain buoyancy in the balloon, then you fire up your rocket engines and drift upward, yet the buoyancy remains the same.
Here is another real life scenario which totally wastes energy: A balloon is maxed out on it's air heat and floats all the way up to the ozone layer. Then it fires it's rocket engines to drift down-ward. Now this balloon is wasting both rocket fuel and keep the air hot but floats half way up to the ozone layer 8)
The changing eye strength looks nice; there is just one thing. Can you make it so the eye strength is unaffected with less light intensity. I want to lower the intensity so plants don't make as much nrg but still have bots be able to see.
Good point, I will have to research this further.
Update:
I am really interested on how sediment effects eye strength only. Therefor, I really hate the idea that 'depth' by itself should have ANY effect on eye strength. How about a formula where ONLY density and not NOT light intensity effect eye strength?
The bot gets 1 nrg when intensity is set to 0 at any depth (when set to repopulate).
This problem goes back all the way to version 2.45.01 as far as I checked. I really see no point in it.
Should I make it zero?
1. Setting sediment level equal or greater than intensity produces run time error 5. (Invalid Procedure call or argument)
Besides 0 intensity and 0 sediment.
You should never set sediment level greater then 200 or less then 0, I forgot to cap the values.
As far as setting sediment level greater then intensity, I was not able to duplicate this error. see picy. Can you tell me the exact values you used? Or better yet, can you send me the settings file?
p.s.
Thanks for pointing this things out. :) I had to dull down the chloroplasts explanations for the next version so the users will have an easier time to transition and no they are not yet implemented :(
-
Hey Tilthanseco, let me know if this works better for you.
If not, I can always adjust the value OR put the old code back and make the whole eye strength thing optional.
-
Yep, I like that better.
I could not reproduce the error either for some reason, wish I did save the settings.
I used (100,100) (100,101) and (1,1) to get the error. But those don't make the error anymore. I will post again if it comes up again.
-
ah I found it, it has to do with dragging a robot to the top of the screen, you may get an invalid procedure or call of argument, fixing now.
-
Now I think you should make the eye thing optional or lower the affect of depth.
My poor bots are almost blind from the middle down using: 96000 x 72000; 100 intensity; 0.1 Sediment.
-
I ended up correcting for screen size, tell me if this is any better...
-
But how will we know what level the bot is floating?? On the other hand, it would be real cool to have different species at different levels, like bottom-feeders and the like.
-
But how will we know what level the bot is floating?
You can tell by looking on the screen;
If you are asking how to return to the dna what is the current buoyancy level, then use 'rdboy'.
If you need to figure out vertical position and return it to the DNA use 'ypos' aka 'depth'.
The button line is the robot needs to check both its rdboy and its depth to figure out the size of the sim.
-
if you are asking what will happen if there is no Y-Gravty:
The robot can move down screen until it reaches a wrap point and go back to the top and calculate the screen dimensions that way. Or, it can go down screen screen and hit the edge, use .edge command to figure out dimensions.
-
Lol, Yes it is better, I think.
-
Don't worry guys, I'll keep submitting updates. Even though people don't really care unless they just see an opportunity to bust my balls.
Submitted version 2.45.02 almost one year ago, did I see a single Youtube video featuring the new version? :cussing:
-
Even though people don't really care unless they just see an opportunity to bust my balls.
Aye, I'm currently saving up corrupted sims with different bugs. I plan to post a dozen of them all at once so I can show how much you suck. ;)
-
No problem, I am taking you seriously on this. Please post them. I bet 75% of them where old bugs that I had nothing to do with. :wacko:
you suck.
That's nothing new, almost drowned for the same reason... :wacko:
-
I only had something weird going on with a sim a few days ago. I can't reproduce it, but I'll let you know when I discover how. The issue was that bots were ignoring some veggies, going through them as if they don't exist.
So, no I don't actually have a dozen bugs ready for posting. :P
-
Sorry Peter, just having a bad day I guess.
-
[Bump]
I will have to stay away for a while :/ Until I convince my mom that the things I do here are worth while. I think her brain wash is beginning to kick in :/
I hope it will not take too long.
-
[Bump]
Resuming work this Saturday.
New people are still out voted 6 to 2, but thanks for participation. ;)
-
[Bump]
I am still looking on making this thing optional.
1.) Basically I totally ran out of control space in OptionsForm.frm
2.) The only way I can shove this into PhysicsOptions.frm is if I tabulate that entire form. That's not really fun. Also, why bother tabulating a whole set of controls just to add two or three radio buttons?
3.) It is different, but I don't think it is worse, I think when people get used to it, it gives more variation to the Pond mode simulations.