Darwinbots Forum
Code center => Suggestions => Topic started by: Botsareus on September 09, 2014, 03:14:35 PM
-
Just putting this question out there for people do ponder (while I mix my first Album) did anyone other than Panda want chloroplasts? And I am not talking about metabolism, many wanted a good metabolism system.
-
Panda, I am not saying it is a bad thing, it thought me a real life lesson at the least. ;)
-
Sure, chloroplasts sound good. Just really hard to properly implement as it got a huge effect on DB2.
I didn't like that the old system was removed while the new was unproven. :sleep:
-
I frankly don't care. Let the plants be plants :D Or perhaps we could include an optional carbon dioxide and oxygen relationship along with them. Just a small value. Depending upon their activity, a bot will raise the overall carbon dioxide level for the simulation, but lower the simulation's oxygen level. Depending upon the amount of sunlight they're getting, plants will lower the carbon dioxide level and raise the oxygen level for the simulation. As a result, the plants ALSO depend upon animal bots to survive, instead of just depending upon the sun. I'm not asking for a gas simulation to accurately represent the carbon dioxide and oxygen. Just a simple number. Then again, it's just an idea.
-
Yes, probably it would have been best to start-up a new line of DB2 called "DB port to Chloroplast " or something like that. My narrow-mindedness got better on me there.
-
Spork, adding a lot of features on top of this little desaster will just make the disaster worse. That was my whole point with this specific thread. Chloroplasts made DB2 waaaaaaaaayyy to slow.
What got me was the premise "Make something sound interesting enough and people will want it."
Usually not a smart development strategy in the long run.
-
The only time it is a smart development strategy is when you want to kill a project very, very quickly and move on to a different project under a new company name.
-
On a final point, this only happens in the situations when the final product turns out to be a complete pease of siht.
-
Such as example here.
-
There are plenty of people selling promises. You can make a living out of it. Marketing, sales, politicians, Nigerian princes, recruiters.
Products often don't get sold on the basis of it being good, but based on the nice looking package.
-
What do you think is wrong with the current chloroplast system, botsareus?
-
Not much. Except I have yet to find use for it. It slows the program down considerably if you want to actually use it properly. And it was a pain to balance out so people can use it properly. And by properly I mean co-evolution.
-
I liked the chloroplast feature and found settings that worked well for me, but I didn't like the decaying of chloroplasts. Setting it to decay a lot slower might be an option that you suggested.
The only other thing I didn't like was that alzheimer always created chloroplasts.
-
Agree on both points, it is going to be 0.05 in next version.
I will make sure it is not allowed in Alzheimer.
-
That would be awesome!
-
hmm strange can't seem to post very well. anyway i like it and am getting some interesting results with it. sim is attached...
i think attachments are broke because i can't attach any thing...
ok so i can attach the sets but not the sim. it's only 18.6 MB
-
Done deal
-
When it was initially suggested there was at least 4 of us who discussed it.
-
And not a single one of us knew the risks, I know.
-
I think having chloroplasts was a great idea, it always bothered me that veggies had to be assigned to be veggies. In real world, the border between plants and non-plants is quite fuzzy.
But the current implementation is driving me crazy because the formulas behind it are not in any way logical.
I understand the challenge that bots with chloroplasts get their energy for free, but that just means that having chloroplasts should come at a cost. They should require quite a bit of energy to maintain, they should make bots become large, etc. If you do it right, then some bots will evolve to have chloroplasts, while others will evolve to eat those that have chloroplasts, and others will switch back and forth depending on conditions.
-
Agree :)
-
Shvarz's they pretty much do that now.
EDIT: I think the way that Botsareus has explained it has made it complicated. He added an "area limiter" as well so there's only a certain amount of energy available in a simulation. Other than that, the features you are talking about (and are all implemented) were the ones that were suggested last year, or the year before, in the first place.
-
I think chloroplasts need to be more simple and crippled more. Every sim seems to jump against 'veggies' with chloroplast limit. These limits should rarely be hit.
Would vision be cheaper to compute if you reduce the line of sight?(not removing it completely)
If so, reducing vision gradually would be a win/win. 'veggies' would be lesss cpu hungry and less powerfull.
Would make them less good at hunting because they can't see prey. And make they better suited to hunt for.
Also, does size automatically increase shooting power? (I expspect not, but wondered)
-
hmmm.... blind vegys? What will be there motivation to evolve for stuff they can not see?
-
Chloroplasts are easy to evolve remember, they are very, very hard to de-evolve.
-
hmmm.... blind vegys? What will be there motivation to evolve for stuff they can not see?
Being able to see. :P
Less chloroplasts, better able to run away or even hunt. That's a incentive to grow less chloroplasts.
-
Very good reasoning. But, I'll wait for more info. from Shvarz on this topic.
-
Panda, if they do that and it's still better for bots to just sit and feed on veggies, then they don't do that enough. Jack up the penalties until running around and feeding on veggies becomes a competitive strategy.
-
Once again I like to point out that these factors should become important with enough robots on the screen, to weed out all the losers.
-
Panda, if they do that and it's still better for bots to just sit and feed on veggies, then they don't do that enough. Jack up the penalties until running around and feeding on veggies becomes a competitive strategy.
Which bit are you talking about?
-
Panda, all of them. I guess bot size has a limit, so maybe not that as much (although you could still crank up the rate at which bot size grows with chloroplasts). You can crank up the rate at which chloroplasts contribute to waste. And you can increase the actual energy cost per cycle of having chloroplasts. You can also tone down the amount of energy that a bot gets every cycle via chloroplasts.
My point is: if the problem with having simple equations for chloroplasts is that bots always evolve chloroplasts and then sit and feed on energy (which is what Bots told me), then there are ways to deal with that via other balancing functions.
-
The plan was originally that all of those would occur and the area limiting part wasn't going to be implemented but that was Bots decision to include. If you have a look at http://forum.darwinbots.com/index.php/topic,3487.0.html (http://forum.darwinbots.com/index.php/topic,3487.0.html). I implemented that at one point; however, I had problems releasing it. I think this is the sort of thing you're talking about more, aren't you?
-
Yes, exactly what is discussed there. I think Nums' suggestion sums it up nicely:
http://forum.darwinbots.com/index.php/topic,3487.msg1383180.html#msg1383180
-
Adding radios to robots with more chloroplasts is not a bad idea. But we should do so in a way not to hit the limit on size.
Never cared for the body<---->chloroplasts slider. How should the program auto adjust for something like that?
Overall, no one actually figured out how much energy a robot gains from using chloroplasts.
As for costs, I always prefer costs a user can set. That is what I did.
Finally, I actually am willing to send a version to Shvarz with no 'correction for robot population by area' for further testing. What I was attempting to avoid is the picy attached. But it seems it happened anyway. All my robots in internet mode enabled survival evolution now use chloroplasts.
edit: Picy not attached because I can not RDP to windows7 from XP.
-
What specifically do the current chloroplasts do?
chloroplast cost nrg to build
chloroplasts give nrg(based on the area formula)
chloroplast decay(fixed amount?)
Make a bot heavier and bigger
Anything else?
-
Make a bot heavier and bigger
i know they make a bot heavier (very obvious) but i don't think they get bigger do they? i've never seen any noticeable size gain even between 0 and 32000 chlr.
-
I have enough to make improvements to chloroplasts, it will also simplify the program. Update soon.
-
Hopefully there'll be less killer plants of doom. I've already made 3.. all of which are dangerous
-
Word up :)
-
I've always avoided ageing costs, I haven't really liked them. :/
Here is what I currently have:
Robots get more radii the more chloroplasts they have.
Robots with less chloroplasts lose chloroplasts faster then robots with more chloroplasts.
A picture of results is attached. (Pond-mode, Weather, and Tides on)
Should I add aging costs to this or not?
Any ideas?
Panda are you cool with that? Or, do you want me to disable point mutations as well?
-
I am going to attempt to crank up point mutations a bit first.
All aging costs I have just attempted where completely op.
-
Robots get more radii the more chloroplasts they have.
What would you say about doing it the other way around? (unless it's too late for more suggestions here) You could make the amount of chloroplasts limited to bot size. A maximum of 500 chlr per 1000 body or something like that. So the bot would have to convert lots of nrg to body to have room to store chlr.
Should I add aging costs to this or not?
I think all costs should be affected by the ageing cost setting, if they aren't already.
-
A little late there because I already started the league :( Looking more bio accurate and less crap to configure.
I may end up playing with age costs...
-
I have an idea, chloroplast efficiency gets lower with age...
-
I don't like it. I think aging costs should be completely independent of other costs, like it is now.
-
Yea, I went with chloroplasts efficiency gets lower with age, to at least get them to reproduce instead of just sitting there.
-
Testlund, bot radios is now based on body + chloroplasts with a cap. But I may try your idea anyway.
-
New idea, chloroplast efficiency remains the same, but robot losses energy mod(age * chloroplasts) kinda like a forced upkeep cost.
-
I think it is as good as it is going to get. Have no time to play with mutation rates.
-
Well finally stabilized the mutation rates. Chloroplasts are evolving... not really, they still only use chloroplasts.
-
Any good news? Well they continue to mutate and reproduce!