Veggie Energy

[Shen]

Suggestion for improvement of the way veggies get energy.

I would like to see a proportional veggie feeding system similar to the old system. The reason I say this is because it would give a lot more control over exactly how much energy goes into the system. Im having a bit of trouble stablising my bot populations because fixed feeding doesnt give as much flexability as the old way, even though its A LOT easier to understand. Something along the lines of this...

A value, say 500, is specified by the user. That value is then distributed amoungst the veggies directly. So if I have 10 veggies they each get 50 nrg per cycle, if I have 100 veggies they get 5 nrg.

I never did understand the old feeding method completly but I knew that more bots/veggies meant less energy, and thats what Id like to see. Maybe we could use both systems? Like the way shot damage has the option of fixed or proportional.

[Numsgil]

The point of stabilization should just be at a different place than before.  I notice that alot of the time the stabilization point is with alot of vegs and fewer robots.  Which is actually not necessarily a bad thing.

Still, if people would like a constant energy input level I guess it could be set up.  But I'm still of the opinion that you'll get more realistic simulations in the new system.

[Shen]

My problem is with the growth/death cycles of alga. Im currently running a size 12 sim with 1000/200 max/min veggies and 3 nrg/cycle and Im having several problems.

When the alga grows to 1000 thats essentially 3000 nrg per cycle which is far to much. But when the alga is 200 thats only 600 per cycle. Its the opposite of what is needed to stablise the population. I need maximum growth of veggies when theres few alga to boost it back to normal levels, and minimum growth when theres to much alga about. What I get in reality is max growth when minimum is needed and minimum when max is needed.

Also IMNSHO I think your wrong about the realism of the fixed system. In real life if there is overcrowding of a population growth is slowed not accelerated.

Whatever the base source of energy veggies are supposed to draw from, be it soil nutients or sunlight or whatever, it might be a renewable source but its still delivered at a finite rate. Think rainforest canopy. You have loads of trees competing for sunlight, the ones under the canopy dont get sunlight and therefore either stay small or die off. But when a hole appears in the canopy there is a spurt of growth from all nearby trees to fill the gap and claim the sunlight.

[shvarz]

Maybe I should reply to Shen, as I pushed for the current system for a long time.

This system is much better than the old system.  The problem you are having stems from a wrong assumption that you need a stable population of predator.  But this is wrong - population of predators almost always goes in cycles: predators grow large in numbers -> wipe out a lot of prey -> their numbers go down due to scarcity of food -> number of prey goes up.  This is normal.  And this is what I am getting right now, also on size 12 sim.  

With the old system that was impossible to achieve.  The most stable situation was when predators became so numerous that they covered most of the field, leading to either constant re-population or to small numbers of veggies receiving huge amount of energy and feeding the whole crowd.

Your analogy to rainforest ecology is not valid, because you don't get even close to 100% coverage of all area of a sim.  Take any other system and you'll see that new system is more accurate.

I don't mind having the old system (or something like it) as an option though.  It might be useful in some simulations.

[Numsgil]

Shen, the field, in a way, is getting a constant amount of energy per cycle.  It's just whether there are enough vegs to take advantage of it or not.

If you pretend the simulation is a top down look at a forest, and the vegs are trees, then I think the new system models real situations very well.  Most simulations I run end up with ~10 times as many vegs as animals, which is actually pretty close to reality.

[Carlo]

I agree with Shvarz and Nums on the fact that the new system is more realistic than the old one. The old one was not designed to be realistic, but essentially for:

1) having stable simulations, with a good bots/vegs ratio (the more vegs, the slower the simulations) by:
2) avoiding starvation by boosting up vegs growth when to few were in the sim
3) avoiding overgrowth of population cutting the energy input when it as too much.

So the idea was something like: the simulation is like a universe: inside it, the total energy is constant, it can only move from vegs to bots and vice versa. This is not realistic, since Earth is not a closed system, as a constant amount of energy comes from the Sun.

If you manage to get such an open system stable and "playable", then it's ok.

[shvarz]

For me, it is actually easier to get this system stable and playable.  I've never been able to run sims for millions and millions of cycles and now I can.

[Shen]

Well the sim im running at the moment is filled with cannibots so the population is usually ~50. The problems happens when usually you would get a predator boom cannibots dont get one because as there numbers increase they also eat each other and are kind of self limiting at around ~150. This causes the veggies to max out at 32K nrg and the bots simply arnt able to eat fast enough because they are limited to low numbers by themselves.

Lowering the max veggies just decreases the total bot population below sustainable levels, and the feedrate is already at 1. Any ideas on increasing my population shvarz?

[PurpleYouko]

What size sim are you running Shen?

Try really big!

[shvarz]

I am assuming that you are running size 12 sim, otherwise you can just increase the size of the sim.  It certainly sounds like your system has too much energy.  You need to increase the costs somehow, so that energy is flowing out of the system faster.  There are many ways to do that.  Just pick one that does not interfere with what you are trying to achieve:
1. Make veggies spend energy (make them running or making shell or something).
2. Switch to kinetic-energy mode
3. Increase costs in "costs" panel
4. Increase friction
5. Introduce day/night cycles
6. Reduce moving factor

In my current sim I see the appearance of cannibots, but they die out fairly quickly.

[Shen]

Ive been slowly increasing the friction but my bots got rid of the condition on the move gene fairly early on so it will have to be gradual. The day/night cycle is an idea though hadn't thought of that. I wonder if the population cycles will end up matching the day/night ones. Should be interesting. Oh yes its a size 12 field.

Cheers! :D

Edit: Hmm it appears that the day/night cycle doesnt work, at least not that I can see. I might just not understand how it works but the energy input seems to be the same for both day and night. Hows it supposed to work? Or is it broken?

[PurpleYouko]

Edit: Hmm it appears that the day/night cycle doesnt work, at least not that I can see. I might just not understand how it works but the energy input seems to be the same for both day and night. Hows it supposed to work? Or is it broken?

Yes it is broken, or more to the point it only works in pondmode right now.

Trouble is that the last few versions (since the new veggie feeding rate) up till 2.36.3PY (in which I fixed pond mode), pondmode is broken too.

It is all fixed in my version now.

[ashton15]

If you wanted this as an option could you not get the already existing advance to dawn/dusk option and create a grduated energy slider which is in sync with the total veggie population so that the same ammount of energy is always given, should just take a bit of maths and some simple code for fine-tuning the gradient

[ikke]

If you wanted this as an option could you not get the already existing advance to dawn/dusk option and create a grduated energy slider which is in sync with the total veggie population so that the same ammount of energy is always given, should just take a bit of maths and some simple code for fine-tuning the gradient

using nrg per body (squared) in stead of per veggie is the major improvement. you can also cap the total energy in the sim, so afaik this is already implemented

[SlyStalker]

i think the veggies should have an internal chlorophyll value which dictates how much energy they get from the sim. this is what I propose:
Veggies should have a piece of code saying that they are actually veggies instead of a check-box that you tick at the start of the sim.
Veggies should be able to produce chlorophyll, which generates energy for the veggy.
Chlorophyll should break down over time, like slime and poison. This forces the veggies to constantly make more chlorophyll.
Chlorophyll can be stolen by other bots. The other bots then have access to a free energy supply. But of course, the chlorophyll will still break down.
Chlorophyll energy production is controlled by a 'sunlight level' cariable within the sim.