The deal is you need a catylist. Either a flamable naturally occuring substance or an external synthetic function. Oxygen allows the sugars (or nrg) to be burned and used to produce kinetic. So to say that life might develop without a catalyst is to deny the exsistance of DB; or in other words, without a catalyst we've crweated a simulation not mimicing life.
Erm... Let's just say that life works with an energy gradient. (Early life didn't use oxygen at all, it used sulfer. You just need an electron donor and an electron receiver.) It performs some
action to
something, and is able to extract useful work from it. What that action and the something
are is a seperate question from what the life
does with that useful work. Darwinbots presently ignores the first part and just concentrates on the second. That is, it just models behavioral evolution. So the question is wether the program should also model that first part, the
how of things, and if you
need to simulate the how in order to say you're simulating life.
It seems that Evolution made a simple system whereby an organism of some compound base can have a toxic/poisonous element.
I think our susceptibility to harm from heavy (and not so heavy) metals has less to do with our chemical base and more to do with the abundance of the heavy metals. If arsenic was a common element at the beginning of life, the whole ATP citric acid cycle wouldn't have developed in the first place. Or it might have evolved to rely on arsenic. You especially should not treat it like...
In Carbon-Based forms oxygen works for burning the carbon, and our poison is Argon; with Nitrogen-based organisms the poison would be Selinium. If you use the elemental # designated by the Periodic Table, the equation for poisonous elemnts becomes wrapped around the elemental base-type of that organism. This has to do with chemical interactions and quantum mechanics.
...that. Very wrong. Well, it's wrong first in that you're doing that whole pseudo-science wrap-around-on-the-periodic-table wrong. I assume you're referring to that thing from the movie
Evolution, where they kill the nitrogen based organisms with selenium because
arsenic is poisonous to us (carbon based). It doesn't work like that. The periodic table is
not a chess board and it
does not follow that patterns on the periodic table will hold for other biochemistries. On top of that, I sincerely doubt that nitrogen will ever be the base for any biochemistry. It's just too boring for a base. Nitrogen
and phosphorus maybe... Silicon has been suggested, too... But for the most part our carbon chemistry is the "best" choice for our given temperature and pressure environment. See
wikipedia.
There is a way to combine both our ideas, but it means making a complex system for elemental interactions at a genetic level. I suppose any organism can appear from any inert gas because of its stability as an element, and ability to compound with others like it. Some thought is needed to decide how this will work, but for simplicity I say use a complete ratio conversion from Evolution, or create a defaulted system using either system.
You could indeed make an alife sim that worked by simulating simple chemistry. I've seen a few papers that explored that route. However remember the scale that biochemistry works on. It took billions of years for the oxygen catastrophe to slowly push biochemistry towards the more energetic oxygen. It's a neat phenomena, but it's a
planet wide phenomena. For any ecosystem that's even relatively plausible to simulate, you have something that isn't a closed system. What comes in does not exactly balance what goes out. Most attempts by humans to create a biosphere smaller than the earth (for example the project "Biosphere" in the desert), have failed. Gaian principles require an extremely large sample size to remain stable. In, say, the last billion years of evolution (which covers all multicellular life), the core biochemistry has not changed at all. And considering it's hard enough to create stability as it is, I don't think we should add something that could add one more dimension that needs to be balanced for a successful simulation.
All it gains is extra ways for things to
not work. A successful biochem sim would work exactly like the present simulation does. An unsuccessful sim would kill off all life in your simulation. Usually features should add something that increases the complexity of possible behaviors.
