I hear that, right now, genuses diverge when a type of cell part is removed or added. If this is the only cause of genus change, then clearly, new types of cell part are getting added and old ones removed with disturbing frequency. Even disregarding this, cells, at least small cells, seem to prefer adding new food sources. In the volcanic vents, I’ll often see the first three divergences being specifically to add thermosynthase, rusticyanin, and sulfur metabolism.
If I had to guess, right now, cells aren’t really pressured to specialise for any particular niche. I don’t know enough about auto-evo to give a firm guess as to why, but some things that seem plausible to me are:
Patches generally go through harsh cycles of mass speciation and mass extinction, especially when the player is a eukaryote. Oftentimes, if I sit in a patch for a few days as the large hunter I prefer to play as, very soon there will be no other cells remaining. This could mean that the lifespan of species is too short for them to really adapt for the conditions of the patch, although when I look at my tree of life, there aren’t many red nodes, proportionally.
I recall an older version of auto-evo having a serious problem with producing omnitrophs. I don’t know exactly why this was, but I think it was something to do with it not weighing competition correctly. It’s possible competition still isn’t strong enough to make cells specialise.
Patches aren’t different enough to incentivise much competition. A cell that works in one patch will work in many others, and while these cells will be constantly diverging, masking the effect, they won’t be dissimilar under the hood. As a result, if a generalist cell is wiped out, another will soon arrive to take its place.
I think the lack of specialisation is something that will need to be troubleshot before 1.0.0. At the very least, I suspect having longer-lasting and specialised cells will result in the food trees being much more sane. (Right now, they aren’t anywhere near stable enough to arrange into good structures.)
Now, I can’t cite large-cell instances of this, because large cells show their parts at a smaller magnification and all cells are effectively interchangeable. That being said, I constantly get recommended to add rusticyanin as a glucose-only cell; if auto-evo thinks eating iron is always worthwhile for the player, that implies a flaw with the generation of the AI, since they should never disobey auto-evo’s suggestions.
Where did you hear this? I was going to say that you are wrong, but I luckily checked the code and sure enough I had forgotten that the new miche changes also added this new function for determining when to generate a new genus name:
So I’m wondering a bit how you found this out…
How would you change this? Because with the new miche-based auto-evo a species only gets energy if it captures a miche. Meaning that a species that doesn’t manage to outcompete the competition for a miche gets 0 simulation energy and thus no population (assuming it doesn’t manage to capture any of the other available miches). So if a species loses by even a fraction of a fitness point for a miche, it gets nothing.
3 Likes
aah31415
(The maker of SitF, Radiostrocity, The Lifenote and TGBing; The Second Ascended...; And just maybe a security warning come alive...?)
3
The suggestion system seems to have many problems at the present. I and some other people also reported this system suggesting myofibrils prior to entering multicellular.
Fundamentally the problem is that that makes competition much less harsh, and that would make the name much less accurate (as they are meant to be a niche + extra stuff and biologically speaking a niche should never have more than one species in it as it doesn’t qualify as a niche then).
Also I’m totally the wrong person to try to poke about auto-evo changes as that’s not my area really.
4 Likes
aah31415
(The maker of SitF, Radiostrocity, The Lifenote and TGBing; The Second Ascended...; And just maybe a security warning come alive...?)
7
I believe 0.6.2. Or 0.6.4 but I don’t think that’s a version.
In my head, it’s stored as ‘two versions ago’, and was, until recently, the latest Mac version. Technically still is since the current Mac version doesn’t work.
1 Like
aah31415
(The maker of SitF, Radiostrocity, The Lifenote and TGBing; The Second Ascended...; And just maybe a security warning come alive...?)
12
That was quite some time ago, so the algorithms may’ve been different from the last non-miche autoevo.
Would it make sense to implement a couple Omnitroph specific Miches separate from single compound specialist Miches? Or would that be to unrealistic or problematic?
1 Like
aah31415
(The maker of SitF, Radiostrocity, The Lifenote and TGBing; The Second Ascended...; And just maybe a security warning come alive...?)
14
The developers probably don’t have enough time left to implement that and rewire autoevo to work correctly again.
Obvious;y, but that doesn’t answer the question. I remember reading on one the Dev topics that they wanted to have more specific niches in macroscopic, something about growing on a pink rock instead of another colored rock. I think in general a couple of additional mitches would add some variety to microscopic, and @thriveuser1454 was complaining about lack of omnivores.
2 Likes
aah31415
(The maker of SitF, Radiostrocity, The Lifenote and TGBing; The Second Ascended...; And just maybe a security warning come alive...?)
16
That would require a consultation with Thim, which is the master behind the miche system. He’d know what needs to be done to get such miches working.
I wasn’t complaining about a lack of omnivores, I was complaining about there being too many omnivores. There’s not enough specialisation, as is indicated by the title of the post.
@aah31415 that feels like a completely unnecessary reply…
I’ve not said anything yet about you being the last poster in literally every thread for multiple days, but I’m starting to feel like you should only reply if you have something to add, not always reply just because someone else has said something.