The Random Run

Well, LNA has been running fairly nicely for a week now. In an attempt to better the performance I reformatted the hard drive and while I don't know if I see a notable difference in the stream, the computer itself is much easier to work with now.

Progress has been slow but steady on BattleStance. I have added in the overlays, CAs, CA linkers, and have worked a little on the sprites for the fruit meant to grow on the tree. In experimenting with norns in the room though, I am worried that I may have made the lifts too large-- in some of the smaller rooms, they seem to grab every norn into it, willing or not-- and because they move so fast, if one norn is insistant on pulling the lift and the other on pushing, the lifts end up bouncing up and down for quite a time. As much of a pain as it was to do those lifts I am seriously considering reworking them.

There's been a lot of chatter over at Creaturetopia about CCSF '09. A couple of polls finally set this year's festival date for November 1st-14th-- partly in honor of the ten-year anniversary of Creatures 3, which was released on November 1st, 1999. Now there's a bit of dicussion about the potential theme. It's nice to see so much enthusiasm. I have high hopes for this year.

With the required updates out of the way, I'd like to make mention of a new sort of wolfling run I'm trying out. I'm calling it a "random run" and essentially beginning it with this block of code injected eight times:

new: simp 3 4 1 "eggs" 8 0 2000

pose 0
attr 195
bhvr 32
elas 10
fric 100
aero 10
accg 4
perm 60
gene load targ 1 "*"
setv ov01 0
tick 900
mvto 443 9167
anim [0 1 2 3]

It is a generic egg-injection script-- but the bolded line ensures that a completely random genome is injected into the egg. I have recently downloaded just a ton of new breeds and thought this would be an interesting way to test drive them.

It looks like the initial line-up consists of some chichis, an astro, a fallow, a harlequin, a couple plant norns, and a weird purple bengal. Not as many of the strange new breeds as I initially hoped, but randomness will have to take its course for the sake of the run rules.

I intend on letting the run go on until the population is mostly homogenous. Nothing special is being done to accomodate those norns with special needs-- the resulting genetic sample needs to be able to survive and breed normally in a typical DS world. I've rigged up the egg finder with the timer in the workshop and killed all teleporters as I do for all my runs, but other than than the world is virtually untouched. Nature will take its course now.

On a somewhat related note, I would like to do a few experiments on norn adaptation in different metarooms. It would take a while, and the results may or may not produce anything, but it might be interesting to try.

I would run several feral runs... perhaps five for each room. Starting with normal chichis, and letting them run until generation 100 or so, then comparing genetic samples to see how much they differ from each other and the base chichi genome. In theory, the norns within the same run would differ very little from each other, but differ some from the norns in different runs within the same metaroom, just due to the randomness of evolution. But theoretically, the norns from all runs done in one metaroom would differ quite noticabley from norns raised in another metaroom, because they have had to adapt to different enviroments.

I'm not sure if 100 generations is really enough for the genome to settle though. And I'm not even sure if the difference in metaroom enviroments would cause the genome differences to be greater than the sheer randomness of mutations that take place in separate colonies. But eh... that's what experiments are for, right? To prove or dispprove theories.

But I'll file that away in my nice little idea folder for now. As if I don't have enough projects to keep me busy!

Darn Genetics.

Well, around generation 300+ in ColorRush, I got tired of seeing a bunch of norns bumming around in the workshop, somehow stupidly happy and alive despite doing hardly anything. They apparently had no desire whatsoever to push the door-- they all liked to crowd around it, but I'm pretty sure no one left that workshop for a good 50 or so generations.

I finally compared the genome of one of the random norns to the orginal Siamese CFE genome and wasn't really surprised to find a ton of odd mutations. There were over 300 genetic differences--more than a third of the genome had changed.

For one, his Hunger For Protein drive was associated with Pain. So I assume if you beat it senseless it would start complaining of hunger, but I'm not about to try it. Furthermore, his Boredom drive was associated with fear. That made me laugh a bit; I can image a norn facing a huge monster only to yawn and start complaining of boredom.

Organ repair rate for a few organs was affected by EDTA instead of Prostaglandin. There was a male-specific gene regarding how to deal with pregnancy. The genes to raise and lower sex drive in certain siutations had mutated so they were ineffective. Hunger for protein backup dissolved into nothing while Hunger for fat backup became coldness, but that didn't matter because hunger for fat would become hunger for carb backup if the need arose.

Hearing creatures speak actually made it lonely instead of crowded. Eating critters apparently relieved pain. Pushing pests relieved boredom.

There were a ton of insiginifcant mutations, just little changes in some of the values so some reactions took place a bit faster or slower. There were also a lot of genes I just didn't really understand because I'm still far from being a genetics expert. But it was an interesting experiment if nothing else.

I'm not entirely sure what to do next. But I probably should wait until after I'm finished with my senior project to worry about that.

If only evolution actually worked like this.

On the way to the current generation of 210 in ColorRush, several things have happened.

For one, the siamese norns seem to have managed to out-evolve the IQ test. That is, they don't seem to have to eat to survive long enough to breed anymore, and many of them are perfectly content to sit around on the top floor of the workshop (at least they can work the elevator) and complain about their hunger.

They're clearly not immortal though-- at least not by drowning >> I'm sure ERFN-types wouldn't be too happy to know that since I do want this to be a pure siamese run, I've been manually culling norns that have mutated bruin heads, grendel legs, etc., by dropping them in the nearest body of water.

On the other hand, I've been loving playing around with the Wolfling Monitor. To make sure my norns don't get off too easy, I have it set to check the run every 15 minutes and export norns that have high hunger/crowded/boredom drives. It's also been exporting norns older than 90 minutes just to keep the generations progressing, but as I slowly lower the drives threshold I am considering removing that criteria to allow the creatures that are fulfilling their drives to stick around and breed a bit more.

Ironically and contrary to the name, I haven't seen a lot of color mutations. There are a few here and there but the general population still has a default pigment.

On a somewhat amusing note, my folder of creatures exported from this run contains five thousand five hundered and thirty seven norns. 332mb of norns. I wonder if that's some kind of record?

I'm going to keep this run going a while, but I have some other ideas for future runs too. I'd like to try an ettin run, just for a change of pace... and maybe a bruin run in which I actually encourage sprite mutations until the entire norn looks nothing like a bruin.

After I've done several runs I would like to take a few norns from each run and put them in different mesos in DS docked and see which ones end up mixing and which ones die out and which ones breed in their own little group.

On a side note, I wish there were more Creatures blogs out there. I think it's interesting to follow the progress of these things. But the CC is tiny enough as it is.. you can't expect them all to be bloggers too :P

They say cats are intelligent

So I've started another wolfling run with purely siamese norns. They have always been one of my favorite breeds, and I wanted to see what I could make of them.

I did something a bit unusual with them this time around though-- I extended the IQ test. All eggs are still teleported to the the bottom of the workshop, but all the vendors in the norn meso have been kill hots'd and the majority of the run instead takes place in the Norngarden metaroom. This adds an extra step that the young ones have to go through to get into their little nornish paradise.

At generation 54 the success rate is quite high, however, and I'm beginning to think I need to make this run a little tougher on them. I have rediscovered one of my favorite wolfling agents and have been using it to automatically export norns over 1:30 of age, but even having been given such a short lifespan, the population is constantly at its max and the creatures are becoming quite crowded in that little garden.

I think I may set up Wolfling Monitor to take fear into account when exporting, since a lot of the norns seem to be abnomally fearful upon hatching...