Sunday, February 18, 2024

Not Out for the Count Yet

I haven't really tried to mess with RAS since the debacle over them getting stupid for no apparent reason, mostly out of lack of will and partially out of lack of knowledge. Considering the original 2017s also had a similar problem, my best guesses are either the CFF Fullness Lobe or the wide variety of instincts RAS possesses messing with their decision making process somehow. Or maybe it's just one of my regularly used agents not being as creature-friendly as I thought it was. 

But after a while of mulling it over, I don't think I'm willing to give up Creatures entirely over it. There's other projects I could possibly pursue in the meantime. 

I still have my attempts at Gizmo-based and CFE-based RAS genomes lying around. If the issue lies with the 2017 base, there's two other options that might have a higher chance of success right there. In particular, there's the idea I had right at the beginning, which was to start with a CFE base and fix as many issues as I could while still maintaining compatibility. Though I'd be focusing on my personal bugbears, which are mostly that base CFEs eat too much, breed too much, and can suffer from irritating lactate-based mutations killing their muscles. 

I've had a particular thought about remaking the Enhanced ChiChi genome with a CFE base, implementing everything they tried to do in a manner that actually works. I've especially thought about doing this with a CFE Grendel base instead of a ChiChi base, because in all honesty I'm really tired of working with norns and want to start messing around with my preferred species. 

And if these Enhanced CFE Grendels work out, I'll have a nice base for future grendel breeds. Then maybe I'll think about taking another crack at base norn and ettin genomes. 

Stay tuned, folks.

Tuesday, February 13, 2024

A New Base Genome: Discouragement

 Right when I thought I was ready to go ahead with making the rest of the RAS genomes, I get a testing report stating something to the effect that the latest test release has problems feeding themselves (more specifically, that they have a tendency to go to the brink, but starvation deaths are rare). 

So I figured that was likely an issue with RAS's ability to learn from the Disappointment stim. Considering I've seen something to that effect even in my own testing, I figured it was probably best to revert that change. 

Not long after that, I hit on the idea of privately journaling a wolfling run world, inspired by an old post of Aiko's. I figured it'd be a nice break between finishing the RAS ChiChi genome and moving onto the next RAS project. For the run, I decided to make a quick, RAS-based Grendel genome; in the process of making it, I found that the eat stims in 2017-based genomes have the secondary hunger stim silenced. I figured that might also have something to do with the earlier observation, and changed that for this grendel genome. 

With the world set up, I hatched a batch of grendels and went...only for the population to forget how to eat and start starving to death en masse before making it past gen 1, in a manner much more akin to my observations about vanilla creatures than anything CFE-based or beyond. This was in C1toDS and I had just ripped out the original elevators and replaced them with elevines, so I figured maybe that had something to do with it.

So I restart the run in a mostly-vanilla DS world, with some food sources added to the hub and workshop. Once again, the mostly first-gen population forgot how to eat and started starving to death; specifically, their hunger for fat was consistently getting maxed out, according to the medical monitor. So it was definitely something wrong with the genome. 

My number 1 goal with RAS was to create creatures who can take care of themselves pretty well, so that they're interesting to watch. If they're having this much trouble surviving in a mostly vanilla DS world, something is very, VERY wrong. 

But...I'm at an impasse here. I have no idea what's causing this, nor any idea on how to fix it. I do know it was a problem extending back to the original 2017 genome; it was why they only had one instinct relating to each type of hunger, and probably why the aforementioned eat stim change was made as well. So was it just because my base was flawed?

I don't know. But what I do know that this has basically shot any motivation I had to keep playing Creatures in the foot, at least for right now. It feels like every time I just want to play the game to relax, this kind of s-argh- happens and just makes me resent the game (this isn't even the first time I've had mass stupefying happening to CFE-based creatures).

So...this is to say that progress on RAS is being halted for the time being. I just don't have the knowledge necessary to get past this brick wall, nor the patience to keep bashing my head into it. 

And as much as I'd like to keep doing things for Creatures in the meantime, chances are that I'm not and this blog's going to fall back into silence once more, unless I have some kind of epiphany.

Wednesday, January 31, 2024

A New Base Genome: Tentative Conversion Plans

 I've been taking somewhat of a break from Creatures in the days since finishing the RAS ChiChi genome. I guess my brain's tired of thinking of dev and wants to do other things for a while. But nevertheless I'm still having thoughts about where to go from here and what to tackle next. 

In my breakdown of the final RAS ChiChi genome, I mentioned wanting to do a full set of RAS conversions for the official breeds. In this time of indecisiveness, I figured I might as well talk about what my plans would be for each breed. Maybe in the process I'll figure out where to go from here. 

First and foremost: when I first spoke about what my plans for RAS ettins were, I mentioned I'd be basing them on the RAS ChiChi genome as opposed to porting all the RAS changes onto the 2017 Ettin genome. Well, it's not just the ettins I'd do that for; it's my plan for all of the official breeds. It might be a bit more work, but in this way I can design the RAS versions as I see fit, avoid some of the pitfalls the original genomes made, and perhaps give them their own unique flair. 

 For the former mall breeds in particular (Bondi Norns, Hardman Norns, Treehugger Norns, Toxic Norns, and Banshee Grendels), I'd also put special focus on replicating their behavior from their respective chapters in the DS story and making sure they can make use out of all the agents in their respective packs. 

As for each breed individually....

Saturday, January 27, 2024

Genetic Mysteries: The Case of GrM0262

 

Ten years ago (why am I so old?), I was running a series of grendel wolfling runs with the goal of reaching generation 1000, after which I'd take the final result and use it as a base genome for a new genetic breed. I still have that goal of reaching generation 1000 on my own someday, but alas generation 300 was as far as I got. Nevertheless, I shared three packs of grendels, one for every 100th generation milestone. 

On the approach to generation 300, the run was suddenly stricken with an epidemic of sliders (creatures who lack pose genes and thus can't move; if you try leading them with the hand they "slide" along the ground, hence the name) and other such maladies. And among the individuals in the run was GrM0262 (a clinical name given to him by Kezune's Wolfing Autonamer, which I used to track generations in the time before I made my Wolfling Nametagger). 

All of the genomes I used to start the run and continued to add new generation 1 creatures from were CFE based and included all of their immortality preventing edits. Yet somehow, GrM0262 was immortal; he was going on 17 hours of age for genomes that died under the five hour mark. I naturally exported him as soon as I realized and set about trying to figure out what went wrong.

At the time, I never did, and the best anyone else could figure out was that he was extremely vulnerable to injury and basically all of his organs were dead. Ten years later, I came to a realization about what the cause could've been...and not only was it figured out, but it lead to the discovery of yet another oversight with the vanilla genomes.

Thursday, January 25, 2024

A New Base Genome: The End is Never the End is Never the End is Never the End....

 After I posted the breakdown of the supposedly-final version of the RAS ChiChi genome, Dragoler offered up a criticism of it. Specifically, she pointed out that the way RAS consumed muscle tissue when breeding probably wouldn't be all that effective in adding a cost to creature breeding without also adding in a gene that prevented breeding outright in creatures who weren't taking care of themselves. 

As it was, RAS creatures can just keep kisspopping and laying eggs even when their muscle tissue is at 0, and not actually suffer all that much for it considering amino acid (and thus muscle tissue) isn't actually used for all that much by default; mostly the production of prostaglandin and also as an emergency backup should glucose levels be low but amino acid levels be higher (this setting aside the fact RAS also uses prostaglandin to heal wounds, which in my experience is a more frequent occurrence than just organ injury). 

Now, the reason I didn't stop RAS creatures with low muscle tissue from breeding to begin with was partially out of concerns over compatibility, but also partially because the levels of muscle tissue in RAS are much less stable than they are in TWBs. I monitored the levels of a group of first-gen RAS ChiChis with the Biochemistry Kit to double-check this, and sure enough they commonly had levels of muscle tissue below the levels that would prevent breeding in TWBs.