Creatures Caves

GimmeCat wrote:
Tested both apps on Windows 7 x64, both load fine and did not crash during my brief playaround with them.

Thanks!

GimmeCat wrote:
I was able to import s16 sprites in the SpriteBuilder, but couldn't figure out what to import into the Map Editor. Both s16 and spr files from C1 and C2 complained about not having the correct headers.

The background files of C1/C2's headers indicate that all images are of size 150x144 and that there are 464/928 of them respectively. If either of these conditions is not met you get the header error.

GimmeCat wrote:
I had no idea this project existed, but I'm excited by your progress with it! What are some of the key things you aim to differentiate this updated version from the classic Creatures games?

To name a few:
[list]
[*] the end result should have a lighting engine
[*] metarooms each have their own coordinate system, so there's no chance of conflict
[*] creatures are diploid, eliminating the slider bug, and making it possible to use XY sex determination instead of genes set to male or female
[*] creatures size can change according to a receptor locus for smoother age transitions
[*] creature movement is controlled by the physics engine, so they should be able to jump and bounce and so on.
[*] objects can freely rotate and scale
[*] sprites are all designed for high DPI monitors
[*] CAs for heat and humidity should be using a real fluid dynamics approach, (this is harder than it sounds, I can't determine what exaclty makes any implementation not go through walls)
[*] I'm hoping for fluids more like rayman-origins style
[*] gravity is controled by an object called a 'physical zone', so you can reverse gravity in the metaroom if you want to
[*] all metarooms can call connect() with another metaroom and join together, this is also used to make ring worlds ala C1 (doesn't seem to work vertically yet)
[*] Neural networks now use a combo of C1/C3 and HTM, which should make it possible for norns to learn sequences of actions and think about what they're going to do.
[*] Chemical reactions now have 4 reactants and 4 products
[*] Chemical receptors can no longer bind to reaction speeds
[*] The VM uses jewelscript instead of CAOS
[*] The program is multithreaded, so it should run better on newer computers, where the current creatures has slowdown from so many agents. I'm thinking that in the original games the game would draw sprites by copying data from the sprite file to the frame buffer. This explains why the RAM requirement was so small compared to the size of the backgrounds, why the rotate/scale commands do nothing, why cameras are so slow, etc. This means the game was ultimately limited by hard drive speed, not CPU speed, so it wouldn't be running better on newer machines. I still remember how I could never run terra because it would make the game so laggy.
[*] The Creatures biochemistry and neural networks are stored in a seperate library i made (available free under the MIT liscence), so you should be able to replace it if someone makes a different implementation of neural networks that is better.
[*] I'd like to implement something like LoneShee's Netures ( http://sheeslostknowledge.blogspot.com/p/blog-page.html ), but I probably won't ever get to that =/
[/list]

C-Rex wrote:
This is awesome! May I ask why the animation for the norn runs a lot smoother than the regular creatures games?

Because the old games updated the limb rotations every .5 seconds, mine updates the animations 32 times a second and should render at 60fps so the hand doesn't feel laggy. Basically I'm using the pose genes as targets for where the limb should be every .5 seconds and adjusting the rotational velocity accordinginly to meet that target.

Doringo wrote:
I do find the animation rather weird, almost ragdoll-like. But maybe the weirdness will be worth it for more versatile content creation if this ever gets to a level where it can compete with the current creatures games.

I noticed that too, it's really hard to get life like movement, at this point I'm unsure of if it's because of the pose genes or a bad interpolation engine; but using the orignal pose genes puts it into the realm of the uncanny and offputting for sure.