Opening the August thread, I’m very close to submit my AI based game creator to the Gemini API developer competition.
In the last update, the AI can suggest an “attack strategy” for the opponent in a fighting game. My vision is quite simple - the AI will assemble and apply some strategy for the opponent according to the latest statistics, then every few seconds new statistics will be sent to the AI and new strategy will be created and applied so the first evolution is quite “damme” but it becomes smarter in later iterations and those strategy evolutions can be saved and used in future games mainly for saving AI services costs.
I’m proud to submit this project and I think that It’s also a good PR for JME and may attract new developers to try using it.
In the below video I used Suno AI for creating the background music. I encourage you to try it. It’s amazing…
Also, this opened a lot of possibilities – e.g changing the way the texture is drawn of decals (to prevent smearing when drawing a decal close-to-parallel to a surface)
You can wishlist on Steam. I’ll very likely release it next week. But available at gamejolt already.
@pspeed
I’m using regular lighting. I’m not jumping in the PBR bandwagon for a long time to come. But I did learn all the fundamentals by now.
I always have the normal/bump and specular textures at the ready. I just don’t always use them while I’m developing. I like to keep it minimal so I can iterate during development.
Made some progress this month - finished up some skill effects and added sounds, worked on some items and tweaked the editor.
There’s still so much more that I need to do before I can make a complete game.
I’m a big fan of working on the Engine. However, Paul is here for his own reasons, not do whatever you or I think would be best.
Perhaps he practices 3-D modeling because he enjoys it. Or because he’s trying to hone his skills. Or share what he’s learned with those who follow him. In that case, would you humbly suggest that he shouldn’t do what he enjoys? Or that he shouldn’t improve his skills? Or share what he’s learned? Why??
I’m sure Pixelapp is just trying to help Paul It is good to know that there are quality assets readily available for those of us who need and want them.
Paul has a unique look in his game, perhaps that is one reason to craft his own too.
Yes, that’s right. I’m looking to have a consistent and unique look.
I also have specific requirements for the models to support later game features like skinning, etc…
Also, that prebuilt asset is very specifically a pheasant and would not look good as a cardinal (which is what my first bird will be)… and would not be easily reskinnable as other birds. I will have game-selected reskins almost immediately after getting it into the game. At least male+female cardinals, robins, and probably a blue jay to start… if I do like I did with the butterflies then I will have color-tintable gray scale skins to have infinite “rare colored” birds to collect.
(Mine also only has ~400 triangles and just maybe a dozen or so bones… which will help when I want to have 20 of them around.)
I wish the Forester plugin was up to date.
I tried running it but without documentation it’s not worth the hassle.
Hopefully I have time soon to create a very simple vegetation loading library.
These are the actual models I’m going to use for vegetation. The previous model was just a placeholder.
Hey @pspeed what are your approaches to create normalmaps from textures? I use AwesomeBump, which works decently but sometimes needs a few tweaks. Did you ever show how you do it in any of your videos? Or perhaps you’ll show it in the next video, since you’re going to talk about UVs?
hope to see some gun action soon 
.
(Thank you)
It is good to know that there are quality assets readily available for those of us who need and want them.
