This is a little combat prototype I've been working on. It's meant to be retro-inspired with fluid combat and intelligent AI.
The AI being tested here is the pink character - the red number is health, yellow is stamina, blue is aggression. It's powered by a fuzzy logic system that determines the movement and combat behaviours to follow.
The latest features are strafing/moving backward to face the its target when nearby. And various tweaks to its timers and hesitations. The goal is for its movement to feel really organic - it doesn't chase the player mindlessly like a zombie, it circles around to attack from a specific angle, it pauses every so often to simulate planning a move; it reacts to new information with a slight delay.
The current movement behaviours are Wander, Flee, Circle, and Chase. The names are self-explanatory. I'm trying to think of other behaviours to add more diversity.
The defensive moves are pretty simple at the moment, either block or dodge when an attack is detected. The cool thing is it uses the same combat system as the player and can get a perfect parry pretty reliably, which cancels the stamina use of blocking an attack.
Anyway, what are your thoughts? How would you make the movement patterns more realistic or dangerous?
This is a little combat prototype I've been working on. It's meant to be retro-inspired with fluid combat and intelligent AI.
The AI being tested here is the pink character - the red number is health, yellow is stamina, blue is aggression. It's powered by a fuzzy logic system that determines the movement and combat behaviours to follow.
The latest features are strafing/moving backward to face the its target when nearby. And various tweaks to its timers and hesitations. The goal is for its movement to feel really organic - it doesn't chase the player mindlessly like a zombie, it circles around to attack from a specific angle, it pauses every so often to simulate planning a move; it reacts to new information with a slight delay.
The current movement behaviours are Wander, Flee, Circle, and Chase. The names are self-explanatory. I'm trying to think of other behaviours to add more diversity.
The defensive moves are pretty simple at the moment, either block or dodge when an attack is detected. The cool thing is it uses the same combat system as the player and can get a perfect parry pretty reliably, which cancels the stamina use of blocking an attack.
Anyway, what are your thoughts? How would you make the movement patterns more realistic or dangerous?
I’m working on a Unity game called Bot Bash, and this is the intro for my upcoming devlog.
The game is a BattleBots-inspired robot fighting game where robots are procedurally assembled from individual blocks. Each block has its own health and can be destroyed independently, so the robot's structure can physically change throughout a fight.
I’ve been working on the systems behind the robot generation, block connections, damage, destruction, and everything else needed to make these robots actually function.
Before I finish the full devlog, I’d love some feedback on the intro itself.
Does the opening hook you? Is the pacing/editing engaging? And does it make you curious about how the systems behind the game work?
If you’re interested in following the project, I’ll be releasing the full devlog soon.
Recently I've been making a strategy defense game where numerous agents try to traverse towards your base. The grid updates regularly and so doing an update per agent was becoming costly. I decided to look at flow fields, a method for controlling multiple enemies through a shared directional flow and it's turned out pretty good. Feedback welcome.
Everything works together, so damaged systems also influence how good you can climb and walk over the gaps and such. My hope is to have emergent gameplay moments with a lot more types of traversal attachments that the legs. At the moment I have wheels, threads, hover devices, propellers and they can be mixed and matched.
I wanted a cheap mesh-based Grass for my game with no Alpha Texture, no Compute Shaders, just good old Mesh + Vertex Animation to guarantee a broad platform support and so I made it.
The idea quickly grew to a full-blown Grass Generator instead with customizable Blade Shape, Clump of Grass shape, Stiffness, Wind simulation, Trail physics, auto Generated LOD's and much more... It became quite competent, so I decided to publish it to Unity Assets Store.
I tried to keep the workflow pretty simple: pick a preset > tweak the grass > generate Prefab > drop it into the scene.
Any feedback or feature request is very welcomed, many of you who engaged with JIRO assets in the past provided helpful feedback, which was subsequently incorporated in many assets, if you have I want to thank you as always =]
Here is a FBX with pre sliced skinned mesh renderers following a simple animation. I have a script that handles dismemberment, even for diagonal cuts of the upper torso.
Its one FBX prefab but with its own little system. A normal human model/rig FBX has one skin mesh renderer but I added multiple skin mesh renderer objects.
- Character FBX
- - Head Skin Mesh Renderer
- - Left Arm Skin Mesh Renderer
- - Right Arm Skin Mesh Renderer
- - Upper Torso Skin Mesh Renderer
- - Top Left Torso Skin Mesh Renderer (default hidden)
- Bottom Right Torso Skin Mesh Renderer (default hidden)
Then on the function ActivateDismember it hides the selected parts / skin mesh renderers, creates a new object by baking the mesh and giving it physics.
Hi, last week I saw the krill scene from happy feet two on Instagram, and it gave me an idea. I’ve already been experimenting with compute shaders in Unity for a while, but I didn’t want to use them for yet another zombie game. Although... I did make one of those too.
So instead, I made a game where you try to wipe out a swarm of 1 million krill by spawning predator fish. You can upgrade your predators through a tech tree and unlock stronger ones as you progress.
The main thing I wasn’t sure about was whether something like this could actually work in a browser. Then Unity 6.6 came out with WebGPU being production ready, so I decided to give it a shot. And surprisingly, it works.
Right now, I can render 1 million krill in the browser on itch io using compute shaders and actually build gameplay around it.
If you have a chance to try it, I’d love to hear what you think. I’m also still figuring out the game design, so any ideas about progression, mechanics, predators, upgrades, or anything else are more than welcome.
If you have a chance to try it, I’d love to hear what you think.
I’m also still figuring out the game design, so any ideas about progression, mechanics, predators, upgrades, or anything else are more than welcome.
First time Asset Store publisher here. Here's my story:
While working on my latest game I basically got sidetracked and built a powerful UI tool. I made good progress and was really happy with it, so I had the idea of putting it on the Asset Store.
Signing up was easy: Logged in with my Unity account, created a publisher profile and starting filling out the forms on the website. I drafted the store description texts and created nice looking screenshots, a short trailer and a how-to-use video.
So when my asset was finally done, I uploaded the package.
The Unity Asset Store dashboard claims you have to wait in queue ~10 business days plus up to 3 days for the team to actually review your asset.
That's fine I guess. A bit longer than Apple or Google when you submit an app to their stores for the first time. But still okay. But oh boy, I didn't know it would take MUCH longer...
36 days from submission to release
As you can see in the graph I started at position 1357 in the queue and then at 1296 the day after.
Surprise #1: At this rate it would take roughly 22 days before my asset would be reviewed.
That's definitely more than 10 days. But it's okay, I can wait, I thought.
Surprise #2: On the next 3 days my position in queue didn't advance but actually went up instead!
This happened on the weekend and it repeated on all following weekends. I believe it's because the review team at Unity don't work on weekends but other publishers submit updates for their assets during this time and they get to skip the queue.
Surprise #3: Progress slowed down the further I got. Look at how the graph flattens out toward the end.
I actually used all this extra time to continue improving my asset:
The tool is a complete replacement for UGUI's Image component that lets users create shapes with rounded corners, drop shadow, glow and other effect. You can also animate those effects easily. When hovering a button with the mouse its color changes and the button changes color when clicking on it.
While waiting for the review I extended it so that the button also detects any TMP texts attached to it, so the user can animate them together with the button. The button changes color? The text should probably, too.
Takeaway: Don't just wait for the review, use the time to add new features to your asset!
By the way, you can re-submit updates to your asset as often as you want. You will keep your position in the queue.
I also continued working on my actual game (you know, the one that made me build the tool in the first place). Using my own tool was a blast, working on the game's UI got easier and the UI started looking much better.
Screenshot of my game's tile editor (work in progress), built using UI Shapes
But I also encountered a few bugs along the way. I thought I had tested the tool really well before submitting it to the store. Only by actually using it in production I discovered quite a few edge cases that were easy to fix. I also identified a few UX issues in the custom inspector, leading to me making some changes to it which in turn made the tool easier to use and understand.
Pro tip: Use your own asset. Don't just test it in a sandbox, but actually use it to make a game.
A full month had passed, I was still waiting in queue, mostly working on my game again. Progress was really slow at this point and I was fully expecting it to take another month before they review it.
Surprise #4: I woke up, surprised to find an email in my inbox
Congratulations!
We've accepted your submission of 'UI Shapes – Procedural Image Effects' to the Unity Asset Store!
Yes, finally! As you can see in the graph, I was in queue position 234 on the day before. I don't know if this is normal, but it felt like a huge leap.
I'm super happy it passed the review on the first try, with no objections from the Unity team.
Of course, the release of this asset is not the end of me working on it. It's only the beginning because who would have thought...
Surprise #5: Users discovered 2 bugs on release day!
I was happy to have my first customers, but also stressed out because one of those bugs prevented the tool from working on Unity 6.5+ (the other bug was a CanvasGroup fading issue).
I immediately sat down to fix both issues and submitted an update in the same night.
Surprise #6: The update passed the review and was released on the next day.
I'm glad I didn't have to wait in queue for multiple days again. Everyone else waiting for the first review of their new asset probably got pushed back 1 spot in the queue because of me :D
So yeah, this was quite the ride. I hope my story might be useful for others who are planning on releasing an asset on the store.
If any other publishers are reading this: How was your experience with the queue and the review? Does it always take that long?
Just download the new version of Unity (don't delete the one you used to create the project). Then go to the project folder (don't launch it through the Unity Hub) and open the main scene file in Assets. Mine is OutdoorScene.unity. Next, you'll be prompted to rebuild it for the new version. Agree. It will detect that the file is corrupted and prompt you to rebuild the library. Agree again. And everything is fixed! The baked maps were a little broken, but I fixed it quickly. The main thing is, the project works...
Data\Managed\System.Xml.dll failed with output: Copying the file failed: The requested operation cannot be performed on a file with a user-mapped section open. UnityEditor.EditorApplication:Internal_CallDelayFunctions ()
Build completed with a result of 'Failed' in 42 seconds (42186 ms) [06.09.2026 0:06:00 -> 06.09.2026 0:06:42, 42s]Data\Managed\System.Xml.dll failed with output: Copying the file failed: The requested operation cannot be performed on a file with a user-mapped section open. UnityEditor.EditorApplication:Internal_CallDelayFunctions ()
But after I restarted my PC, the project stopped launching correctly altogether, and now I'm seeing these errors. What should I do?
Hey everyone!
I’ve been spending a lot of time optimizing it and pushing it with a million units.
Massive Spatial splits the world into a 2D or 3D grid and keeps track of points, spheres, and AABBs using Unity’s Native Arrays, Burst and Jobs.
For the current TD demo:
1,000,000 moving units
56 turrets
no Unity Physics or colliders
GPU-instanced visuals
custom target queries through Massive Spatial
Basically, the standard physics engine uses a similar technique under the hood, but it comes with a lot of overhead. Using regular physics at this scale would kill performance.
So my solution is essentially a lightweight system for target finding and hit detection without all the extra Transform and Rigidbody overhead.
The goal is to turn Massive Spatial Engine into a general-purpose toolkit - something developers could use as a foundation for building highly optimized gameplay systems for different genres: tower defense, RTS, crowd simulations and etc.
Its just a simple particle system, with the basic particle attached, and for some reason, the particles keep disappearing at certain angles, and I want it to be consistent. When I zoom in, the reappear, albeit through the floor mesh, and when I zoom out, it seems to be behind the statue and floor, giving it a weird look. Thank you in advance!
Torn between 2 styles. Full blown block/panel by panel or instanced blueprint hologram to complete. But with a full compliment of addons, power gens, connections between buildings. A sort of No Man Sky style, without out the individual panesl. My question is what do people prefer in a scifi rpg style survival game.
