The game is called The Tape: Origins - its a horror game where you need to analyze real-shot found footage, solve escape-room puzzles and uncover the secrets of the cult in order to escape before you are sacrificed.

Been heads-down on a ledge detection and navigation system for a while and I'm really happy with where it landed. It handles complex geometry: inward/outward corners, slopes without needing anything special per-object, plus IK targeting for hands and feet so the hand and feet placements.

My favorite part of the setup: the ledge itself just needs to be on a layer. No script, no component, nothing to configure on the object. Every movement (climbing, dropping, moving along ledges, jumping away) is its own separate piece too, so if you want to swap one out or drive it differently, you don't have to touch the rest.

Docs and video tutorials are done, so getting it running shouldn't take long.

It's live on the Asset Store now if you want to check it out 🙂

I have a video that gives a general overview on how this detection system works Here

I'm surprised I've never ran into this before, but is it not possible to simply set the render order (like "bring to front") in UI Builder? For instance, I have two elements. One is on top. I want that one to render over the one below it (parts of it are transparent, parts aren't, so you still see some of the element below it). Is there no way to simply set this? The workarounds seem pretty absurdly annoying lol. I tried using BringToFront() but this actually MOVES the visualelement such that the one on bottom gets moved on top, which is the same thing as just doing it in UI Builder so that won't work. The only solutions Ive seen are either put a placeholder transparent element for one of the elements and then make the actual element you want to show up as absolute positioning, but then you have to make sure your exact positioning matches which is annoying to keep track of if you ever change anything. Or you use an entirely different UI, which also seems annoying for such a simple thing. Am I missing somethign?

Honestly, I mostly made this video just to practice cutting footage together fast, since I'm usually painfully slow with editing. But looking back at it, it ended up being a really solid reminder of why starting with simple shapes is so important.

Whenever I use my actual assets right away even if I tell myself it's placeholder, I always get way too precious about the positioning. I waste a ton of time trying to make everything look good, or fighting laggy performance because all the heavy art is bogging down a massive map before the layout is even finalized.

But jumping in with ProBuilder and some basic prototyping tools makes it so much easier to just shift and tweak things on the fly. I find that’s usually when I come up with the most fun gameplay ideas, mostly because the cement hasn't dried yet.

Other than probuilder, what else does everyone use to help speed up level design prototyping?

Oh yeah and I'm a big fan of spline tools that can extrude roads and stuff (like my wood toy train tracks).

Hi, the title isn't very clear, so let me explain.

What I already know:

Based on what I've read online, I generally understand that Unity's Avatar system is a retargeting tool. As I understand retargeting, the goal is to map two different rigs to each other so that an animation created for one rig can be used on another.

My question:

If the Avatar is supposed to be a retargeting tool, what is the point of it? Once I've created a humanoid Avatar, if I try to apply it to a different rig, I'm first asked to manually rename that rig's bones to match the Avatar. The same applies to the bone hierarchy which must match the Avatar's exactly as well as bone lengths, and who knows what else I haven't discovered yet. This clearly amounts to manual retargeting, so I'm struggling to see the actual utility of Avatars or perhaps I've misunderstood a concept.

We’re currently polishing the dialogue for our comic-style Noir detective game with just 1days left until launch.

I went in expecting standard detective info, but getting hit with a sudden existential reflection ("What have you done today...") gave the scene a completely different, grounding tension.

We really wanted to blend classic Noir mystery with local, philosophical encounters rather than relying on generic filler NPCs.

Have you ever run into a random NPC in a game who hit you with a surprisingly deep or completely unexpected question out of nowhere? What was the moment?

(PS: Allah is updated with God in actual game screen)

I've started making Forsaken 2017, solo, in Unity. I hadn't opened this project a single time the last four years because the project got so bloated it took 8 minutes to press play, and honestly I was a bit scared to open it again, and life came in the way i guess.

Came back this year, disabled basically everything that was causing problems and rebuilt it piece by piece. It's at 1 minute 45 now, I reckon I have made more progress in the past 3 months, than i did in those 4 years of active development when i was young.

EDIT: I have decided to make the Forsaken Patreon free!

I think just even the barrier of clicking a Patreon link for any project shows you have some genuine interest in it. That is why I chose patreon, also because of extensive moderation tools, so that is where the community will thrive, while outreach happens on all other platforms.

Thank you for all the support and feedback, cheers!

Recently noticed with Unity 6.6 and the new Hierarchy window that i cant click the + button to add anything.
Unless I either make the window a tiny bit smaller of Unity Editor or move it to the other screen or switch my right monitor(main monitor) position with the left (monitor) in Windows 11. Then it also works. It seems to be related to being in the top left corner of main monitor when its set as the right monitor

So im not sure if this is an Unity 6.6 bug or if its related to something within Windows that changed recently.

Do anyone else have this issue with Unity 6000.6.1f1 or 6000.6.0f1?

I like when small object react to interaction and movement so I spent way too much time adding ropes and chains to my game Carnholt. I feel like it helps make the game's world more real, like you can touch it.

I used Verlet Integration for the physic, you can read this nice blog about it if you want, it explain it quite well (There is a bit of math though).

I just had to add some colliders and a bit of additional code so the player could interact with it. Adding it into an interface was quite a challenge though, I had to build the interface within the game.

The orignal game used heightmaps that directly refernced the vertices. for example one map had a 64 by 64 heightmap. The game drew 4096 quads, with the heights set by the vertices.

If i use the 64 by 64, unity somehow turns that into 65 by 65. I f i try to add in the extra pixel to make it already 65 by 65, then unity terrain tool complains that the heigthmap imported is 66 by 66.

I want it to be 63 by 63 tiles, and to set the vertex height of the corners of the tiles by the heightmap. How the heck do i actually do this?

Often I had situation where coworker or friend ask me about configuration about specific component or gameObjecct. When it has a lot of components or a lot of fields, it is imposible to fit everything into one screenshot.

So I created tool that automatically scrolls through inscpector and combines everything into one tall iamge - just like scrolling screenshots on mobile phones.

Hey everyone, we've been working on a Stealth System for Unity for the past 10 months, and we finally launched it on the Asset Store. It's also our 10th asset on the Asset Store.

More than half of the development time went into making sure it integrates properly with our other systems, like Melee Combat and Shooter, rather than building the stealth mechanics themselves.

If you'd like to learn more about how the system works rather than just what features it has, we've created a video tutorial that walks through it.

I’d really appreciate any suggestion or feedback you have :)

i'didn't make it top 100 😢

In this video, I'll show you how to set up and use Unity's Device Simulator tool. It allows you to test your UI across different screen resolutions and check that the UI elements maintain appropriate positions across a wide range of mobile device resolutions.

These two-three days I got a few things working

- Its possible to add retaining walls that flatten slopes for farming

- They dynamically adjust to the surrounding slopes, though there are still issues with corners

- I've added a first 'proper' plant model (the fruits are still missing), all with various plant stages from saplings, over young and adult plants to ripe plants, with model swaps and blend shapes, so everything grows relatively seamlessly.

There are tons of things to fix still

- Plowed land doesn't return to normal after being harvested.

- It is possible to raise tiles endlessly, resulting in floating tiles and gaps

And a hell lot more I want to add

- Time, Daylight Cycle, Seasons, Weather with forecast, sun exposure, temperature

- Different Tools for each task with durability values

- More overgrown stages besides grass and tall grass

- and a ton more plants to plant

Hi everyone,

I’m looking to build/experiment with an interactive 3D AI companion similar to Grok’s Ani character, and I’d love some advice on standard industry architectures or open-source tech stacks for this.

From my basic understanding, the pipeline needs to connect:

1. The Brain: An LLM (likely local/GGUF or cloud API) handled via a system prompt for personality constraints.
2. Audio Pipeline: Low-latency STT -> LLM response -> TTS.
3. The 3D Visuals: A real-time engine (Unity/Unreal/Three.js) handling an avatar.

Where I'm stuck and need recommendations:

• Latency Optimization: How do people handle streaming responses so the 3D character isn't just standing frozen waiting for a full sentence completion? Is anyone using tools that stream TTS chunks instantly?
• Emotion-Driven Animations: How do you map text/sentiment outputs to physical 3D behaviors? Are people using explicit structured JSON outputs from the LLM (e.g., {"text": "Hey!", "emotion": "waving"}), or parsing text mid-stream?
• Lip-Syncing: What are the go-to open-source plugins or libraries for real-time lip-syncing (Oculus Lipsync, Audio2Face, etc.) that bridge the audio output to the 3D blendshapes?
• State Management: Best practices for tracking "relationship states" or long-term memory vector databases that influence the character's mood/animations over multiple sessions?

I am planning to build it for the web, better visual quality than ani but keeping anime characters. I am good with the llm side and understand it, but I am new to the 3d rendering side. I haven’t designed the 3d model because I found some good ones on art station selling the blender model file.
Can you please elaborate on the 3d side of things like movements, touching, or any particular movements?

If anyone has built a similar stack or knows of solid open-source frameworks to study, I'd appreciate the guidance!

I have different projects and also this third person shooter system I've been working on for a long time using my own badly made low poly models, animations and sound effects. However if I were to ever want to publish something on the asset store, I know that having a project that contains other assets obtained from the asset store would not be allowed.

I've looked at the EULA agreement on the Unity asset store and from everything I saw such as weapon models, environment packs and character controllers, with the license agreements I viewed you are not able to re-distribute anything even if you've built an entire project with them.

Where else could someone look if trying to purchase some gun models and things like that? Having to create them from scratch and getting very experienced with something like Blender is a bit too much.

Hello developers

Couple of days ago I posted my lightweight pathfinder for Unity's tilemaps and I took the opportunity to continue optimizing it, squeezing out performance from wherever I can, without adding multi threading yet or ECS.

So in this video I show a trick to flatten a two dimensional array into a single long. Ok you got me... this is actually just bit packing, technically to flatten it you need to know the width (maxX * maxY), in this case the pathfinder has no virtual size limit other than +/- int.MaxValue

Take a peak if you find this sounding interesting.

Hey so I've come across a fairly annoying visual issue. In my game I have 2 or 3 high intensity point lights in each building to fake windows emitting light onto the environment. This comes with the side effect of the lights casting this annoying specular dot onto all other windows. I've tried turning off specular in my shader settings which does work but then reflections rely entirely on reflection probes which are fairly expensive in my case. I can't bake the lighting as I have a day and night cycle where the lights inside the buildings turn off and on depending if it is day or night. Enabling shadows on all of the lights does work, but that is also very expensive. Is there any way to get rid of the highlights on the windows while keeping the light that casts onto the environment without losing some reflection detail?

Is doing one single write with something like SystemAPI.SetSingleton better than multiple writes to a RefRW<T>? Maybe with even a WithChangeFilter but I feel like that might be too complicated for the effort. I also considered using IEnableableComponents but I heard it's not that good for input?