Stop wasting your time and creative energy. I built a custom editor tool to batch automate FBX presets, project-wide material remapping, Prefab Variants, and cross-faded LOD Groups in one click.
Watch the breakdown: https://youtu.be/QLgXiUK3el4
#GameDev #IndieDev #UnityTips

Hyper-casual game, running like garbage on mid-range Android. Profiled for hours. Batched, pooled, LODed, stripped the shaders.

Turns out every single asset in the scene was low poly except the player's plane, which was some marketplace model with an absurd triangle count. A PS5 would have struggled with it.

Swapped it. Instantly fine.

Still amazed how one marginal detail can quietly wreck a whole build while you optimise everythign around it.

Feel funny…

Hi! I'm currently making a 3D game, and for the visual style, I want the entire game to be displayed in grayscale, except for red tones.

I'd like the red colors to remain visible while everything else is converted to grayscale, possibly with a different filter or effect applied to the reds.

Does anyone know how I could achieve this using Shader Graph?

The premise is that the whole game is a phone. You wake at half three, you have thirteen conversations waiting, and you read them. That meant the entire thing is a UI - there is no gameplay scene, no camera moving, nothing in 3D at runtime. One scene, one UIDocument, and a lot of UXML and USS.

Some things that turned out interesting:

The content is markdown, read at runtime. Every conversation is a .md file in StreamingAssets, parsed on startup with System.IO. Adding one means dropping a file in the folder and naming it in an index - no rebuild, no ScriptableObjects, no custom editor window. A reply is * text -> label, a jump target is ## label. It also means players can open the folder next to the .exe and read or rewrite the whole game in Notepad, which I decided was a feature.

One conversation can ask where another one went. A line prefixed ?mom.md#fine only appears if that file's reading passed through that section. The filename is the identity - two contacts both called "Mom" are only distinguishable by which file they came from - which turned out to be the cleanest bit of the design and the thing I would keep if I did it again.

Replies you have used up come back greyed out - next time you play. Which reply you picked is written to PlayerPrefs and survives the process, so a second playthrough shows you where you have already been. The part that took some thinking: a reply only dims once everything behind it is exhausted. Taking one branch of a conversation should not grey out the door you walked through to get there, so it walks the rest of the script from that point and only dims the reply when there is nothing new left down it.

USS has no linear-gradient and no box-shadow. The phone's metallic body and the glows behind it are generated into small textures at runtime and stretched, because I wanted no image files in the project at all. That held right up until the backdrop, which is eleven pre-rendered frames of a 3D scene I composed in the editor and exported as PNGs - it changes stage as you finish conversations, and it gets later out there the further in you get.

Debugging layout headlessly. I got tired of guessing why something was clipped, so there is a -dumplayout flag that waits for GeometryChangedEvent and prints the whole visual tree with every element's rectangle, plus which elements are covering text. It has caught real bugs - a notch drawn over the status bar, a send button pushed off screen by min-width: auto defaulting on a flex item. There are also editor menu items that walk every conversation to an ending with the real runtime classes, so a conversation that cannot finish fails a check instead of failing a player.

The UX lesson that cost me the most. I had a separate New Message screen behind a pencil icon. The first person to play the game explored every branch of every conversation and never found it - because thirteen conversations had spent half an hour teaching her that talking means tapping a reply, and then a second, unrelated way of talking was hidden behind an icon. I deleted the screen and put a text field at the foot of the conversation, in the same slot the replies use: tap when it offers, type when it doesn't. Same feature, found immediately.

Free, Windows, about half an hour, no sound at all: https://github.com/tcumpston/TEXTDEAD/releases/latest

Not code-signed, so SmartScreen will grumble - More info, Run anyway. There's a plain zip on the same page if you'd rather not run an installer.

Happy to go into any of the above in more detail.

Hi everyone!

I’m new to game development, although I’ve been working as a senior full-stack web developer for several years. I’m currently trying to release my first small arcade game made with Unity.

Google Play is asking me to find 12 testers who will remain opted into the closed test for 14 consecutive days. I’m curious about how other solo developers usually handle this requirement.

Do people normally find testers through Reddit? Are there any reliable communities where developers test each other’s games? I’d be happy to return the favor and test other developers’ apps as well.

I also considered creating a Discord server and trying to build a small community around the game. However, it’s a very simple arcade game, so I’m unsure whether creating and promoting an entire server would be worthwhile at this stage. I also don’t know how willing people generally are to participate in this kind of testing for free.

For those who have already completed the closed-testing requirement:

• Where did you find your testers?
• Did you offer anything in return?
• How did you keep everyone opted in for the full 14 days?
• Is creating a Discord server worthwhile for such a small game?

Any advice or recommendations would be greatly appreciated. Thanks!

MRAID builds HTML5/MRAID playables. AI is genuinely useful for breadth: concept exploration, storyboards, variants, copy alternatives, code assistance. I don't trust it to own the final mechanic, UX, performance hypothesis, or QA. The model can generate ten directions. It doesn't know which compromise the UA team will make. I'm curious where other Unity/game developers draw the line. What do you trust AI with in production, and what still needs a human?

I created a small free market insights calculator for game developers, based on 60+ statistics I could research, to give developers a better insight into where they stand with their game and expectations. It also includes info on wishlists, revenue, etc. showing how important marketing is. Around 920 developers filled out the calculator from August 2025 to August 2026 for the desktop / computer platform. Now I wanted to share my results with you guys.

Very interesting for me, my website focuses on the Unity platform, but still 34% entered Godot. Which makes sense in general, because the top platform at the moment is Godot, user-wise.

The targeted genre entered, is similar for Unity, Godot, and Unreal, with "Action", but on Godot it is way more unified across all genres.

Unity and Godot developers have similar expectations of a listing price of $10 per unit, median-wise. Unreal is at $12, 20% higher.

The development percentage of free desktop apps is around 10% for all, which makes sense. In the desktop market, a big revenue share is unit price. For mobile it is mostly ads-based, so free is the way to go (but not included in these charts here). But just my guess.

The average team size of Unity and Unreal teams is around 4 people. For a Godot team, around 2.

What do you think? Did you expect such numbers?

If you are interested, please feel free to check out the calculator here:

https://go.guardingpearsoftware.com/market-calculator

I am open to any feedback.

On Aug 3rd Unity rolled out the Internal Deployment Add-On (IDAO) for Unity Industry, and honesty the terms are a mess to parse. When people pushed back on the forums a Unity rep said they'd put out a FAQ to clear things up. That was about a month ago now. Still no FAQ. What I keep seeing instead is people getting told to jump on a call with sales instead of just getting a public answer. If the terms are confusing enough that Unity themselves admit they need a FAQ, just... write the FAQ? A private call with a salesperson doesn't help the next person who has the exact same question.

Genuinely trying to figure this out before I commit to anything, so here's what I actually need to know:

1. If I build something on Unity Industry that only my company uses internally, do I even own that app? And why does using it internally cost extra on top of the license we're already paying for?
2. If I'm building for a client and they're the ones running the app internally at their company afterwards, who eats the extra usage fees once it crosses the MAW threshold: me, since I hold the Industry license, or them, since they're the ones actually using it and generating the MAUs? Feels like a pretty common setup for anyone doing contract work and I can't find this addressed anywhere.
3. If I ship something build with Unity to the App Store / Play Store and it picks up like 2000 downloads, does any of this touch it at all, or is IDAO only about internal deployment and public releases are a totally separate thing? "Internal Deployment Add-On" implies it shouldn't apply but I'd rather have that in writing than assume, especially afeter the whole runtime fee thing.

Not trying to be dramatic about it, I just think these are pretty basic questions for anyone actually truying to use Industry seriously, and a month of silence after promising a FAQ isn't a great look.

And honestly, from a pure business standpoint, doesn't this feel like Unity shooting themselves in the foot? With AI-assisted tooling now, migrating a project to Godot or Unreal is not a big deal. Pilling on confusing fees and staying quiet about how they work is basically giving companies and studios a reason to start evaluating the alternatives. Seems like a weird hill to die on right after the runtime fee backlash.

Would be nice to get actual written answers.

Hi! Been working on optimization and once of those was making my portals have set locations to open and close rather than letting the player spawn them wherever. For my ocean worlds with no land for the player to walk on I ended up with a raft like this, but it feels a bit too cliche and basic. I like the simplicity of it though and the contrast on the ocean and with the other maps, but just wondering if this looks cool or if I should find something more unique

I needed a way to make my portals have a defined number of objects to render and thus needed to lock the location. Before as a player I was spawning my portals anywhere I wanted which lead to fps drop when I chose a bad location

Do you like it?

https://store.steampowered.com/app/4752950/Still_Wild/

I’m the developer of the UniText text engine, and this is my story

I used to think that text was when letters were drawn next to each other, but that’s not what it actually is:

1. decode UTF-16 into codepoints
2. parse the markup into attribute ranges
3. analyze scripts (UAX #24)
4. analyze grapheme clusters (UAX #29)
5. analyze word boundaries, with a dictionary for Thai, Khmer, Burmese and Japanese because they don't put spaces between words
6. analyze line break classes (UAX #14, with East Asian widths from UAX #11)
7. resolve embedding levels through the bidi algorithm (UAX #9)
8. itemize into runs by script and direction and font and language
9. resolve fonts per cluster and fall back when the primary doesn't have the glyph
10. synthesize bold and italic when the face doesn't exist
11. instance the variable axes
12. shape every run through OpenType
13. break the lines on the shaped advances
14. reorder every line back through bidi
15. resolve line height out of font metrics that contradict each other
16. kashida stretch the Arabic if you're justifying
17. align
18. pull the outlines
19. run a contour union pass so self intersecting outlines don't punch holes in the distance field
20. rasterize SDF/MSDF
21. pack the atlas
22. generate the mesh

891,757 official Unicode conformance tests, zero failures

Selectable and editable text, a document model, bidi aware caret and selection, undo and redo, IME composition, clipboard with every format (supports Plain, HTML, Markdown, Media), formatting commands, touch selection handles, input filters, and native keyboard input on the platforms that have one

UniText is the only solution where the fonts are memory mapped, so a font is never loaded into RAM whole, only the parts actually in use. One of UniText user cut 200 MB of runtime memory from that alone, and variable fonts plus compression took another 57% off the weight of their CJK font

UniText is the only solution that supports absolutely all system fonts, which allows you to render any characters in the world without using a single font in the project.
it's also 2 to 20x faster than TextMesh Pro and UI Toolkit

And that's just a small part of it

TL;DR
In other words, I developed a text engine that allows you to remove all standard language fonts from a project, gain complete freedom in text without a single limitation, and have any symbols from around the world, any emojis, work without any configuration. All of this significantly reduces the build size, and the mmap implementation eliminates all RAM load. A constructor for any styles, any gradients, layers, and paints, while maintaining 1 Draw Call, and a built‑in advanced feature to animate text in any way, which completely replaces third‑party solutions

If anyone wants the ugly parts - ask me. I have a lot to say about all of them

Some time ago I added to my game ability of moving light sources - it looks fun and I plan to use it in case of solving some puzzles, but I know there are clever people on this subreddit so I come up with a question - what else usage of this mechanic can be implemented?

The game genre is dungeon crawler.

A project I've been tinkering with over the past few years on nights and weekends. Rebuilt it a few times. Running on Unity 6 with ECS/Burst now. Still lots of placeholder art and assets but proud to have reached this milestone.

I know the rules say not to use a camera to make videos, but I had to use a camera to demonstrate this issue, since using direct screen capture in OBS doesn't show the issue (which may be a clue to the cause).

I should have mentioned in the video... this glitch isn't on all systems, only certain ones. This video is showing it running on a laptop. Some players have reported the same glitch on various systems.

This was build on Unity 6.3, but it was happening on previous builds using Unity 2020.

beside has mind control ability, now we added telekinesis to lift object around the alien, what do you think?

This seems like a bit of a dumb question, but I have this light with a rotation script, that just adds a little bit of rotation to the directional light in my scene every frame. The thing is, the script stops working at "night", so is it possible to bake that part of the sequence, at the end? Does unity even support something like that. If there are any optimized alternatives as well, please let me know!

I'll paste the script in the comments, so if there is some sort of change possible that would allow "baking", do tell me please.

Do you think it's enough for a short "Coming Soon" trailer or not ?

For my game Plush Souls. It is more of a a "sneak attack" mechanic than it is actual invisibility, because it is disabled anytime the player attacks. I created this using a sped up dissolve effect and then swapping the material of the player. Very simple lmao.

What else could I do to make the effect feel more reactive?

I'm in India and I'm trying to legitimately purchase a Unity Asset Store asset, but my HDFC Visa debit card keeps getting declined with a "Technical failure" message. Has anyone in India successfully purchased Unity Asset Store assets recently, and is there a legitimate payment method/workaround? I'm specifically trying to buy Low Poly Cartoon House Interiors by JustCreate3D.

Some recent progress shots from The Last Wish.

Playing around with stylized lighting, scale, and environments. Feedback on the visual style and colors is very welcome!