Alex Goldring

Graphics engineer, not a cat

[Alex Goldring @SoftEngineer]

@OWL_OF_INGENIUM Thanks!

[Alex Goldring @SoftEngineer]

Foliage in WebGPU
* tree generation
* wind simulation

[Alex Goldring @SoftEngineer]

@nishantcy This is skinned mesh, simulation is full fluid simulation in a 3d clipmap.

The simulation results are accumulated via a side channel next to the foliage "bones".

Roughly a mixture of "God of War" wind tech and "Alan Wake 2" foliage tech.

[Alex Goldring @SoftEngineer]

@BaalBaell I'm using Opus 5.5 generally, but with my limited experience on Sol 6.1 - yes, it's worse. It's very lazy and pretty awful at using tools.

[Alex Goldring @SoftEngineer]

Pretty disappointed in Google's Gemini 4 Argon.
Antigravity as a harness also feels like Claude Code from about 7 months ago

[Alex Goldring @SoftEngineer]

@cybereality Too many pet names.
The voice is interesting.
Quite a few noticeable hitches, though I gather it's the current tech limitation.

[Alex Goldring @SoftEngineer]

@bg1987 Fully agree that tuning gameplay and distilling your ideas into something that feels great is the core challenge

[Alex Goldring @SoftEngineer]

Not really. People judge Vampire Survivors like it's a simple game or one that's easy to make - I don't believe so.

As a software engineer and a game maker myself - I see Vampire Survivors as a very complex and elaborate piece of software with deliberately primitive graphics.

[Meringue Rouge @MeringueRouge]

@SoftEngineer @Shabba_Es I refuse to believe you spent 20k on an AI to make a Vampire Survivors clone that looks like that.

This has to be outrageous bait to get Twitter engagement, right?

[Alex Goldring @SoftEngineer]

@ReadOnly4048 Hey, happy to hear someone else is tackling this problem!

Best of luck!

[Alex Goldring @SoftEngineer]

Just released my AI coded game on Steam:
https://t.co/R1pQwG31p3

What it took is:
- a few hundred prompts
- about 6 months of development
- just over 1500 commits.

[Alex Goldring @SoftEngineer]

Hey @MeringMerong ,

You wrote a thoughtful comment, so I'll engage in kind.

First on templates. You seem to be under the assumption that my game is a clone of anything - it is not. I was, likewise, inspired by many games, but all of the mechanics are 100% original.

You also seem to be assuming that the game is simple mechanically - it is not. It is deliberately made easy to play, but there are layers upon layers of interacting systems in there.

Next, and I don't mean this as a flex - the engine is my own, and I've been working on it for over 10 years now.

Your game look fun, and it's colorful, and I could see myself playing and enjoying it, so kudos to you for achieving that.

Broadly speaking my game has 3 major layers:
1. ballistics + stats + elements, this makes up the core of the gameplay
2. the level generator + the game director, this is resposible for creating moment to moment challenge for the player
3. meta progression, that's the long-term motivation for the player

The technical complexity is one thing, but those "little things" you mentioned - each part of the game is full of them. And all of those disparate systems are tuned to work well together, so you don't get a technical demo but a fun game.

You mentioned UI looking very basic. That's true. But, here are a just a few things off the top of my head that go into the combat UI:
* currency and kill-count display bounces procedurally when you earn some, more and more quickly earned - the bigger the bounce
* weapon outlines and ammo counts are color-coded by weapon element
* objective markers stick to the edge of the screen for main objectives
* ability cooldowns play flash animation when they are ready
* ability cooldowns have a pie-type cooldown and desaturate while inactive
* attempting to use an ability comes with a UI animation when it's not ready
* the edges of the screen flash in proportion to the damage you receive
* status effects with timers are displayed when the player has buffs/debuffs

It is easy to dismiss my game on the look of the screenshots, or even the video. But I promise you - I didn't make the game for pretty screenshots or even a video, I made it to be fun to play.

As someone who has played likely every half-way decent survivors-like game on Steam, I believe mine is both entirely different from the rest and has the best late-game elasticity, allowing the player to enjoy multi-hour runs if they wish to. That's something no other game in the genre offers, to my knowledge.

You can think my game is trash and low effort, or that I'm entirely unskilled. Most people here seem to 😅

I'm fine with that, I wrote this because you offered, what appears to be, a well considered argument, so I thought you might be interested in a proper response.

Whichever way your views swing, thanks for taking the time to comment and for checking out the game, even if at the very surface level, I appreciate it

[Alex Goldring @SoftEngineer]

@ekyra69 @MeringueRouge @Shabba_Es Not sure what "larping" means here, but yes, it's an estimate based on per-token price.

It's hard to price accurately on a subscription, and I was doing other things during that time, it wasn't 6 continuous months.

[Alex Goldring @SoftEngineer]

@primord_io "I'm not going to denigrate" - proceeds to denigrate.

[Alex Goldring @SoftEngineer]

@JasonHa35545521 If you say so 🤷‍♀️

[Alex Goldring @SoftEngineer]

@JasonHa35545521 No joke, the game is coded by AI, the 3d modelling is ~99% human work, including the capsules.

[Alex Goldring @SoftEngineer]

@animiotics @MartinJIFuller They are not stable at all, or rather - there's nothing that guarantees their stability.

VRS is a clever and good tool, but it's not magic.

rates stay relatively stable because of TAA and because of the underlying viz buffer being quite stable.

[Alex Goldring @SoftEngineer]

Variable Rate Shading in WebGPU

primary reference:
VRCS in DOOM: The Dark Ages by @MartinJIFuller

[Alex Goldring @SoftEngineer]

@AgustinCordes It is, no doubt about that. I've been doing game development for an embarrassingly long time, and I don't recommend it whenever I talk to people.

I'm happy though, I know what I'm getting out of it and why I'm doing it

[Alex Goldring @SoftEngineer]

@EckoNode Thanks! I appreciate it!

[Alex Goldring @SoftEngineer]

There are different kinds of motivation behind what anyone does. I didn't do this for money.

I did it because I was interested in the tech problem behind pathfinding for a massive number of enemies initially.

Then, I was interested in seeing how far I can push that hand-made prototype to a real fully-featured game, and then I just found it fun to continue.

I already got what I wanted from this, so don't worry about me

[Alex Goldring @SoftEngineer]

@taiyolang The look was not the goal, the gameplay was. But I appreciate the feedback.

On the engine - yes, it's a custom engine, I didn't develop it for this game, but it is what the game is built on

[Alex Goldring @SoftEngineer]

Games tend to have a hockey stick sales graph, and it doesn't look like I'm getting much here, that's true.

I didn't make this for the money, at least not as a primary motivator. And I'm going to keep supporting it for a long while, so I'm honestly not disappointed.

I released the game because I believe it's a fun game, and it's been a fun journey to this point, no too much complexity behind my motivation.

That said, money would be nice

[Alex Goldring @SoftEngineer]

@cybereality I'd like to think I have thick skin, but it's almost funny how brutal some people are 😅

But yeah, give them a chance and they would knock out every major AAA game in an afternoon with a single-shot prompt, I'm sure