top | item 27634582

(no title)

Jakobeha | 4 years ago

I really wish internet games were the future. Both from the producer and consumer side - as someone who tried to make games before, and currently owns a low-end mac (great for productivity but bad for gaming) - I really believe the ideal is web-games that are easy to start and work everywhere.

The issue is that at least in my experience web game engines suck. JavaScript is slow and awful for large-scale software, even today. Frameworks which compile into JavaScript and WASM (e.g. libGDX) compile incorrectly, leading to obscure awful bugs, and they're still slow because JS. Your best bet is a general-purpose game engine like Unity or Godot which can export to web, but still, the web export is often broken, missing features, and just slow. Even tiny games I've played on itch.io either don't load or play super slow, and sometimes that might be the dev's fault, but often I think it's the state of web-gamedev in general.

discuss

order

socialist_coder|4 years ago

> Your best bet is a general-purpose game engine like Unity or Godot which can export to web

Totally disagree about Unity. Unity's web export is awful. If you want a good web game you need to use a web-native engine like Pixi, PlayCanvas, or Phaser. I guess the Haxe & Godot tools are good too but I've never used them.

The huge advantage with using a Javascript/Typescript based game engine is that you can use the latest & greatest build tools like Webpack, and the latest ECMA/javascript features, instead of some proprietary tool chain that hasn't seen a major updated in years.

I think mobile game companies who have only been using Unity and have all their games built in Unity will have a hard time transitioning to mobile because you just cannot use Unity to make good web games. You have to rebuild them in a new engine.

But, if you use a web game engine, you can still target mobile. So, I would not use Unity for any new mobile/web game projects.

pjmlp|4 years ago

Except Unity on mobile will be able to take advantage of Vulkan, Metal and GL ES 3.2, while WebGL is stuck in ES 3.0 subset and who knows when WebGPU (with a MVP 1.0) will come out into stable.