For anyone interested in WebAssembly and the future of gaming in the browser - my team and I at Wonder Interactive are bringing the full power of native gaming to the web. We're building out a platform and suite of tools that allows developers to publish, host, share, and monetize their games directly to their players online, without 30% fees.
The current focus is on the Unreal Engine (4.24, 4.27) and UE5 support which is coming later this year. Other engines will follow such as Unity, Godot, Open 3D Engine, and custom engines we can provide porting for on our paid plans. We're building out a WebGPU backend for UE5, to really enable high end desktop and console quality games in HTML5.
Given the boundary condition that a web browser can certainly not allow a game to make more out of the hardware, that is, the best case is that in-browser gaming is equal to native gaming, and the probable case is that it's somewhat worse; what is the rationale between targeting a browser?
Yes, an example is the .io games, which kids love to play, because they just go to the browser and load the page and they are playing with other kids, just like that. No installation, no setup, no network debugging, etc. These are simple games, but more advanced games would also benefit from nearly frictionless entry to the game.
How would you even "run the numbers"? How about try the idea and see if it works? All it would take is one hit game for devs to see the potential of web gaming (and they already see it. Why do you think flash games were so popular?) If Fortnite could run in the browser it probably would have instantly become the biggest game in the world.
Nice. How will you fix big download times? I imagine it sounds to cool to be able to start a game in a browser straight away until you realise you still have to download it like any native game, so in reality other than avoiding installing a marketplace like Steam it's not much of a difference experience to play in the browser.
The two main ways we address this are with our asset streaming system (only loading in what is needed, everything else loading in the background at runtime) and compression (both data compression and texture compression).
Sound promising, but I tried the sequencer and the town game on Firefox android, both of them loaded but almost every texture was black. And in both examples, the requirements were all "undefined", this doesn't look good at all.
The current focus is on the Unreal Engine (4.24, 4.27) and UE5 support which is coming later this year. Other engines will follow such as Unity, Godot, Open 3D Engine, and custom engines we can provide porting for on our paid plans. We're building out a WebGPU backend for UE5, to really enable high end desktop and console quality games in HTML5.
Further reading, with demos attached:
https://theimmersiveweb.com/blog