Lack of CoDs or CK2s that use HTML5/JS

[karneekarnay]

I've dabbled with HTML5 for a while now and I'm continuously surprised with what HTML5 and Javascript can achieve now. It's got me thinking why we don't see things like Call of Duty/Halo'esq or Grand Strategy games like CK2 using the language?  Are we still restricted by technical limits or is it just the lack of adoption from the wider Games Industry? 

 

[mattstyles]

For rendering a browser gets pretty much the same access to the GPU as any other application, albeit with less control (and via an older spec, webgl isn't quite as good as you'd have with different avenues).

For running calculations JS is impressively fast for a JIT compiled language but can't match any AOT languages such as most modern games would use. There is no getting around this, WASM (web assembly) is mooted, but, it may not quite get to the same place.

There is also an argument that JS is not a particularly good language for larger teams, such as those involved in more complex application (i.e. AAA games). Again, there are some solutions for this, such as Haxe, TypeScript, Elm or even using LLVM to convert your code in to WASM (or JS), as always, there are trade-offs.

The browser has some interesting restrictions that make achieving optimal performance tricky i.e. reliant on a platform canvas implementation, no UDP sockets, cross-platform differences are much trickier than most AAA games have to deal with etc etc.

Websites have different load requirements than local applications i.e. most games require a hefty download (or a large storage device such as a CD or memory stick) to a local storage device, would you accept this wait every time you wanted to play? Because the browser is not permitted to store so much information locally (there are hacks around this, but, they are nasty) so you'd have it for every play.

Browser users have expectations of browser-based applications that they typically do not for a local application, expectations can be broken down, but its hard.

I still think there is one issue that trumps all of these above: monetisation.

Closed application stores and closed source applications are opposed to the fundamentals the web is built on so a new method is required, which hasn't really happened yet.