Hey, what's up with WebGL games?

[flerokoo]

According to http://webglstats.com/ more than 90% of all devices now support WebGL!

Does that mean that it's (finally) time to make 3D games? Has someone had an experience in selling/distributing webgl based games? 

And what do you guys think about 3d web games in general?

[mattstyles]

You might want to change the title to mention 3d gaming, almost everything you hit on the internet currently will use webGL and most frameworks will favour webGL with canvas rendered fallbacks, been that way for years.

Even though JS is a bit rubbish for complex calculations the browser has the same access to the GPU as any other application so if you're able to manage the gpu well enough from JS and the bulk of your rendering happens there you can create some really stunning and complex scenes and applications right in the browser. Still its a question of difficulty, 3d gaming is generally far harder to do and get right and with limited clear monetisation options its a bold call to target web over other platforms.

It's not so much the webGL stats as the impending stability of WASM that should be getting you really excited about creating richer more computationally expensive games right in the browser.

[pyalot]

16 hours ago, flerokoo said:

According to http://webglstats.com/ more than 90% of all devices now support WebGL!

The number is based on 12-15 million visits/month that come mainly from sketchfab.com, mrdoob.com and playcanvas.com (around 80% of the traffic contributors). So the statistics has some bias because it's based on what WebGL centric sites see.

I would like to diversify the sample base if possible, and it would help tremendously if less WebGL centric sites (like this forum for instance) could embed the tracking code:

<script src="//cdn.webglstats.com/stat.js" defer async></script>

If that would be difficult for a site, please contact me and maybe we can work out an alternative way to contribute your traffic.

16 hours ago, flerokoo said:

Does that mean that it's (finally) time to make 3D games?

There's no yes/no answer to this question, it depends. WebGL support itself is not an on/off thing, it comes with a large number of capabilities and some caveats. For instance the major performance caveat usually indicates that an implementation is software/swiftshader, which is far slower than actual GPUs. Then there's the various parameters you need to observe (like how many textures you can use in a fragment shader), and if your needs exceed the commonly available capabilities, you need to think of ways to render in those cases (this is commonly known as implementing alternate renderpaths). The same goes for the variety of extensions that can give you access to floating point textures, sRGB FBOs, VAOs etc.

Then there's some reservations about I/O latency (user input until stuff shows up on screen). For very twitchy games (like say a counterstrike clone or a hectic jump&run), this tends to be a problem, because the latency is a bit higher than it is natively. This is improving, in particular because of efforts around WebVR but for the time being it does somewhat inhibit that class of games.

So it all kinda depends on your needs and also on your audience, a bit more on the later point below.

16 hours ago, flerokoo said:

Has someone had an experience in selling/distributing webgl based games? 

And what do you guys think about 3d web games in general?

I have no personal experience with web based 3d games, but I've been in contact with a few devs. You'll usually face around 3 issues:

1. You will need to integrate your own payment solution. Depending on where you live this can be hard to impossible. It's also the case that app-stores have pre-established payment relations with customers, and you would not have that benefit.
2. People might be less willing to pay for a web-based game than for one from an app-store, even if the app-store game is just an app-container wrapped around a browser with your game in it. It's weird, but that's what it is. I've got this reported from several developers who went through just that process both on the iOS and Steam app-store.
3. People might even be less willing to play a web-based 3D game on general principle than an app-store based one. One developer did some A/B testing on this, and found that when he put two buttons on his landing page, one goes "Play Now" and the other "Get it on Steam", that around 90% of visitors pressed the link to get it on steam. There might be a need to do more research on this topic, but it sure does seem like the web is at a disadvantage there. The only recommendation that I give there is, don't make people press buttons on landing pages to get to your game, just throw them in immediately. If you give them a choice, they'll chose wrong