Update from our project / Post mortem
Hey, I’m just posting this to give feedback about our experience with WebGL with our project. A lot of people seem to be insecure with WebGL, so maybe this post will help. We are a working on a port of a multiplayer 2D game, the game is already published and have players around the world. The main goal was to focus to port the current engine to Unity. We focused standalone/mobile for the development, WebGL was secondary. But we were trying to always have a semi-working WebGL version.
At the moment, our Android build is ~50mo and WebGL is 85mo Release and 25mo Compressed.
(without the asset bundles)
What went well:
-Almost no WebGL specific code, the only specific part in the project is the class that handle websocket request. We use the jslib on the asset store for that.
-I have almost 0 experience in web development and 0 interest, but I was to able to build and deploy easily.
-IL2CPP, it’s suprising how it work well. Our codebase have a lot of code with Linq and (unfortunately) some really complex part. And I think we needed to modify only 2-3 things to make it work. And the problem were related to AOT error. So it’s was not WebGL specific.
-No optimisation to do. Since it’s a 2D game and we had mobile in mind, it’s not surprising but I tought we whould have some optimisations to do.
-A lot of update/fixes from Unity team since January 2015 (the moment we started using WebGL)
What went not so well:
-Build time/Debugging: The iteration time is really slow. And some time only you can only debug with Debug.Log.
-Build is not consistent on different browser. Currently our build only work with Firefox. Since a lot of development is still done on WebGL for the browser, it’s probably the point that add the most insecurity to our project.