I get your concern about supporting the web; I think it's because a lot of people really want us to be a web game engine, but we aren't. We're a mobile game engine. We simply chose "HTML5" as an API layer because of the excellent tooling around the ecosystem. It just so happens we support the web as a side-product, and that's pretty cool I think, but not really the point of the GC DevKit.

Great feedback on the fuzziness of the license WRT contributions! Overall our intention is to let people keep their source code proprietary, but protect clear bug fixes and enhancements to the core engine. We never, ever intend to come up with some "gotcha" and start charging money. Any suggestions on changes or clarifications to our license that let us accomplish that are very welcome! (Though please read our first attempt in Exhibit A "Modifications" first)

The good news is that the core native engine (c/c++/java/obj-c) is very obviously separated from the game code and so that's easy; it's just the JS libraries that we have to be precise about. Perhaps the MPL has the right wording and guidelines around that?

("The MPL is a simple, file-level copyleft license. The MPL’s “weak” copyleft is designed to encourage contributors to share modifications they make to your code, while still allowing them to combine your code with code under other licenses (open or proprietary) with minimal restrictions.")

But games are all different. An improvement to the engine for one kind of game may not be an improvement for another kind of game. Most games would probably only use a small part of the engine.

Will the engine become a monolith with support for pretty much everything, and all games using the engine will only need 10% of it? Or something else?

I think you need to state very clearly what the goal of the engine is (best 2d engine, best tile game engine, best ??? engine, general purpose engine, etc?), which can help define exactly what kinds of contributions are actually useful.

>But games are all different. An improvement to the engine for one kind of game may not be an improvement for another kind of game. Most games would probably only use a small part of the engine.

I don't really see them as all different. Especially at the engine level. Especially considering most 2D engines (Cocos, Game Closure, etc) are mostly the same and use similar concepts.