Superpowers – The extensible, collaborative HTML5 2D+3D game maker
sparklinlabs.com
sparklinlabs.com
I'll be happy to answer any questions, obviously.
We want Superpowers to be as "Web native" as possible. TypeScript is just JavaScript + types, they're working to align with ES6 and maybe even merge with ES7+, they have a lot of mindshare, their compiler services are awesome (having intellisense-like in the editor is a big deal) and the whole compiler can be run in the browser, which means it'll work everywhere just like the Web we're building on.
(I've talked in more details about it in our devlog - http://forums.tigsource.com/index.php?topic=46317.0)
Demo doesn't work well in Firefox (latest, Developer Edition) on Win8.1 x64, though. Z-buffering looks totally busted and the framerate is bad.
https://www.dropbox.com/s/jymr2vr8ip52tw5/Screenshot%202015-...
That's quite the weird rendering bug. Everything works fine here in Firefox stable (tested on both Windows 7 and 8.1). Probably a driver issue of some sort? We're really not doing anything fancy, just using the default basic mesh material of Three.js and simple meshes.
I'm curious why I'd choose this over say Unreal (free), Unity3D (free), Source 2 (free). ok, they aren't 100% free but they're effectively free until you make money.
They all have a track record of shipping games on multiple platforms (some more than others). Some already have huge communities. 2 of them export to HTML5 if that's what you want. They have large detail animation system, cutscene systems, visual programming system, shader editors, quality importers for tons of formats...
Script changes are marked as drafted until you explicitely choose to apply them to the current build (Ctrl+S), so you won't break the build because you're in the middle of typing a line of code. There'll be a membership and rights system, right now there's only a server-wide password. Real-time script editing uses the same tech behind google docs and such (operational transform). There's no special history support right now, the server owner can version the projects in Mercurial / Git manually. Later on we'd like to have a project-wide revision system right in the UI.
> I'm curious why I'd choose this over say Unreal (free), Unity3D (free), Source 2 (free). ok, they aren't 100% free but they're effectively free until you make money.
Unity, UE4 (and probably Source 2) are multi-GB download and fairly heavyweight tools with sooo many options and drawers. Don't get me wrong, I love Unity (haven't spent much time with UE and Source 2 isn't out yet), but Superpowers is competing with Game Maker, Construct and the likes, not the big engines.
Great to see Superpowers making progress. I'm interested in your choice of Three.js. Did you consider other WebGL game engines? Like PlayCanvas[1] (where I'm a co-founder) which is also open-source[2] but more designed towards games than three?
We went with Three.js because it's pretty versatile and does everything we need in the near future. Superpowers is built in 3 layers: the core (which handles projects, storage and networking), the system (which is basically Three.js + game instance/actors/components) and all the editor & asset plugins built upon the system. The system can be swapped out with something else, and a new set of plugins can be built on top of it, or the existing plugins adapted for a new engine, so the community will totally be able experiment with other engines including PlayCanvas!
We spent a lot of time designing the Superpowers platform so that this would be possible :). In fact, Superpowers could even be used to make movies or write novels collaboratively, with the appropriate system and set of plugins.
This sounds exactly like an idea that I've had before. What my friends and I called it was 'GameTube' which would be a platform for making and sharing games online. That was part of what motivated me to work on the Duck programming language, which would be a simplified language to develop in, along with some web IDE's for game making and common content that aspiring game developers could work with.
I think creating a platform for collaborative creation in games is a great idea and something that is as-of-yet untapped in online communities. The main limiting factor is creating tools and platforms that outperform what already exists and creating an ecosystem for users where they have support to work on ideas that connect.
Links: http://github.com/teamduck/gametube http://www.team-duck.com/space/
Any plans for shaders?
I've also been toying with the idea of making a "Superpowers Minecraft modding edition" when Minecraft 1.9 comes out with an actual 3D model + skeletal animation format (everything has been hardcoded until now), so that could go hand-in-hand.
EDIT: Missed your question about shaders. We do plan to have a material editor and postprocessing shaders!
You can load up your games in a recent mobile browser or package them into APKs with the Intel XDK (we didn't have a chance to test iOS export yet since we don't have an iDevice but it should work). Performance-wise, I don't have hard numbers or anything. This is the Web, you won't get as much perf out of it as a native app, but the games we've built so far run great.
You don't need funding to make an open source solution to game development.
There is lots of competition out there for what you are doing. Have you compared your project to others just like it?
Best bet is to continue to work on the open source project, but if you need funding, do that separately, and don't confuse it with open source.
I think you are confusing open source with free as in beer. A lot of open source projects are funded by companies, being open source doesn't exclude being paid to develop the software.