bnjm / webgl-framework-comparison Goto Github PK

Trying to understand how "Proprietary Tools" is a weakness :)
Any other engine out there, either has no tools at all, or have own proprietary tools. And it is great, as those tools are tailored for engine precisely.
Other negative, is regarding community questions regarding editor or engine. It is actually 95% engine, and rarely related to editor.
Regarding one-off experiences, I actually strongly disagree. Tools and platform, are built with focus on long-term teamwork. Real-time collaboration in Editor, Publishing process, Dev Logs with timeline, Asynchronous assets loading, and many other features are made to allow long-term development of projects.
There are projects that been in development for months and months, and approaching years. TANX for example is constantly revisited, and now being worked by single developer constantly. And project evolves as being released and played all around the web. In the future will have monetisation with in-app purchases, which perfectly shows that this is not one-off experience. :)

So considering real weakness comparing to some other tools out there, I would say lack of node-based material editor, is a weakness. But none of WebGL engines have that either.
Or node-based logic editor - some people might find it as weakness. Goo at the same place has that, and has a big focus on enabling non-developers being able to create apps.
Some other weakness and strength at the same time, comparing to lets say three or babylon, is that everything is in a cloud, so it is different workflow and approach to development. Weakness here is lack of classic version control systems. But that will be addressed in a future.

Looking at strengths and weaknesses, they are a bit weird.
For example all engines have animation, but only goo mentioned that.
All engines have scene graph. Although threejs does not follow component system paradigm, while goo and playcanvas does.
Goo has node-based logic editor, without need to write code. Goo and blend4web is the only webgl engines out there with such capabilities.
Three does not have built-in physics.
Three comparing to all others is built by web guys, not 3D industry people. So concepts, API design and generally usability of three is more focused on web developers. While for example playcanvas is built with industry veterans with high focus on performance and understanding real-time rendering complexities, it tries to address very unique challenges of webgl platform.

Here is fair and good comparison with a bit in-details of many engines: https://playcanvas.com/project/337758/overview/comparison-of-webgl-3d-engines

As well, I think BabylonJs has very good place within those 4 leading tools.
Actually could even add blend4web to this list. Totalling with 5 decent webgl engines.
Big separation here is by the fact of which engines have tools and which don't.

It is really good to have such projects as this, comparing nicely different engines and tools :)

To compare really difference between engines, users would love to see close code wise examples, that implement same thing, without code-style differences.
So to judge more about API of engines, and concepts, rather being distracted by different code-style of three pieces of code.

Hello, i own a little framework that wraps three.js API and physics. What do you think about adding it to your list?
Github: https://github.com/WhitestormJS/whitestorm.js
Website/Docs: http://whitestormjs.xyz/

Some medium articles about developing a game using WhitestormJS:

• https://medium.com/whitestormjs-framework/developing-a-street-basketball-game-part-i-getting-workflow-ready-f4f6968e4d10#.fnotag75y
• https://medium.com/whitestormjs-framework/developing-a-street-basketball-game-part-ll-throw-a-ball-into-a-basket-261f123d6e9c#.anlueww4i
• https://medium.com/whitestormjs-framework/developing-a-street-basketball-game-part-lll-level-selecting-and-stats-b076747b13b3#.t8pc039jz