Roadmap
Planned features and improvements for the next major release:
-
Support Windows Internet Explorer versions 8 and 9(issue #12, r21; version 8 won't be supported. Please notify me, if your opinion is that IE8 support is important.) -
Replace "for/in" by "for" when traversing arrays(issue #13, r24) - Replace b2Vec2 by WebGL arrays internally (Fall back if WebGL is not available)
- Reduce function calls
- Reduce references
- Improve minification by evaluating "private" and "public" modificators
- Keep documentation comments in the non-minified source code
- Replace properties by methods
- Provide an own API-documentation
There are some reasons why I don't like the idea of converting Box2dFlash using a compiler anymore. Box2dFlash shows no activities since I started writing my converter. Their version still is in an alpha-state. The main reason why I decided to write a converter was to easily keep my port up to date. After finishing my work and using the port I realized some bad things about Box2dFlash. E.g. debug drawing is implemented in a bad way. Parts of the drawing code are defined in b2World and b2Controller, whereas Box2D only provides debug drawing in the external test-bed. Furthermore, they are iterating through arrays using for/in at many places which leads to performance problems.
So, why not creating a new port by hand?
I considered to do this, but there are some problems with this idea as well. Javascript doesn't support types and modificators like "public" and "private". I think that these informations play an important role for optimizations. For instance, I'm able to define a namespace "b2Inline" and make the converter processing inline functions. In Javascript you have to copy-and-paste code if you want to inline functions in order to improve performance.
So, what should I do, instead?
I have no concrete plans, but if I write a port by hand, then it will be a C++ to ActionScript port. As for the Javascript version, I'll use my converter again.