This looks promising, I've been working with the Emscripten[1] port of eSpeak and porting it to the browser and node.js [2] as a contributor (not author)

[1]: https://github.com/kripken/emscripten

[2]: https://github.com/christopherdebeer/speak.js

I'm curious how best one could participate/contribute to spec drafts like this, seems it's being steered by Google at the moment.

Is Google going to open-source their speech recognition engine? Without that, this feels like standards white-washing.

I assume, given the authors of the draft, that Chrome will have support for this soon?

Edit: Apparently, it already has it: http://news.ycombinator.com/item?id=3428329

Pretty exciting, particularly for doing command/control via speech recognition. I wish they would support BNF-style grammars the way Sphinx does.