Acronym soup aside, Semantic Interpretation for Speech Recognition is now a Proposed Recommendation.
Here.
If you've ever developed speech applications (haven't we all), you're probably aware that the grammars are crucial. Whats more, the semantic interpretation pieces often need to be rewritten across platforms. And thats generally OK, until its time to switch or support more (voice) browsers.
So this could be good.
GRXML or
ABNF camp welcome.
And along with some alignment to web applications as we know them, à la
Reusable Dialog Components (RDC) tag library (at
Jakarta), speech application authoring might even start coming up everyone's alley.