I'm sorry y'all.

I'm sorry y'all. Our TTS engine will be frozen at 2007 pronounciation of Espeak. If we want true, true decoupled multilingual support. I wonder if it's worth maintaining two "flavors" of the driver, one which just uses existing Espeak phoneme data as built into NVDA's Espeak, and another that lets it live standalone to be useed for compiling and other projects. The fact is, this engine cannot become "free" until the phoneme data can live independently within it.