Hi I've been recently confronted to these questions for the tuxdroid project (which i intend to run on a NAS/Home router).
About text-to-speech, eSpeak seems really interesting for embedded devices, plus it's more or less the only 100% free sotfware of this category (others, like festival / flite * for embedded *, are free, but not their voice models..). http://www2.tux-is-alive.com/wiki/Text-to-speech As for in-app integration: http://www.speechio.org/ About speech recognition, i'd say that it's today very hard if not impossible to get a working dication feature. The best option may be to skip it, and concentrate on "command launching": more basic pattern-based recognition. A great option (if we manage to make it run) is CVoiceControl, but it needs maintainers: the project is down. http://www.kiecza.net/daniel/linux/ See http://www2.tux-is-alive.com/wiki/Speech_recognition for our preliminary evaluations and resources/links _______________________________________________ OpenMoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community