I am also interested in this package. Unfortunately, this synthesizer is known to segfault while deallocating memory when compiled under Linux x86-64, whereas this does not occur on 32-bit Arm and x86 architectures. There is a need for someone with considerable background and experience to try to track down the cause.
Based on the code which I looked at, it appears that large parts of the Lingware files are probably derived from data by means of machine learning algorithms. Consequently, there may be no "source" in any conventional sense for this material. There is a press release indicating that tools for building new languages and voices were part of the contribution by SVOX to Android, but they haven't appeared publicly as yet. >From a technical point of view, based on my (admittedly limited) knowledge of tts, this is a very serious technological contribution to free software. I am interested in it for accessibility reasons, in particular its potential to be used with the emacspeak, gnome-orca, speakup and yasr packages, given the necessary integration efforts. -- To UNSUBSCRIBE, email to debian-wnpp-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
