[sword-devel] Speech synthesis voices?

[Karl Kleinpaste]

Greg Hellings <greg.hellings at gmail.com> writes:
> TTS never struck me as something that is properly handled by the
> application but should only be offered on the system level

Bear in mind that I implemented Read Aloud back in the days of Fedora 5
(2006).  I have no memory of there being _any_ system-level TTS support.
The fact that I found Festival available at all was an accident.

That said, how TTS is done "at the system level" is a pretty funny
concept.  That is, if Xiphos isn't doing Read Aloud directly, but the
user is expecting it "at the system level," what shall he expect to
happen?  Every time he navigates a verse, the entire Bible chapter, as
well as at least the commentary verse and possibly the entire commentary
chapter, is displayed.  Shall TTS read both Bible and commentary verse
in their entirety?  NETnote Ecc 1:1 is 8Kbytes of text, and takes nearly
12 minutes to speak.  How does one, at the system level, tell the system
to read one pane and not another, and within one pane, to read only a
certain slice of text?  I simply don't know.