[sword-devel] Speech synthesis voices?

[Karl Kleinpaste]

I too have worked with a blind sysadmin, a fine fellow named Henning
with whom Martin Grüner is acquainted.  A very competent fellow who also
uses a Braille reader.

But we're talking in circles by now.  Xiphos provides Read Aloud because
there wasn't system-level support for it when it was implemented, and it
has proven disturbingly popular.  (For something that started out
somewhat as a joke, and as a proof-of-concept for myself when I was
getting my feet wet in the code, it's really quite odd.)  If the user
has system-level TTS support, then he can ignore Xiphos' internal
support and use that; if that's key sequences to emulate mouse sweeps
and clicks, fine, Xiphos doesn't have to care.  As it is, Xiphos
provides ongoing verse reading if Read Aloud is simply turned on, and
the user can sweep-and-select a region for reading on demand.  I don't
see that anything more than this is in order.  Xiphos doesn't provide
voice selection, or any other aspect of Festival configuration, because
that was considered underneath the level at which the user interacts
with Xiphos.