[g-a-devel]GNOME Speech 0.2.4 Release
- From: Marc Mulcahy <Marc Mulcahy Sun COM>
- To: GNOME Accessibility Development List <gnome-accessibility-devel gnome org>
- Cc: Marc Mulcahy <Marc Mulcahy Sun COM>
- Subject: [g-a-devel]GNOME Speech 0.2.4 Release
- Date: Wed, 30 Jul 2003 00:28:27 -0600 (MDT)
Hi All,
GNOME Speech 0.2.4 is a major rewrite of the included drivers. I'm
happy to say that this version of GNOME Speech includes a driver for
Cepstral's Theta text-to-speech engine. the version in 0.2.4 does not
work with the current download available from Cepstral's store, but we
hope to have this issue resolved soon.
Here's a quick look at what's new:
* The .server files for all drivers which support speech callbacks now
include the GNOME_Speech_SpeechCallback:0.2 in the repo_ids field.
* The SynthesisDriver object is the central point where clients can query
for information about, and create speakers with, specific voices
provided
by an engine. therefore, for consistancy's sake, I have removed the
"voice" parameter from all speakers.
* Added Theta driver which supports word-level speech
callbacks. * Added Speech Interrupt test and wait test to test-speech.
* Rewrote the Fstival driver. This rewrite includes:
* One instance of Festival is now shared by all speakers-- no
longer is Festival spawned for each speaker which is created.
* The version number returned by getSynthesizerVersion is obtained
from the Festival binary.
* driverInit is implemented and returns FALSE if the Festival
binary can not be found.
* Rewrote Viavoice Driver. This rewrite includes:
* All speakers now use one instance of Viavoice.
* Added speech progress notifications after each word which is
spoken.
* AS mentioned above, removed the voice parameter from Viavoice
speakers.
* Rewrote the DECTalk driver. This rewrite includes:
* All speakers share one instance of DECTalk software.
* As mentioned above, removed the voice parameter from DECTalk
speakers.
* Added speech status callbacks after each word spoken.
* Rewrote the FreeTTS Driver. This rewrite includes:
* All speakers share one FreeTTS Voice. This eliminates the bug
which caused the FreeTTS driver to run out of memory when
multiple
speakers were created.
* Specifying the path to DECTalk now works.
* Cleaned up IDL build system
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]