(top)
gnome-speech 0.4.10:

* Fix for bug 409573 to have eSpeak driver also look for LANG when
  determining the default language. (Gilles Casse)
	
gnome-speech 0.4.9:

* Fix for bug 398916 to help select the Viavoice Chinese voices. Also
  fixed a memory leak with the driver. (Gilles Casse)

* Fix for bug 399512 to handle config parameters better. (Leonardo Boshell)

* Fix for bug 403763 to include math library when building driver
  for Swift. (Willie Walker)

gnome-speech 0.4.8:

* Brand new driver for eSpeak (http://espeak.sourceforge.net).
Many many thanks to Gilles Casse for his contribution here!!!

* Fix for bug 321658 to allow ibmtts is unable to select/change voice
of a different language.  Many many thanks to Gilles Casse for his
contribution here!!!

gnome-speech 0.4.7:

* Add support for Cepstral Swift driver.  There are known problems
with this support.  In particular, there are audio artifacts 
(chirps and pops) when speech is interrupted.  We have a support
request into Cepstral for help with this: Support Ticket #6297.

gnome-speech 0.4.6:

* RFE 350389: add in Loquendo driver from ONCE (Thanks Juan Ramon
  Jimenez from ONCE!).

gnome-speech 0.4.5:

* Fix for bug 353322: fix crash/hang in DECtalk synthesis driver.

gnome-speech 0.4.4:

* Fix for bug 326818: escape the \ character so it is not interpreted
as a line continuation character by the festival shell.

* Fix for bug 352165: use "-f" flag to rm files when doing a clean.

gnome-speech 0.4.3:

* Fix for bug 349730: do not abort festival synthesis driver when it
cannot convert a string to the encoding expected by the voice being
used by festival.

gnome-speech 0.4.2:

* Fix for bug 347357: convert UTF-8 to ISO8859-1 before sending text to
  DECtalk engine

* Fix for bug 348240: convert UTF-8 to ISO8859-1 before sending text to
  IBMTTS (a.k.a., ViaVoice) engine (thanks Javier!)

* Fix for bug 347124: remove LT_VERSION_INFO since it is only available
  for libraries (thanks dmacks!)

gnome-speech 0.4.1:

* Partial fix for bug 144457: add "punctuation mode" parameter to DECtalk
  synthesis driver. The supported modes are as follows:

    0 = none
    1 = some
    2 = most
    3 = all

* Fix for bug 341405: don't core dump if festival is not installed on
  machine.
	
* Fix for bug 141516: allow character encoding to be dynamically
  determined from the voice.  Credit to Milan Zamazal as well for the
  creation of the "coding" attribute convention for festival voices.
	
* Fix for bug 341990: eliminate compilation warning in festival synthesis
  driver.

----------
gnome-speech 0.4.0:

* Fix for bug 321216: support UTF-8 encoding (Chaitanya Kamisetty)

* Fix for bug 337151: quit when the festival process dies (Remus Draica)

* Fix for bug 341744: turn phoneme mode off in dectalk driver (Willie Walker)

* Fix for bug 332250: dectalk driver compilation warnings (Willie Walker)

* Fix for bug 332247: dectalk driver can hang on exit (Willie Walker)

* Fix for bug 329444: add speech dispatcher driver (Hynek Hanke)

----------
gnome-speech 0.3.9:

* Fix for bug 319421: ibmtts hardcoded to en_US locale (Bill Abt)

* For for bug 319875: Add eciInsertIndex to viavoice driver (Bill Abt)

* Fix for bug 320013: Add more parameters to ibmtts driver (Bill Abt)

* Fix for build problem on Solaris (Rich Burridge)

* Fix for bug 142354: FreeTTS driver exits when refcount goes to 0 instead
  of living forever (Willie Walker)

* Fix for bug 313652: FreeTTS build fix (Bernard Leak)

* Fix for bug 319980: test-speech now uses selected gender instead of
  hardcoding to male (Willie Walker)

* Fix for bug 314736: Cygwin ports patch (Cygwin Ports maintainer)

* Fix for bug 314840: Corrected the use of the rate and pitch
  parameters for ibmtts driver so that they are absolute and not
  relative values (IBM)

* Provide different method for finding the Java virtual machine for
  freetts-synthesis-driver (Willie Walker)

* Fix for bug 319168: FreeTTS streaming audio player could not be used
  (Willie Walker)

----------
gnome-speech 0.3.8:

* Fix for bug 305489: Festival driver in gnome-speech-0.3.7 seg faults
  (Willie Walker)

* Fix for bug 308468: #include <sys/socket.h>, and do not declare 
  variable after statement (Behdad Esfahbod)

* FreeTTS driver now says "FreeTTS synthesis driver running..." instead
  of saying it is *not* running (Willie Walker)

----------
gnome-speech 0.3.7:

* FreeTTS driver now includes and uses a user_addenda.txt pronunciation
  lexicon addenda to support characters from the extended ISO8859-1 character
  set as well as the unicode currency block - bug 163805 (Willie Walker)

* Automatic detection of voices for festival driver - part of bug 141516
  (Fernando Herrera)

* Fixes for bug 301086 (Willie Walker), bug 141516 (Bill Haneman and Fernando
  Herrera), bug 169805 (Bill Haneman), and bug 167810 (Oana Serb)

----------
gnome-speech 0.3.6:

* Festival driver now supports callbacks and end-of-speech markers - thanks to
  Oana Serb of BAUM Retec.

* FreeTTS driver is less intolerant of nonstandard JAVA environments at 
  build time (Callum McKenzie).

* IBM ViaVoice driver updated to work with new Linux ViaVoice runtime
  (Marc Mulcahy)

* Critical fixes for FreeTTS 1.2 driver which could cause speech to fail
  (Marc)

gnome-speech 0.3.5:

* Improved locale matching when creating a speaker -- thanks billh

----------
gnome-speech 0.3.4:

* Better LOCALE matching for getVoices queries,
  for FreeTTS, Theta, and DecTalk drivers.
* CLASSPATH fix for FreeTTS driver.
* Use volume range of 0 - 100% instead of 0 - 1.0, for
  consistency, in FreeTTS Speakers.
* FreeTTS driver getVoices() command now works (formerly, 
  it seems only getAllVoices() worked for FreeTTS.)
* Better error reporting in test-speech.
* test-speech now asks for gender and locale choices
  before presenting voice list.

gnome-speech 0.3.3:
* FreeTTS driver now requires FreeTTS 1.2beta.

----------

gnome-speech 0.3.2:

* Drivers now take into account the language specified in getVoices
  and the createSpeaker calls.

-----------

gnome-speech 0.3.1:

* Fix some driver strings to fix activation errors.

-----------

gnome-speech 0.3.0:

* Updated VoiceInfo structure such that language is a string rather
  than an int.

-----------

gnome-speech 0.2.8:

* Fixed FreeTTS build issues
* Fixed compiler warnings

-----------

gnome-speech 0.2.7:

* Fixed libtool versioning
* Enhanced build checks
* Fixed bad escaping of strings in the Festival driver.
* Fixed refcounting issue in the FreTTS driver.

-----------

gnome-speech 0.2.6:

* More build fixes.

-----------

gnome-speech 0.2.5:

* Fix build.

-----------

What's new in gnome-speech 0.2.4:

* Driver added for Cepstral Theta.

* All existing drivers were rewritten to share one instance of the TTS
  engine among all GNOME Speech Speaker objects created by the driver.
  Drivers which were rewritten include:
	* IBM Viavoice
	* Fonix DECTalk
	* Festival
	* FreeTTS

* All drivers were hacked to shutdown the server executeable when the
  SynthesisDriver object is destroyed.

* Memory leaks were fixed in test-speech.  Also two more tests were
  added:
	* Interrupt test-- tests speech interruption
	* Wait test-- tests the driver's ability to wait for speech to
	  finish.

-----------

What's new in gnome-speech 0.2.1:

* Improvements to prevent installation of .server files for drivers that aren't built.

* Work with newer autotools.

* Added a new driver for Eloquence, DECTalk.

* Build fixes and bug fixes :-)


-----------
No news is good news.
