2005-05-09 Willie Walker <william.walker@sun.com>

	* configure.in, README, NEWS: Updated for 0.3.7 version.
	* drivers/freetts/java/org/GNOME/Speech/Makefile.am:
	commented out SOURCES line in Makefile as the EXTRA_DIST line
	appears to manage the inclusion of the java source in the dist.

2005-05-09 Bill Haneman <bill.haneman@sun.com>

	* drivers/freetts/freetts-synthesis-driver.in: Use
	@datadir@ for the user_addenda path.

	* drivers/freetts/Makefile.am: Use $datadir to install
	the pronunciation addenda.

	* drivers/freetts/java/org/GNOME/Speech/Makefile.am:
	Reverted change to Makefile, to reinstate inclusion of java source
	in the dist.

2005-05-06 Willie Walker <william.walker@sun.com>

	* drivers/freetts/user_addenda.txt: New pronunciation addenda
        for the extended ISO8859-1 extended block as well as the
        unicode currency block.  Fix for bug 163805.

	* drivers/freetts/freetts-synthesis-driver.in: Add support
        for the pronunciation addenda.  Also add additional options 
        to speed cancel time for streaming audio player.

	* drivers/freetts/Makefile.am: Add support to install the
        pronunciation addenda.

	* drivers/freetts/java/org/GNOME/Speech/FreeTTSSynthesisDriver.java:
        Fix for bug 301086.  Also modified Makefile.am to properly generate
        a Makefile that works.

2005-05-06 Bill Haneman <bill.haneman@sun.com>

	* tests/test-speech.c:
	(select_server): Correct off-by-one error in select-server
	logic - if you selected a server that was "one greater" than the
	actual number of servers available, test-speech could crash.
	
2005-05-06 Bill Haneman <bill.haneman@sun.com>

	Patch from Fernando Herrera for bug 141516.
	
	* drivers/festival/festivalsynthesisdriver.c:
	(get_voice_list_with_spec): New.
	(festival_getVoices, festival_getAllVoices,
	festival_createSpeaker): Use get_voice_list_with_spec instead 
	of get_voice_list/prune_voice_list pair.
	(festival_driverInit): Get the voice list from the festival
	service, and wait for the reply in festival_response_sock to fill
	the voice list.
	
	* drivers/festival/festivalspeaker.c:
	(festival_speaker_new): Set the voice from the voiceinfo name.
	
2005-04-08 Bill Haneman <bill.haneman@sun.com>

	* drivers/festival/festivalsynthesisdriver.c:
	(festival_start): Set io channel encoding to ISO-8859-1, to 
	pass non-ascii latin chars correctly.  Bug #141516.

2005-03-22 Oana Serb <os@baum.ro>

	* drivers/dectalk/dectalksynthesisdriver.c:
	(dectalk_synthesis_driver_add_index):
	(dectalk_callback):
	(dectalk_driverInit):
	(dectalk_synthesis_driver_say):
	When a marker comes, all the markers before it in the waiting queue are 
	considered as received. This way all the markers are presented by the driver.
	When DECtalk has finished to speak, the waiting queue is flushed.
	(dectalk_synthesis_driver_say):
	Skip leading spaces before the words.
	(bug fix #167810)