ANNOUNCE: Unofficial 0.6.8~rc1 release of speech-dispatcher.

[Luke Yelavich]

Hi all
I have decided to go ahead and release an unofficial 0.6.8~rc1 tarball of 
speech-dispatcher. There have been a fair few changes since 0.6.7 was released, 
some of which are bug fixes, and some new features, as well as some code 
cleanup. There may be a few commits I missed out from this shortlog for changes 
since 0.6.7, but I am pretty sure they were bug fixes anyway.

I must thank the various folks from the community who have sent me many a patch 
this last few weeks and months, the most significant of which, is the rewriting 
of the pulseaudio driver to give better audio output performance. I think this 
output code will still need tweaking, but it is already quite usable, and I 
myself use it from day to day.

Please note that this is not an official release, i.e the freebsoft 
project/Brailcom haven't given their endorcement/approval for this release, or 
many changes that are encorporated in this release.

You can download the tarball from 
http://packages.themuso.com/speech-dispatcher-0.6.8~unofficial~rc1.tar.gz

Please report back with any issues you have with the tarball, and of course, 
patches are welcome.

also be on the lookout for another email from me, outlining a roadmap that I 
personally have for speech-dispatcher, which will focus on future additions and 
functionality changes that I'd like to see encorporated into speech-dispatcher.

Here is the shortlog:

Andrei Kholodnyi (3):
      don't install session files while --disable-session-integration
      removed 'struct Func', typedefs for plugin functions
      remove automatically generated config/speechd.conf

Gilles Casse (3):
      add a few curly braces
      modules, config - ibmtts: list of installed languages
      modules - ibmtts: fix the maximum number of possible languages

Hynek Hanke (34):
      (Tests.audio_try_play): Use proper os.path.join for constructing path to 
the testing wav file.
      *** empty log message ***
      *** empty log message ***
      Prevent bash expansion of $DATA.
      (module_speak): Added mutex to protect espeak_state != IDLE
      * src/modules/module_main.c (PROCESS_CMD_NRP): Mutex unlock moved to 
correct place.
      Respect ${DESTDIR}.
      *** empty log message ***
      (_alsa_close): Protect stop pipes with mutex. (alsa_stop): Handle 
correctly write return value.
      (alsa_pipe_mutex): New mutex.
      (speak): Added debugging messages.
      *** empty log message ***
      *** empty log message ***
      (Communication Protocol for Output Modules): Synchronization of index 
marks explained in more detail.
      *** empty log message ***
      (maintainer-clean): Maintainer-clean does clean.
      *** empty log message ***
      Maintainer-clean deletes configure.
      (clientconforigdir): New location for .desktop file.
      (EXTRA_DIST): Include speechd.desktop in distro.
      *** empty log message ***
      (all): Handle speechd.desktop.
      (Tests.diagnostics.decide_to_test): Set port in ~/.profile and place 
speechd.desktop in ~/.config/autostart
      (EXTRA_DIST): Include speechd.desktop in distro.
      New file.
      Create conf/desktop/ directory for installation.
      *** empty log message ***
      *** empty log message ***
      Testing CVS -- blank line.
      (_pulse_open): Added PA_STREAM_ADJUST_LATENCY
      (load_default_global_set_options): Pulse Audio parameters modified 
according to Lennart's instructions.
      *** empty log message ***
      * src/audio/pulse.c (_pulse_open): Typo.
      *** empty log message ***

Luke Yelavich (28):
      server - Send log level configuration setting to speech modules
      modules - Receive and set log level from speechd server
      audio/alsa, audio/oss - Take the log level into account before logging a 
message.
      audio/alsa, audio/oss - Assign all log messages a log level
      audio, modules - Set libspdaudio's internal log_level variable
      modules - Support graceful audio fallback
      modules - Fix compiler warning
      config - Set the audio output to be "pulse,alsa" by default
      server - More tightly integrate the server into a user session
      C API - Remove function definitions for functions that do not exist
      server - Add support for retrieving current voice and current output 
module over SSIP
      C API - Allow clients to define their own version of SPD_DEBUG
      config - Load system wide and user client configuration files
      Revert "C API - Remove function definitions for functions that do not 
exist"
      C API - Clean up spd_set_synthesis_voice function definitions
      server - Only check for $HOME/.speech-dispatcher if the current user's 
home directory starts with /home
      Revert "server - Only check for $HOME/.speech-dispatcher if the current 
user's home directory starts with /home"
      audio/pulse - Attempt to connect to pulseaudio server when determining 
audio output
      modules - Fix audio fallback logic
      C API - make spd_open threadsafe
      config - Remove quotes from references to the speech-dispatcher 
configuration directory
      config - Reflect default pulse configuration choices
      audio/pulse - Clean up properly if an attempt to connect to the 
pulseaudio server fails
      session - Add auto-spawn functionality as part of session integration
      Add .gitignore files to prevent autotools files from being considered
      C API - Only spawn the speech-dispatcher server if session integratino is 
enabled
      audio - Add libao output support
      Prepare 0.6.8~unofficial~rc1 release

Milan Zamazal (2):
      Fix socket leak on connection error in libspeechd
      ./ChangeLog

Rui Batista (6):
      Audio - Rewriten pulseaudio driver using the pulse-simple library
      Audio - Converted previous patch to read some parameters from 
configuration
      Audio - Changed fdset.h to reflect the new pulseaudio backend settings
      Audio - pulse.c: send NULL when server string is default
      Configuration - Changed default config file to reflect new pulseaudio 
settings
      audio - Changed comments for pulse.c to reflect authors

William Hubbs (1):
      fixed getline prototype

Willie Walker (1):
      OpenSolaris fixes

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: 
<http://lists.freebsoft.org/pipermail/speechd/attachments/20100202/056e1091/attachment.pgp>