speechd.. 
- implements the /dev/speech device -- all text written to this file will
  be spoken aloud.
- was created by Kyle R. Burton <mortis@voicenet.com>, Darxus@op.net, and
  Michael Matsumura <michael@limit.org>
- is currently maintained by Darxus@op.net
- is released under the GNU General Public License
- is written in Perl

The home page of this application is
http://SpeechIO.undef.net -- you should probably check
there for updates if you were not just there.

To subscribe to the SpeechIO mailing list, send email to 
speechio-subscribe@lists.undef.net 

You are also incouraged to join us in the #SpeechIO Internet Relay Chat
channel on EFNet.


Setup:

1) Verify that your sound card is functional with some other
   application.

2) Install either the Festival or rsynth speech synthesis
   packages (Festival with a high quality (16 bit) voice is
   strongly recommended), or the "morse" program (to hear
   morse code instead of speech).

2.1) Festival can be obtained from
     http://www.cstr.ed.ac.uk/projects/festival/ 

     I suggest using one of the high quality (16 bit) voices.  I
     prefer the British voice.

     Both Debian and RedHat packages are available from the main
     package archives of each distribution.

     If you are installing Festival in Debian, I suggest you try
     doing it via an "apt-get install festival festvox_rablpc16k"
     command.

2.2) rsynth can be obtained from
     ftp://ftp.cdrom.com/pub/linux/sunsite/apps/sound/speech/
     This thing is tiny, but the sound quality is not near that
     of Festival.

2.3) Alan Cox's "morse" program... there's a Debian package, I dunno
     where else it can be found.

3) Verify that your speech synthesis package is functional:

3.1) Festival:
     echo "this is a test" | festival --tts

3.2) rsynth:
     say "this is a test"

4) Run "make install".  This will probably do something like:
   Copy speechd to /usr/local/bin
   Copy catspeech to /usr/local/bin
   Copy speechdrc to /etc
   Copy speechd.sub to /etc
   Change ownership of speechd & catspeech to root
   Change permissions of speechd & catspeech to 755
   Create /dev/speech

5) If you would like to use rsynth instead of festival, edit
   /etc/speechdrc, comment out the '$synth = "festival";' line, and
   uncomment the '$synth = "rsynth";' line.  If you would like to use
   speechd with esd (the Enlightened Sound Daemon (and you have it
   installed)), uncomment the $use_esd line.

6) Execute speechd.  You should nolonger ever have to run speechd as root,
   as the creation of /dev/speech is handled by the install script.

7) Test speechd:
   echo "this is a test" > /dev/speech

   If you hear "this is a test" spoken aloud, it's working.

8) You will probably want to put speechd in a startup script
   somewhere in /etc/init.d


catspeech is a seperate application created for use by speech.irc (also at
http://SpeechIO.undef.net).  It simply takes everything from STDIN and
writes it to /dev/speech.  It is included with speechd incase it may be
useful in any other applications that make use of /dev/speech.


Current applications supporting /dev/speech:

* speech.irc - http://SpeechIO.undef.net
  full speech synthesis support for all Internet Relay Chat programs
  supporting the ircII scripting language.  Created by Darxus.

* slashes - http://SpeechIO.undef.net
  slashdot.org headline ticker w/ full speech synthesis support.  Created
  by Alex Shnitman (alexsh@linux.org.il), speech synthesis support added
  by Darxus.  I have sent my patch to the creator, but I do not believe he
  has included them yet.

* mouth.tcl - http://www.limit.org/speech.php3
  "A package for AOL's TiK that echos text to /dev/speech"
  Created by Michael Matsumura <michael@limit.org>.


Command line options:
Usage: speechd [-qQf] [-H <host>] [-P <port>] [-s <festival|rsynth>]
    or speechd [-h]
       -e  Use the Enlightened Sound Daemon
       -q  Quiet mode (Supresses STDOUT)
       -Q  Very quiet mode (Supresses STDOUT and STDERR)
       -f  Run in foreground (Don't daemonize)
       -H  Set the host to use (Default: localhost)
       -P  Set the port to use (Default: 1314)
       -s  Sets the speech synthesis package to use
       -h  This help text


If you have any updates/suggestions/requests, please subscribe to the
SpeechIO list by emailing speechio-subscribe@lists.undef.net 
