Initial commit.

1 files changed, 64 insertions, 0 deletions

diff --git a/README.md b/README.md
new file mode 100644
index 0000000..faecfa0
--- /dev/null
+++ b/README.md
@@ -0,0 +1,64 @@
+# Speak
+## SPO256-AL2 based Text-to-Speech Network Services
+
+http://ibiblio.org/pub/linux/apps/sound/speech/speak-1.0.tar.gz
+
+This package comprises a new stand-alone release of the standard 'text to 
+speech' interface service and tools as first covered in the SPO article 
+appearing in Linux Journal, January 1997.  Generic support for using the 
+SPO256-AL2 based board, and potentially other alternate network accessible 
+text-to-speech resources, is now more cleanly supported in a portable manner
+that should allow use under virtually any UNIX'oid OS.  This release 
+supersedes the initial WorldVU based source release originally described
+in Linux Journal.
+
+This new release includes a new and portable SPO256 text-to-speech server, 
+some glue scripts for init.d in Linux to use the SPO as a normal system
+'service', and many of the standard utilities re-written in a more 
+generic form.  The new server should do a better job of parsing textual
+documents into natural sounding speech as new rules for handling common 
+abbreviations and to improve word pacing have been added.
+
+The source for the SPO256-AL2 server is now more modular and more fully 
+documented.  The utilities now include standard man pages.  The server 
+now supports alternate forms of operation for use in client scripts.
+
+If you do not have the Linux Journal article in question, the SPO256 is
+essentially a serial based text-to-speech board of limited utility.  The
+SPO256-AL2 (Vocoder) chip was originally used in many products, including
+the Matel speak-and-spell(tm) toy, and offers limited text-to-speech
+translation services.  The SPO256 server augments the limited translation
+capabilities of the SPO256 chipset and offers a network accessible resource
+for text-to-speech notifications, such as system alerts, user logins, etc.
+
+This new server assumes a standard TCP 'service' has been defined in
+/etc/services.  You should edit the /etc/services file to include an entry
+like the following (though a different port number may be used):
+
+speak	800/tcp		# SPO256 text-to-speech network resource
+
+You will also need to copy speak.conf to /etc and modify it as needed.  
+
+While the SPO256-AL2 is the primary hardware interface supported in this 
+package, other alternate text-to-speech devices may be added later that 
+will also use the "speak" port service.  In particular, a re-implementation 
+of rsynth is being considered.  Since the utilities talk to a standard
+tcp port interface, they remain unchanged regardless of the underlying
+hardware used to produce text-to-speech output.
+
+To build the software package, first run ./config.  Config will genorate
+the Makefiles.  Config will then start 'make'.  The configuration system
+will attempt to identify the OS it is running under, compiler options, etc.
+This may work under most UNIX operating systems, so the SPO package should
+not be assumed to be a Linux-only utility.  Once make completes successfully, 
+perform a 'make install'.  This will update the /etc/services file and install 
+the application binaries.  If you have difficulty with the new source 
+configuration tool, please review build/config.doc.
+
+Please examine the speak.conf and speak.init files, and relocate them as
+needed.  Further documentation may be found within the supplied man pages.
+Comments may be addressed to David Sugar <dyfet@tycho.com>.  The SPO256-
+AL2 Text-to-Speech board ("Computalker") may be ordered through B&G Micro,
+P.O. Box 280298, Dallas, TX 75228 (214) 271-5546.  Linux Journal may be
+found http://www.ssc.com/lj.
+