1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
|
SPO256-AL2 based Text-to-Speech Network Services Version 0.2
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.
|
