Re: Process speaker not running?
That was an excellent overview. Yea, since those first inquiries, I
picked up the idea of using dtk_tcl from some of Jim's later work on the
DoubleTalk driver. I also like your other comments concerning shell and
redirection; there's some good hints there. I will probably save them and
btw, they might be good hints to stick in your emacspeak documentation
even though they don't apply directly to emacs, they are quite necessary
as backup speech and before emacs and emacspeak get to taking well.
Also, I am making good progress now with it; I've even used it a bit this
On Mon, 14 Apr 1997, T. V. Raman wrote:
> My reply to this last week was terse because I was away at WWW97.
> Let me try and explain what I think is happening in your istuation; it's
> basically the result of
> misunderstanding how the various emacspeak environment variables work.
> Here is a summary:
> Emacspeak uses three environment variables.
> a) DTK_TCL --pathname to a tcl interpreter
> (or in general an interpreter that will be called with a synthesizer specific
> b) DTK_PROGRAM --the name of the device specific script e.g. dtk-exp
> c) DTK_PORT --port where the speech box is connected e.g. /dev/ttyS0
> Upon startup, Emacspeak launches the driver by executing
> DTK_TCL with DTK_PROGRAM as the first (and only) command line argument.
> The above is expected to examine environment variable DTK_PORT to decide where
> the speech device is connected; if that environment variable is not set it
> picks a suitable platform default,
> e.g. /dev/ttyS0 on linux or /dev/ttya on sparcs.
> With this summary, it's probably clear why things went wrong for you.
> 2) Killing a hung emacs, and getting minimal spoken feedback from a shell:
> This is not really an Emacspeak question/answer, but you may find what I use
> useful for your own situation.
> 1) In your /usr/local/bin create a simple
> shell script called speak that sends its arguments to the speech box ie a
> talking version of the shell's echo.
> Assuming you use bash:
> b) In your .profile, add a command at the end to echo a message to the speech
> box indicating succesful login.
> c) bash lets you specify a command to execute every time a shell prompt is
> I have a sound card so I typically set bash to play a sound file like so:
> EXPORT PROMPT_COMMAND="play ding.au"
> (replace play and ding.au with appropriate pathnames for your setup.
> If you dont have a sound card, set PROMPT_COMMAND to something like
> echo "go. " > /dev/ttyS0
> d) Now, if you are stuck, you can go to another virtual console, and get
> sufficient feedback to do what you need in terms of killing a stuck emacs, or
> I typically have shell scripts called "spot" and "slay"
> that accept a process name, e.g. emacs and display the pid --
> so to locate an emacs session I would do
> bash: spot emacs
> and to get the output spoken
> speak `spot emacs`
> note that the above backquote trick is really useful when you have no other
> e.g. you can tell which directory you are in by typing
> speak `pwd`
> Finally I also typically have a script called slay that nukes processes by
> name so I can do
> slay emacs
> once I have detected the offending process with spot do
> kill pid
> Hope this helps alleviate some of your frustration--
> Steve Holmes writes:
> > Why am I getting this? I am attempting unsuccessfully to use emacspeak
> > 5.0 on my linux box. I do not own a dectalk so am trying to build a
> > driver for the Speakout from GW Micro. When I test it using TCL or when I
> > attempt to use a C program modled after Jim Van Zant's d-exp program, I
> > get reasonable results; more work is yet required but at least they more
> > less work. When I try to load up emacspeak, I keep getting the message
> > "Process speaker not running". After that, I cannot get out of emacs no
> > matter what I do. I usually end up rebooting the system. Jim told me
> > recently I could do a c-e followed by a c-s but that doesn't appear to do
> > me any good subsequent key strokes give me the same damn old thing. I
> > have set DTK_PROGRAM to my replacement module for the speakout correctly
> > in the environment but still to no avail. I am running out of luck and
> > patience with this bloody thing. It sounds so simple but I can't get any
> > thing worthwhile out of it thus far. I suppose if I had a dectalk all
> > this would be moot. We need compatability to other synthesisers and I
> > want to do my part but not having good luck so far.
> > Can anyone help me out? Jim Van Zant has been a good help so far, but I
> > think I might need more:).
> > Thanks in advance,
> > <Steve> Holmes
> Best Regards,
> Adobe Systems Tel: 1 (408) 536 3945 (W14-129)
> Advanced Technology Group Fax: 1 (408) 537 4042
> (W14 129) 345 Park Avenue Email: firstname.lastname@example.org
> San Jose , CA 95110 -2704 Email: email@example.com
> http://labrador.corp.adobe.com/~raman/raman.html (Adobe Internal)
> http://www.cs.cornell.edu/Info/People/raman/raman.html (Cornell)
> Disclaimer: The opinions expressed are my own and in no way should be taken
> as representative of my employer, Adobe Systems Inc.