[Emacspeak] Re: Introducing myself and a request

[Devin Prater <r.d.t.prater@xxxxxxxxxxx>]

I'm using Emacspeak under Fedora 34, which comes with Pipewire by default. The only thing I'm missing is the tones from the TTS engine for things like indentation reporting and such.

Devin Prater

r.d.t.prater@xxxxxxxxxxx

gemini://tilde.pink/~devinprater/

On Thu, Jul 22, 2021 at 7:32 PM Tim Cross via Emacspeak <emacspeak@xxxxxxxxxxx> wrote:

Hi,

Welcome to the list. I'll start with a couple of questions which might help
provide more accurate information when you have more questions.

1. Have you used Emacs before? Emacs is a bit different to most editors and can
   be a little daunting at first. Understanding what level of familiarity you
   have with the editor can help with deciding what level of detail/information
   you need.
   
2. Are you running under a GUI (i.e. X or Wayland) or just running inside a
   Linux console. If a GUI, which window manager or desktop environment?
   
3. Are you using any other assistive technology, like Orca, speakup, etc?
   

I"ll start by saying I have no experience with pipeWire, so what follows has a
lot of guesswork and assumptions in it.

The only part of Emacspeak which needs to know anything about the underlying
audio infrastructure is the text-to-speech synthesizer and playing of auditory
icons via some 'generic' play program, like sox or pulseAudio paplay or Alsa aplay
programs. For the TTS synthesizer, if you can get espeak/espeak-ng to work, your
80% there. If you have some program on your system which can play *.wav files,
then your 90% there - the rest will just be a little configuration tweaking.

Of course, how well it works with pipeWire is another question. A lot will
depend on the latency of the system. Early pulseAudio versions had some problems
in this area, but I've been using it with Emacspeak now for years without any
problems at all. As pipeWire is still fairly immature, you are likely to run
into similar issues initially.

Emacspeak relies on external programs for speech and playing of auditory icons.
It doesn't really need to know about the underlying sound infrastructure.
Provided you can get these external programs working, you have a good chance you
will be able to get Emacspeak working. However, you may need to do some
additional configuration - especially for auditory icons, in the sense of
tweaking the variables in Emacspeak which control what programs and arguments
are used to play the icons.

Here is what I would do -

1. Verify you can use espeak from the command line. This is critical. If you cannot get espeak to work from a CLI, you are sunk.
   
2. If espeak works from the command line, you next need to verify you can build
   the espeak TCL interface library in servers/native-espeak by switching into
   that directory and running make. This assumes you have all the necessary
   dependencies installed. The default setup of the Makefile is configured to
   work well under Debian/Ubuntu and may need a little tweaking for Arch. The
   main problem people run into here is not having the necessary build
   dependencies installed or in the paths searched when compiling - this
   includes the espeak dev libraries and the Tcl dev libraries.
   
3. If you are able to successfully build the espeak server, you then need to
   verify it is working by changing into the servers directory and running
   ./espeak. You should here the server speak, saying that it is running and
   be left at a Tcl prompt where you can enter some commands to queue some text
   to speak and then speak it. You can queue some text for speaking by doing q {Hello world} [enter] and then d [enter] and you should here the speech
   "Hello world".
   

If you get to this point, you have a mostly working Eaacspeak setup. The
next thing you will need to do is set various variables to play auditory icons.
Personally, I just use the sox program to do this. However, both alsa and pulse
audio have CLI programs to play basic *.wav files. I expect pipeWire probably
has something equivalent - you will just need to configure Emacspeak to use it.
Note that Emacspeak works fine without auditory icons - they are an optional
enhancement which provide some valuable feedback, but are not essential.

Good luck!

Regards,

Tim

â??
Tim Cross

For gor sake stop laughing, this is serious!

From: Cisco Tissera via Emacspeak
Subject: [Emacspeak] Introducing myself and a request
To: emacspeak@xxxxxxxxxxx
Date: Fri, 23 Jul 2021 06:29:03 +1000

Hello everyone,

My name is Francisco, and I am a highschool student who started getting into Linux a few months ago.

I started with Arch, switched to Ubuntu, Fedora, and then returned to Arch Linux.

I would like to get up and running with emacspeak so I can see what it can and can't do, and see if it suits my needs, however, I cannot, and here is why: I am using pipewire on arch linux as of now.

When I wrote on the blinux mailing list, linux for blind, I got told that sadly emacspeak does not work with pipewire yet.

<I do not have any intention of removing pipewire and installing pulseaudio, because that would disrupt my workflow for far too long.

As of now I am using arch linux with gnome 40 and orca master.

Now, after all that has been said, I'd like to ask if anything could be done about emacspeak to make it work with pipewire.

I have no coding knowledge so I won't be able to help in the development, if any will be done, but I will surely test out the git package in the aUR or whatever if you so ask of me.

Best regards.

Francisco.

_______________________________________________
Emacspeak mailing list -- emacspeak@xxxxxxxxxxx
To unsubscribe send an email to emacspeak-leave@xxxxxxxxxxx