Forwarding to the list to keep the discussion archived.
--- Begin Message ---
Hi

Sorry, we've changed Internet Service Provider, so things are up in the air but 
hopefully, you'll get this reply.

I started implementing your directions about speech support on the CD. 
If I understand correctly, they were:

Apply speakup kernel patch.
Set sane volumes by default.
Install a software synth (you recommended Espeak, but the default 
speech-dispatcher configuration file apparently wants to use Flite, so I 
installed Flite-1.3 with the patch from 
http://homepage.hispeed.ch/loehrer/flite_alsa.html).
Install speech-dispatcher (depends on the dotconf library).
Install speechd-up.

Wow! I didn't expect that.

After that, I am able to load the speakup_soft module, and the CD starts 
talking.
Great!

You mentioned the issue about messages like this one filling the screen:

Sat Sep 29 17:23:16 2007 [244176] ALSA: Start of playback on ALSA

This is actually a bug in speech-dispatcher, not ALSA - they 
unconditionally print debugging messages to stdout. See 
speech-dispatcher-0.6.4/src/audio/alsa.c (look for "MSG"). This can be 
sidestepped by redirecting stderr to /dev/null, or patched out.

I'm not sure what this means.  Does this send all stderr from every application 
to null?  Don't know what patched out means?

So, I have a working configuration (not committed yet, just testing 
locally), but also I have some questions.

1) What are the benefits of Festival, Flite and Espeak over each other?

Festival is big.  Flite is nice and small.  Espeak is small and easier on the 
ear.  Espeak is continually being updated, John the developer has spent a lot 
of time on the speakup list and implemented features that speakup users have 
requested.
2) Could you please mail me the configuration of speech-dispatcher that 
works for you? I don't want to ship unconfigured programs on the CD.

Will do under separate cover, shortly.  I do have a lot of catching up to do 
but will make it a priority.


3) What are the normal means of enabling speech output in other distros 
and LiveCDs? I.e., is there some string that a typical speakup user 
expects to be able to add to the kernel command line to get speakup with 
hardware or software synth? I ask  because I am afraid that that LFS 
LiveCD will "look strange" if I implement something custom.

There are a number of params that should be documented.  I guess the major 
issue is going to be what is the kernel image to be called?  Or are you 
sticking to 1 kernel image fits all?  For my disk, I kept it called linux and I 
had the default line as "linux speakup_synth=sftsyn" which would load the 
software synth so I didn't have to touch the boot prompt.  However, if I had my 
hardware synth attached and wanted to use it, I'd type: "linux 
speakup_synth=ltlk" and this would look for the Lighttalk on ttyS0.  If it was 
on ttyS1 I'd likely need to specify the port. 

As you can see with brltty and speakup, commands there's a large margin for 
error.

I think that the grml disk, they called the speakup kernel speakup, so the user 
just entered speakup and the speakup params.

I created my disk to install lfs on my Mac mini but it's not as easy as I 
thought. <tut tut>


4) Even with "draw_arrow=ON" appended to /etc/tin/tinrc, speakup doesn't 
announce the current group in Tin. How to configure Tin to be more 
friendly? Or should I drop Tin and use a patched Mutt (patch from 
http://mutt.org.ua/download/mutt-1.5.15/patch-1.5.15.vvv.nntp.2.gz)?
I'm sorry, I haven't any idea on this.  I use to use mutt but never heard of 
tin, nor recall any discussion on the speakup list.  When I rejoin my lists, 
I'll ask other users about tin.

I hope you get this post.

Take care, thanks!

Gena
-- 
Alexander E. Patrakov

__________ NOD32 2559 (20070929) Information __________


--- End Message ---
-- 
http://linuxfromscratch.org/mailman/listinfo/livecd
FAQ: http://www.linuxfromscratch.org/faq/
Unsubscribe: See the above information page

Reply via email to