On Wed, 2008-04-23 at 23:41 -0500, William Hubbs wrote:

> Speechd-up is still the way to link software speech to speakup.
> Just do an emerge speechd-up to install the packages.  The default
> synthesizer is espeak, and it and speech dispatcher will be installed as
> dependencies of speechd-up.  Once you have everything emerged, add
> speech-dispatcher and speechd-up to your default runlevel.
> 
> Yes, the files that were in /proc/speakup have been moved to
> /sys/module/speakup/parameters.

In order to get this working with the new system I had to do a little
hacking in /etc/init.d/speechdup.

Specifically at top level
SPEAKUPDIR=/sys/module/speakup/parameters # Not /proc/speakup

and in load_speakup()

  echo soft > ${SPEAKUPDIR}/synth #which used to read echo sftsyn

I have no idea how, but people who want this to work for kernels
>=gentoo-sources-2.6.25-r1 need these in /etc/init.d/speechd-up,
and people who want the pre 2.6.25 need the way it is as of this morning
Which had the following at the top if it helps.

$Header: /var/cvsroot/gentoo-x86/app-accessibility/speechd-up/files/speechd-up.rc,v 1.6 2007/05/30 04:05:27 williamh Exp $

Hope this is more helpful than annoying. 
Best, 
Robert