Maintaining speechd-up

[Bill Cox]

I prefer to keep things in terms that impact what we do.  Here's what I see:

- speakup is the most popular console screen reader for blind Linux
users, at least in Vinux
- speakup is hard to maintain and extend because it's old code,
running in kernel space
- In Vinux, we gain no benefit from being a kernel module
- Speech-dispatcher doesn't support speakup
- SBL will not gain popularity in Vinux without some key features
speakup users enjoy
- Running SBL or speakup globally continues to be the most practical
approach.  The same goes for the speech-dispatcher instance that reads
text from speakup.
- gnome-terminal with Orca functions poorly.  There may be an
opportunity to merge functionality there with the console screen
readers.

SBL seems like the front-runner for replacing speakup in Vinux.  This
would be a good thing in several ways.  However, it's not important
whether it's better or worse than speakup, or if I like it or not,
because I can't replace speakup with SBL until Vinux users want it
that way, and they don't today.

It seems to me is that Halim should work with us Vinux guys to get SBL
to a point where Vinux users prefer it over speakup, though I
understand some of the needed features are hard, and Halim doesn't
agree with Vinux uses on some features like how cut/paste works and
reading of scrolling text.

Braillcom seems committed to speech-dispatcher, and to keep current
speech-dispatcher consumers happy, I would guess that merging
improvements made in ospeakup back into speechd-up would be a good
idea.  When I broke speechd-up in the Vinux testing repository, all
hell broke loose!

Bill