software synthesis support

[Roman Vanicek]

Hi all,

I have one more question to ask. Have you considered to add the support
to the software synthesis into speakup?

I am from the Czech republic and people don't use hardware synthesis
here that much, usually because of price, which is quite high. But on
the other hand, we have one or two software synthesizers which are free
software and which can be used.

We have managed to use these synthsizers with emacspeak which helps
visually impaired users very much. But having such a support in speakup
would be even better.

This is just an idea at the moment but I would like you to think about
it and maybe shape it a bit more. My notion is to send the speech output
to the pipe (eg. /dev/speech) which would be read by the speech daemon
which would know how to work with given software synthesis.

The protocol should be as simple as possible, the best would be to use
some well known (eg. apollos) syntax.

I feel I could do the deamon part but I'm a bit scared about hacking
kernel sources ;-).

What do you think about it?  Would it be feasible/hard/easy?
I would appreciate any comments?

Regards
Roman