[BRLTTY] qemu/vmware/... virtual braille devices?

[Samuel Thibault]

Dave Mielke, le Sun 08 Apr 2007 02:01:36 -0400, a écrit :
> [quoted lines by Samuel Thibault on 2007/04/08 at 01:25 +0200]
> 
> >However, this requires some not-so-trivial configuration 
> 
> Wouldn't it only be network configuration?

Yes, but also proper configuration of the second brltty: copying the
authorization key, and run
WINDOWPATH=7:0xc200002 brltty -b ba -B host=10.0.0.2

Where 0xc20002 is the window ID of qemu. This is not very trivial, and
having all that done automatically would really be a plus.

> >and it doesn't permit to test another screen reader than brltty in the
> >emulated system.
> 
> For this solution to be viable, how many different versions of the virtual
> driver would need to be implemented? If even more than one then I'd say that
> the simplest protocol would be in order.

What do you mean by "version"? If we emulate a device that is supported
by all well-known readers, and that is simple enough, we don't need to
be able to emulate other devices.

> I believe you said that one of the groups is pushing for the best documented
> protocol. The vendours aren't in the habit of publically documenting their
> protocols.

Isn't there at least one that did publicly document their protocol?

> >I mean, you are saying "why does qemu emulate a video board while we
> >could use vnc?" :)
> 
> Because kernels think they have full access to video cards.

Not necessarilly. They can be happy with a serial port too, and then the
configuration gets even more tricky.

Samuel