Speakup in kernel

[Janina Sajka]

Alan Cox writes:
> From: Alan Cox <alan at redhat.com>
> 
> > in a matrix of devices requiring support from the bios through the
> > desktop. Indeed, there is a releated FSG project now getting under way
> > in the new Accessibility WG I chair under FSG.
> 
> The linux kernel is run by Linus. Thats the challenge. To get stuff good
> enough Linus will take it. If its modular you are a large part of the way 
> there because you can get the base to the point he takes it and one nice
> clean driver and then work from it.

Well, and that's were we will take things eventually. However, coming
from the FSG point of view, I don't want to go to the kernel group
(Linus) with one thing today, and another tomorrow, and a third yet
thereafter. We'll coordinate the user requirements and suggested
approaches that cover Speakup as well as any other assistive technology
that might need to hook in via kernel speace before doing this from FSG.

So, why bring it up to FC? Because Speakup was already there in RH
8.0--and then was summarily pulled, even though the I18N issue which
triggered the retraction was resolved. I may add it was resolved before
Speakup became modular. There is a fairly extensive user base, small
compared to the whole, but not as compared to users of eyes-free.

> 
> > If video belongs in the kernel, so does Speakup.
> 
> Not neccessarily. There are good arguments that kernel mode handling
> beyond adding working select() on /dev/vt* isn't needed. Its also a deeply
> political issue it seems - some blind folks are deeply pro speakup others
> deeply anti. Future questions like the vanishing of serial ports on PC's
> make it even harder to keep kernel side.

I don't understand what user preference has to do with it. As far as I
can tell, some folks want KDE, for example, and never want to see a bash
prompt. Does that mean the console is unimportant? So what if some want
to use emacspeak, yasr, or some user space assistive technology. They
are entitled.

Surely we don't expect all flowers will be roses?

So, if user interfacing is supported at bios, bios should be
accessible--if reasonable to do so, technically. Ditto for kernel,
console, and desktop. Whatever mechanisms are adopted to provide user
access should support a wide range of user interface devices, imho.


> 
> 
> --
> fedora-test-list mailing list
> fedora-test-list at redhat.com
> http://www.redhat.com/mailman/listinfo/fedora-test-list

-- 
	
Janina Sajka
Email: janina at rednote.net		
Phone: (202) 408-8175

Director, Technology Research and Development
American Foundation for the Blind (AFB)
http://www.afb.org

Chair, Accessibility Work Group
Free Standards Group
http://accessibility.freestandards.org