Re: keyboard events

[Peter Jay Salzman <p@dirac.org>]

On Thu 15 Mar 01,  2:24 PM, Erik Mouw said: 
> On Wed, Mar 14, 2001 at 01:36:06PM -0800, Peter Jay Salzman wrote:
> > i have a small project in mind.  xvoice is an application that uses IBM's
> > speech recognition API.  currently, it uses "synthetic events" to send
> > information to X applications.  this is how X gets gets "words" from your
> > speech.  many applications don't like synthetic events since they're
> > considered a security risk.
> > 
> > i was thinking of looking into a module which takes xvoice events and inject
> > them into the kernel as keyboard input.
> 
> Hmm, so you want to solve a policy problem in the kernel? This is
> wrong, a kernel only offers mechanisms, policy has to be done in
> userspace.
 
hi erik,

"has to be done" and "should be done" are two very different things.  :-)

of course, i know you're right.  however, i think i can provide an argument
to show you that such a module would be a good thing.

currently, there is an application called xvoice which provides speech
recognition in X apps.  i would like to nudge xvoice into "openvoice".
something that would work in both X and console.

seems like the best way to do that would be to intercept data from the
speech recognition engine and provide it as keyboard input in whatever
context (X or console) the user is in.

viewed in that way, i'm not trying to get around an X policy, but to provide 
a system mechanism.  do you agree?  think of it as an extension to the sound
card driver, since ultimately, that's where the data is coming from.

the flip side of the coin is that my wrists get tired of typing.  sometimes
even hurt (like they do right now).  *i* would benefit from the module a
great deal, which means more to me (and i suspect other people) than proper
module writing (although i can see how that wouldn't fly with many people).

it also occurs to me that if the developers ixnay ibm's API and go with a
GPL engine, it would be the perfect thing to include with the kernel.

and before anyone attacks, let me confess that a kernel webpage server makes
less sense to me than something that would benefit the millions of people
who suffer deeply from repetitive motion syndome.  :(   one is beneficial to
handicapped people.  the other, although ultimately useful, was probably
born from lost pride in the mindcraft fiasco.

hope i didn't make any enemies!  just wanted to explain my thinking.  :)

pete
-
Kernelnewbies: Help each other learn about the Linux kernel.
Archive:       http://mail.nl.linux.org/
IRC Channel:   irc.openprojects.net / #kernelnewbies
Web Page:      http://www.surriel.com/kernelnewbies.shtml