Key rate limit

[Paul Vojta]

On Thu, Mar 29, 2012 at 04:26:15PM +1000, Peter Hutterer wrote:
> On Tue, Mar 27, 2012 at 12:43:49PM -0400, Paul Vojta wrote:
> > On Mon, Mar 26, 2012 at 11:20:03AM -0500, tsuraan wrote:
> > > > I am not sure but xset has a [r rate delay [rate]], would that be any help ?
> > > 
> > > I believe that's for the autorepeat rate, so the rate/delay tells X
> > > that once the key is held down for <delay> ms, it should start firing
> > > that key at <rate> events/s (my units might be wrong though).  That's
> > > sort of the inverse of the problem I have, which is that the keyboard
> > > seems to be sending a bunch of keypress events really quickly for a
> > > short period of time.
> > 
> > 
> > I have a possibly related problem.
> > 
> > Sometimes I get keystrokes occurring via autorepeat even when no keys are
> > pressed.  They stop when I hit any other key (even shift).
> > 
> > I've looked into it, and found out that the problem is that the keyboard
> > (or control chip) is not sending the "break" code until the next key is hit.
> > Then, since X apparently keeps track of which keys are pressed on its own,
> > I get the repeated keystrokes.
> > 
> > Is there a way of telling X to just use the autorepeat provided by the
> > hardware?
> 
> not anymore. we've removed this in server 1.6 (iirc) because that and
> software autorepeat had some weird side-effects. sometimes a key would start
> hw-repeating before sw-repeat kicked in, causing bursts of keys. and iirc we
> had the same problem with keys not getting released correctly.
> 
> that was a time when the whole input system was in a turmoil so it's
> possible that we could fix this now but I still expect it to be a rather
> large (and nasty to test) amount of work to get selective hw repeats.
> 
> Cheers,
>   Peter

OK, thanks for the information.

Sincerely,


Paul