[PATCH xf86-input-synaptics] Turn down the default sensitivity for MacBook touchpads

[Peter Hutterer]

On Wed, Jun 06, 2012 at 03:40:21PM -0700, Chase Douglas wrote:
> On 06/05/2012 06:35 PM, Peter Hutterer wrote:
> > On Tue, Jun 05, 2012 at 09:46:40AM -0700, Chase Douglas wrote:
> >> The touchpads are too sensitive by default. A touch is activated when a
> >> finger hovers a millimeter off the device. This change turns down the
> >> default sensitivity so that accidental touchpad interactions are
> >> lessened and pointer interaction accuracy is increased.
> >>
> >> Signed-off-by: Chase Douglas <chase.douglas at canonical.com>
> >> ---
> >>  conf/50-synaptics.conf |   10 ++++++++++
> >>  1 file changed, 10 insertions(+)
> >>
> >> diff --git a/conf/50-synaptics.conf b/conf/50-synaptics.conf
> >> index 9e86a7a..5b357ea 100644
> >> --- a/conf/50-synaptics.conf
> >> +++ b/conf/50-synaptics.conf
> >> @@ -42,3 +42,13 @@ Section "InputClass"
> >>          MatchDriver "synaptics"
> >>          Option "SoftButtonAreas" "0 0 0 0 0 0 0 0"
> >>  EndSection
> >> +
> >> +# The trackpads in Apple MacBooks are very sensitive. This input class rule
> >> +# turns down the default sensitivity.
> >> +Section "InputClass"
> >> +	Identifier "MacBook touchpad sensitivity"
> >> +	MatchProduct "bcm5974"
> >> +	MatchDriver "synaptics"
> >> +	Option "FingerLow" "70"
> >> +	Option "FingerHigh" "75"
> >> +EndSection
> >> -- 
> >> 1.7.9.5
> > 
> > what's the pressure sensitivity on those? calculate_tap_hysteresis sets
> > those defaults and the calculation is odd to say the least. can we fix this
> > to use a sensible calculation?
> 
> Pressure is a bit of a misnomer on these devices. It appears to actually
> be the width of a touch, even though there is a separate field for the
> width, too. It may be that this is the main reason why the current
> calculation doesn't work.

at some point in the past pressure was calculated from the width, and I
believe this is still the case in hardware for some touchpads. the kernel
just abstracts it differently now. figuring out why we have pressure and
width is surely of archeological interest, but I didn't bring my helmet
today. 2acbd6e67c859290a37cf3134b120c5ca7577268 is a good start though.

> I have created two patches that add a new model type and tap hysteresis
> calculations for these trackpads. I just hard coded a value, since I
> think it will always be equal to this for all bcm5974 devices, and since
> it will likely be different from any other devices.

thanks, I'll have a look.

Cheers,
  Peter