On Tue, Sep 18, 2012 at 11:57:12AM -0700, Adam Williamson wrote:
On Tue, 2012-09-18 at 08:35 -0400, John.Florian(a)dart.biz wrote:
> > From: Adam Williamson <awilliam(a)redhat.com>
> > Oh, I should also note that, IIRC, the intent is that the driver
> > detect if there are no physical buttons and enable tap-to-click in
> > case. So touchpads which have no buttons and are only supposed to
> > with tap-to-click should be OK.
> Where does my notebook's touchpad fall in this continuum? At the
> bottom corners of the touch-sensitive area are two "buttons" which
> click with tactile feedback, but yet are still part of the
> touch-sensitive surface. In other words, the bottom corners can
> actually be deformed/depressed. FWIW, I enabled tap-to-click -- did I
> just answer my own question? -- simply because my wife and I both
> found the mouse to be moving off target too often when tried using
> these "buttons".
As far as evdev is concerned those are almost certainly just perfectly
normal buttons, i.e., they send a 'button press' event. The fact that
they also function as part of the touch-sensitive surface is probably
irrelevant. So evdev would see your touchpad as one with buttons, and
wouldn't enable tap-to-click.
the device looks like a single-button touchpad with four-finger
capabilities. important is the INPUT_PROP_BUTTONPAD property which we
translate to the ClickPad option on driver startup (on older kernels that
option/property needs to be set manually).
if clickpad support is enabled in the driver, we have a number of different
code paths that handle this type of device to support the basic
functionalities like drag&drop. that's also the reason why we didn't
backport this to F16, it's just too much effort.