On Thu, Aug 19, 2010 at 09:05:12AM +0100, Magnus Kessler wrote: > On Thursday 19 August 2010 06:14:08 Peter Hutterer wrote: > > With 4e0e53fcba6fd99d458df1905d055d63360155c0 the driver got it's own > > acceleration mechanism. This slowed down the pointer movement a lot, > > especially on ALPS touchpads. Increase the default acceleration factor to > > accommodate for this. > > > > The number itself is chosen through guesswork and informal tests on one > > Synaptics and one ALPS pad. > > > > Signed-off-by: Peter Hutterer <[email protected]> > > --- > > src/synaptics.c | 2 +- > > 1 files changed, 1 insertions(+), 1 deletions(-) > > > > diff --git a/src/synaptics.c b/src/synaptics.c > > index 3344d75..9d0bf14 100644 > > --- a/src/synaptics.c > > +++ b/src/synaptics.c > > @@ -455,7 +455,7 @@ static void set_default_parameters(LocalDevicePtr > > local) tapMove = diag * .044; > > edgeMotionMinSpeed = 1; > > edgeMotionMaxSpeed = diag * .080; > > - accelFactor = 50.0 / diag; > > + accelFactor = 200.0 / diag; /* trial-and-error */ > > > > range = priv->maxp - priv->minp; > > Tested on a Lenovo T500 (Synaptics Touchpad, model: 1, fw: 7.0, id: 0x1c0b1, > caps: 0xd04791/0xb00000/0x20000). Works well for me. > > There seems to be an awful amount of ad-hoc magic factors in that part of the > code though. And comments like "Again, based on typical x/y range and > defaults". Does anyone know of a way to express the factors in a more > rational > way?
numbers that decide on the user experience are often guesswork. much of this code came to be to provide the same-ish feel regardless of touchpad dimensions (they change slightly within models and greatly between models). I admit though that it's due for an overhaul to see if we can't just hardcode them on a per-model basis. Then again, this requires some sort of evaluation afterwards and that requires way more time than I can expend at this point. > For this patch, though: > > Tested-by: Magnus Kessler <[email protected]> > Reviewed-by: Magnus Kessler <[email protected]> thanks, much appreciated. Cheers, Peter _______________________________________________ [email protected]: X.Org development Archives: http://lists.x.org/archives/xorg-devel Info: http://lists.x.org/mailman/listinfo/xorg-devel
