Re: New touchpad still has some jumpyness

2008-11-25 Thread Jim Gettys
On Tue, 2008-11-25 at 18:54 -0500, Richard A. Smith wrote:
> During SJ's demo meeting today at 1cc I used my test machine with the 
> new touchpad.
> The touchpad on my laptop is much deeper inset than
> I still experienced jumpyness.   However, a much different jumpyness 
> than what happens with the alps device.  In my case it appears to happen 
> if the tip of my thumb gets up into the active area. (Creepage from 
> using it to click thThe touchpad on my laptop is much deeper inset than 
> e X button)   What I saw was the cursor would leap to the edge of the 
> screen which is similar to what happens with the current pad, but unlike 
> the current pad its does not continue to leap around. It a one shot and 
> then returns to normal movement.  This happened to me several times.
> 
> So then I played with my HP laptop which also has a synaptics touchpad 
> and I'm able to duplicate the same behavior.

It's worth a gander at the upstream synaptics driver; I've seen some
patches go by recently.

> 
> So even with the new touchpad we still may have to have some sort of 
> criteria for discarding packets with large deltas for the movement.
> 
> By default in mouse mode the values reported by the device are relative. 
>   So some sort of edge effect must confuse the controller.  I need to 
> enable some logging and see what data the device actually sent.
> 
> I've also noticed that the acceleration of the new touchpad is much less 
> than the alps device.  When you switch from new to old the you can 
> really tell.  We should reduce the acceleration of the cursor for the 
> alps device. Its not necessary to be that fast.
> 

Yeah, the acceleration stuff in X sucks.  Fixed in next release, IIRC
(after something like 20 years...)

> Is that something we can set defaults for in the X config or do we have 
> to detect what touchpad we are running?
> 

Unfortunately, we'll have to detect it.  This whole are of X is being
reworked as we speak  Upstream, everything is hot plug, and you can
then set things on a per detected device basis.
  - Jim


-- 
Jim Gettys <[EMAIL PROTECTED]>
One Laptop Per Child

___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel


New touchpad still has some jumpyness

2008-11-25 Thread Richard A. Smith
During SJ's demo meeting today at 1cc I used my test machine with the 
new touchpad.
The touchpad on my laptop is much deeper inset than
I still experienced jumpyness.   However, a much different jumpyness 
than what happens with the alps device.  In my case it appears to happen 
if the tip of my thumb gets up into the active area. (Creepage from 
using it to click thThe touchpad on my laptop is much deeper inset than 
e X button)   What I saw was the cursor would leap to the edge of the 
screen which is similar to what happens with the current pad, but unlike 
the current pad its does not continue to leap around. It a one shot and 
then returns to normal movement.  This happened to me several times.

So then I played with my HP laptop which also has a synaptics touchpad 
and I'm able to duplicate the same behavior.

So even with the new touchpad we still may have to have some sort of 
criteria for discarding packets with large deltas for the movement.

By default in mouse mode the values reported by the device are relative. 
  So some sort of edge effect must confuse the controller.  I need to 
enable some logging and see what data the device actually sent.

I've also noticed that the acceleration of the new touchpad is much less 
than the alps device.  When you switch from new to old the you can 
really tell.  We should reduce the acceleration of the cursor for the 
alps device. Its not necessary to be that fast.

Is that something we can set defaults for in the X config or do we have 
to detect what touchpad we are running?

-- 
Richard Smith  <[EMAIL PROTECTED]>
One Laptop Per Child
___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel