On Tue, Aug 31, 2010 at 10:37 PM, Peter Hutterer
<peter.hutte...@who-t.net> wrote:
> On Tue, Aug 31, 2010 at 10:02:38PM -0500, Chris Bagwell wrote:
>> On Tue, Aug 31, 2010 at 12:44 AM, Peter Hutterer
>> <peter.hutte...@who-t.net> wrote:
>> > On Wed, Aug 25, 2010 at 06:28:53PM +0200, Ping Cheng wrote:
>> >> On Wed, Aug 25, 2010 at 5:04 PM, Chris Bagwell <ch...@cnpbagwell.com> 
>> >> wrote:
>> >> > Sure, I'll jump back in once the touch-only model support is
>> >> > committed.  Assuming Henrik's patch is submit, there is some things to
>> >> > discuss on X side; which I will be happy to help with as well.
>> >>
>> >> Thank you for your help.
>> >>
>> >> > Henrik's current driver on linux-input is both MT and synaptic-like;
>> >> > which is actually my preference as well but brings up important issue.
>> >> >  Do we modify xf86-input-wacom to be compatible with synpatic-like
>> >> > input events or do we work with distributions to develop rules to
>> >> > re-direct wacom "touch" input devices over to xf86-input-synaptics?
>> >>
>> >> I'd like to know Peter's answer to this question. It is not purely a
>> >> Wacom driver decision, I think.
>> >
>> > AFAICT, the device is exported as a separate device, right? If so, there's
>> > no need to hook the wacom driver onto it - we just treat it as a touchpad
>> > and hook the synaptics driver onto it.
>> >
>>
>> Yes, on both Bamboo's and Tablet PC's, the exported device is nicely
>> isolated.  In Bamboo's case, its very much like a standard touchpad
>> (x/y/pressure+4 buttons).  For Tablet PC's, its very much like a
>> buttonless touchpad.
>
> I missed the tablet PC case. Can you point me to the commit that added this?

Yes.  This commit adds Tablet PC.  In this context, its not
xf86-input-synaptics compatible because of way it uses
BTN_TOOL_DOUBLETAP.  Since its buttonless though it doesn't have the
additional problem of how buttons are reported.

http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=545f4e99dee7284ed57c79384c5c1d5ac58dcd59

The next commit adds multi-touch and further prevents being able to
use with xf86-input-synaptics because of way it uses
BTN_TOOL_TRIPLETAP and sends unrelated X/Y/Pressure with BTN_TOUCH.

http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=ec67bbedcf290ef182a897017f65a2707106c7f8

>
>> Some pros/cons for having xf86-input-synaptics handle wacom touch devices.
>>
>> Pros:
>> * Get synaptics gesture support immediately (edge scrolling, two
>> finger scrolling).
>> * Get gnome-mouse-properties (and similar) GUI configuration support
>> (I can finally enable Lefty mode from a GUI!)
>> * It will force us to be Synaptics compatible.  This allows switching
>> to xf86-input-multitouch driver and get their gestures (swipes,
>> pinching, 3-fingers, etc).  Or at least it keeps us closer to
>> compatible to what ever client-side gesture engines are coming up.
>>
>> Cons:
>> * ISDV4 devices requires some touchpad logic in xf86-input-wacom
>> anyways (not a code savings)
>
> 2.6.37 will have kernel MT slot support for ISDV4, the patch for this has
> recently been committed. However, this does not split into multiple devices,
> I need to look into Henrik's patch again to see how that works.
> either way, wcmISDV4.c is a dead end, or at least will be soon.
>
>> * User would lose xf86-input-wacom's advanced button remapping.
>
> meh. I'm not convinced yet this button remapping is solving more issues than
> it causes...
>
>> My suggestion?
>>
>> * Make touchpad-like wacom devices to use events that are
>> synaptics-like (no multiplexing, no serial/channel #'s, change way
>> BTN_TOOL_*'s are used, use MT events for 2 finger HW, etc).  This
>> means specifically Bamboo and Tablet PC's.  Since Tablet PC code has
>> already shipped in a kernel, we need a roadmap of how and when to
>> convert over.
>> * Modify xf86-input-wacom so that it can handle todays ISDV4/TabletPC
>> version of touchpads and also make it compatible with synaptics-like
>> touchpads.  Since its backwards/forwards compatible, that roadmap for
>> TabletPC can be change it today.
>> * Decide which X driver we prefer as default and then let user do any
>> xorg.conf.d modifications if they prefer a specific pro/con item
>> above.
>
> IMO, wacom is a _tablet_ driver. This means it should handle tablets as good
> as possible but it should not become a touchpad driver just because some
> tablets start looking like touchpads.

Personally, I agree as well.  As a person thats been submitting
changes related to multiple touchpad-like devices (including a
touchscreen), I prefer to understand a single set of events from
kernel instead of the two set I understand right now.

> Long term we're looking at true multitouch anyway and I'm not sure how
> synaptics will hold up in this regard.

Agree.  In MT, BTN_TOOL_DOUBLETAP and TRIPLETAP will actually get in
the way if we are not careful (which is why xf86-input-multitouch
started from xf86-input-evdev I'm sure).

>
> Cheers,
>  Peter
>

------------------------------------------------------------------------------
This SF.net Dev2Dev email is sponsored by:

Show off your parallel programming skills.
Enter the Intel(R) Threading Challenge 2010.
http://p.sf.net/sfu/intel-thread-sfd
_______________________________________________
Linuxwacom-devel mailing list
Linuxwacom-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxwacom-devel

Reply via email to