Hi Nicolas, hi Hendrik,

I do not have an E6 system to test with. So, I tested it with an E3 system,
which has the same pen and touch models as E6. Except item 4, everything
else looked fine.

Even for item 4, it is not a bug, at least from driver's point of view ;-).
It was designed that way.

We did not report touch in AbsMTPosition{X,Y} since we thought clients
would get the details from the new Abs{X,Y} format.

We do understand there could be reasons for clients to expect AbsMTPosition
for MT enabled devices. I would like to get a second opinion from Peter,
who is cc'd, for pros and cons on reporting MT events in AbsMTPosition{X,Y}
format. The support will be added as soon as we are sure that is the way to
go.

Please see my explanation for your other items inline.

On Tue, Aug 20, 2013 at 1:58 PM, Hendrik Woidt <hendrik.wo...@tuhh.de>wrote:
>
>> Hi,
>>
>> While testing the wacom tablet driver with xinput2 on the Lenovo Thinkpad
>> x230t (ISDv4 E6) running xf86-input-wacom-0.21.99.1 we noticed some things
>> that seem strange:
>>
>> (Testing was done using 'xinput --test-xi2 <pen_device_id>' and "Option
>> 'Gesture' 'Off'" in xorg.conf)
>>
>> 1. When moving the pen into proximity (without touching the surface) and
>> then
>> pressing the pen-button, no RawButtonPress events are generated, while
>> releasing the button gives us a RawButtonRelease event. The
>> xf86-input-evdev
>> driver raises that event.
>>
>
My testing shows both ButtonPress and ButtonRelease events when I pressed
the tablet then releaseed it.

2. Another issue can be described as follows:
>>
>> =============== =============================== ==========================
>> Action          Event(actually)                 Event(expected)
>> =============== =============================== ==========================
>> pen-down        RawButtonPress(detail=1) RawButtonPress(detail=1)
>>
>> pen-button-down RawButtonRelease(detail=1) RawButtonPress(detail=3)
>>                  RawButtonPress(detail=3)
>>
>> pen-button-up   RawButtonRelease(detail=1) RawButtonRelease(detail=3)
>>                  RawButtonRelease(detail=3)
>>
>> pen-up          RawButtonRelease(detail=1) RawButtonRelease(detail=1)
>> =============== =============================== ==========================
>>
>>
>> =============== =============================== ==========================
>> Action          Event(actually)                 Event(expected)
>> =============== =============================== ==========================
>> pen-down        RawButtonPress(detail=1) RawButtonPress(detail=1)
>>
>> pen-button-down RawButtonRelease(detail=1) RawButtonPress(detail=3)
>>                  RawButtonPress(detail=3)
>>
>> pen-up          RawButtonRelease(detail=3) RawButtonRelease(detail=1)
>>
>> pen-button-up   RawButtonRelease(detail=3) RawButtonRelease(detail=3)
>> =============== =============================== ==========================
>>
>> 3. Why does the pen-button generate RawButton{Press,Release} events with
>> detail=3 by default? The evdev driver emits events with detail=2. This
>> is not
>> that nice when trying to support both out of the box.
>>
>
We report the button as button/detail 3 since on tablet PC, there is no
middle (detail 2) button. Plus, most tablet PC styli only have one side
switch, which is translated into the next most useful button (left
click/detail 1 is the most useful), right click/detail 3.

If you think this default is out of date or not a good idea on Linux, you
can initiate a discussion in the mailing list. We can change it if most
people agree.


> 4. Additionally, when working with the Qt5-Library the Touchpanel does
>> not work.
>> We tracked that down to the fact, that Qt watches for event-valuators with
>> atoms AbsMTPosition{X,Y} while the driver labels the valuators with
>> Abs{X,Y}.
>> This is probably implmented that way because the xf86-input-evdev uses
>> these
>> atoms. Is this a bug in xf86-input-evdev or in the xf86-input-wacom
>> driver?
>>
>> We hope to hear some thoughts on these issues, and will gladly provide
>> more
>> information if needed.
>>
>
We are open to suggestions too. Our goal is to provide consistent support
to Wacom users.

Ping
------------------------------------------------------------------------------
Introducing Performance Central, a new site from SourceForge and 
AppDynamics. Performance Central is your source for news, insights, 
analysis and resources for efficient Application Performance Management. 
Visit us today!
http://pubads.g.doubleclick.net/gampad/clk?id=48897511&iu=/4140/ostg.clktrk
_______________________________________________
Linuxwacom-devel mailing list
Linuxwacom-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxwacom-devel

Reply via email to