@Medievil As far as i know dunfell is the most stable branch.i used yocto from dunfell and meta-qt5 from dunfell.I did not make any modification to package configurations.Everything is working fine

@Christian-Ehrlicher
Done, Sir :-)
I was just not sure, whether the fix happened in 5.15 or 6.2.
The latter it is.

It's a QT bug ( https://bugreports.qt.io/browse/QTBUG-79254 )

If somebody every needs it again I implemented a workaround which I posted here:
https://stackoverflow.com/questions/62152845/touch-freeze-in-qtwebengine-after-opening-a-dropdown-menu

@jsulm said in switch button problem! qml:

@imhs It is, see "The following members are inherited from Item." section in the link you posted.

Thanks for your help. I saw it.

I figured out my issue.

I set the following environment variable:

QT_QPA_EGLFS_DISABLE_INPUT=1

Of course, the cursor dissapears, but when pressing around my Gui running on the Compositor I'm not noticing any buttons being pressed except when I press in them, so that's definitely preferred to a wonky cursor!

May I suggest to change for Python 3 ?
Python 2.7 will reached end of life next year so it's not a good idea to start a new project with it.

Hi,

Well, evdevtouch plugin seem to be completely buggy.
The rotation woks very well but afin 1 to 10 secondes, the touch does not woks anymore.
Even with example programs.

I have this information :
TouchPointPressed without previous release event QQuickEventPoint(accepted:false state:Pressed scenePos:QPointF(249.631,798.957) id:2000002 timeHeld:0)

I have to reset the application to work.

Is there any solution to workaround this bug or a solution with libinput ?

Thanks again for your input @tekojo. I'm not 100% sure on the gloves just yet, likely a mixture of synthetic, leather and suede. I'll keep the sliders for now, and will give them a thorough test once my hardware is setup.

Thanks for your help.

It's an application that runs on a big screen that people will use standing up. So they will be using touch rather than a keyboard.

The reason for Windows 7 is that the application will need to attach to hardware, D2K-DASK http://www.adlinktech.com/PD/web/PD_detail.php?pid=350. The hardware drivers are only guaranteed for Windows 7 and 8. Its possible it might work on Windows 10 anyway but it's not clear.

Also, they want to use a touch number pad for number only entry. Not sure how much extra effort that would involve? I just found this artical http://www.cplusplus.com/forum/windows/128603/ , is ok?

@chaithubk
Are you reading from your device

continously
or just reading from it when device notifies you that data has arrived

In the scenario 1 device will misbehave like you have mentioned

For your screen mouse area, have you tried this?

MouseArea { anchors.fill: parent onPressed: { // your stuff here, e.g. global timeout reset mouse.accepted = false; } }

@etiennedm Why don't you just use layouts?

@Silveyra Probably, libqtslibplugin.so file under directory qt/plugins/generic/
Please correct me if I'm wrong.

Hi,

You have several wiki pages about Qt and the Raspberry Pi, see here for a start.

Driver was the the right way to go, thanks for pointing this out.

It turns out that the driver had a calibration method built in, but the transform matrix was not set to change the coordinates. I calculated the new matrix values based on the max coordinates displayed by evtest and the actual coordinates of the framebuffer, plugged them in to the driver code, compiled and ... it works!. Evtest displays correct coordinates and Qt works correctly with no tslib.
I don't like the fact that calibration matrix is hardcoded into the driver, but I consider fixing this an extracredit activity to be done sometime in the future.

@dzimiwine I tested 5.5.0 beta and it works fine

I'll try that.
Thanks