Hi, @alvinhochun!

I have tested your build. Here are the issues:

1. Bended lines problem is here again

2. Rotation sensor rotates into an inverted direction and wraps in two points (looks like it wraps in 2 * M_PI points)
3. Tilt sensor looks as if it is not normalized: most probably 90 vs 60 degrees issue, since Krita normalizes the stuff by 60 deg
4. Tilt direction sensor seem to work correctly (at least it doesn't wrap)

And "bended" lines problem is the major one

That is curious though, since it doesn't seem to happen on my sp3. You're testing on the intel sdp, right?

In D7441#138917, @rempt wrote:

That is curious though, since it doesn't seem to happen on my sp3. You're testing on the intel sdp, right?

I'm testing on "Wacom 5 Touch M". Though connected to the SPD

I have tested the build on the embedded SDP's digitizer. It works even without special wintab drivers (obviously), but has the same bended lines as Intuos connected to this SDP.

In D7441#138915, @dkazakov wrote:

2. Rotation sensor rotates into an inverted direction and wraps in two points (looks like it wraps in 2 * M_PI points)

From https://msdn.microsoft.com/en-us/library/windows/desktop/hh454909(v=vs.85).aspx

rotation
Type: UINT32
The clockwise rotation, or twist, of the pointer normalized in a range of 0 to 359

QTabletEvent's docs doesn't state the direction and range of the rotation, but by taking a look at the WinTab code more carefully it seems that it should be anti-clockwise and in the range of [-180, 180]?

rotation = 360.0 - (packet.pkOrientation.orTwist / 10.0);
            if (rotation > 180.0)
                rotation -= 360.0;

The current calculation used is: penInfo.rotation > 180 ? penInfo.rotation - 360 : penInfo.rotation. It seems like there are two issues:

1. Flipped direction
2. I overlooked the fact that penInfo.rotation is unsigned and there is an unsigned underflow in the subtraction :(

The proper calculation would seem to be:

int rotation = 360 - penInfo.rotation; // Flip direction and convert to signed int
if (rotation > 180) {
    rotation -= 360;
}

Not sure what's the wrapping problem you mean, but could it be the underflow that you're seeing?

3. Tilt sensor looks as if it is not normalized: most probably 90 vs 60 degrees issue, since Krita normalizes the stuff by 60 deg

The end result is still an integer, so I'll just do an integer multiplication and division here: penInfo.tiltX * 2 / 3

1. Bended lines problem is here again
   
   And "bended" lines problem is the major one

I have tested the build on the embedded SDP's digitizer. It works even without special wintab drivers (obviously), but has the same bended lines as Intuos connected to this SDP.

It's probably either issues with the driver or the system... @rempt's SP3 works fine and my Intuos Pen and Touch works fine too. @scottpetrovic tried an earlier test build and said it worked fine on the SP4 (don't know how smooth it is though).

This shouldn't be a show-stopper, since the feature is not enabled by default in any case.

(Making a test build with the changes)

Were you testing with no smoothing? I think I always had basic smoothing enabled when testing.

I guess I also tested with the smoothing enabled, I just didn't touch
the default settings. I will check again when testing your new build

I tested this on my Surface Pro 2 (Wacom) to see if there were any regressions. For me it looks like the Win8+ "Pointer Input Message" produces jitter compared to the wintab version. I've attached an image of two reasonably straight lines, drawn in two instances of the 4.0-prealpha_winink build. The first one with the Win8+ Pointer Input Message support enabled, the second without. I merged the images to show the difference (top: "winink", bottom: "wintab"):

Edit: Both are without smoothing enabled

I've also attached two tablet logs for each session, with a "krita.tabletlog" filter:

The only thing I found of note outside of the "krita.tabletlog" lines was that the winink one also says it's using that API:

00000044	10.00228977	[4216] krita.tabletlog: Loaded Windows 8 Pointer Input API functions	
00000045	10.00232697	[4216] Using Win8 Pointer Input for tablet support

In D7441#140672, @alvinhochun wrote:

Fixed the incorrect localPos calculation.

New build: https://drive.google.com/open?id=0B1i7Q5Isuy2JeTJpczNvRHV4MlE

Verified fixed here, nice work! I double checked so it was using Win8+ Pointer Input instead of wintab too.

It should be noted that the above build is 3.2.1-based and not 4.0, though that just means one more data point in favour of the feature going into the 3.2 branch! :)