HOW TO: Set up the HP TX2z and Dell XT & XT2 (N-trig digitizer) in Ubuntu

[markginter24]

No luck. First it complained about multiple N-Trig Pens so I made a copy of the script, changed all N-Trig refs to the id's - then it didn't complain but it didn't do anything either.

Are we sure I don't need to update my hid-ntrig kernel module?

[Favux]

Try changing the ID #'s to the other pair.

Are we sure I don't need to update my hid-ntrig kernel module?

Are you running Rafi's 5-4-10 version? And do you know which N-trig firmware you have?

[markginter24]

No patches at all. I'm just running the stock suse kernel.

How would I tell what firmware I'm on?

[Favux]

I don't think, from what Rafi recently said, that the 2.6.34 kernel has the 5-4-10 hid-ntrig.ko in it. I think it's an earlier version. I think he was going to tell us what versions were in what kernel, if he could figure out how to extract that info.

If you're dual booting, Windows would be the best way. You could try Rafi's utils linked near the top of the HOW TO. But I think the version number it returns for linux is different from the Windows firmware version number.

[markginter24]

Interesting. I may poke around a bit and see if there's any repositories for Suse and 2.6.35 .... not sure if I feel like compiling much more

[enneract]

Here it is updated source for XT2 WMI driver.

It needs more work, namely Fn+F2 does not work, it needs another key entry (a switch, as the key is a toggle, scancodes are 0x022f and 0x0230), and I don't know how to do that (compile dies, so key names have to be defined somewhere and I am completely in the dark here), and Fn+F8 needs to be ... aliased, I think. scancode coming off it is 0x0070, which spits out of the driver as keycode 33, which is 'p' on my kb layout. I'm hoping there is a more elegant way of handling than than hardcoding an alias into the driver.

Fn+F9 reports the same keypress as the wifi switch does (0x0011), but without the hardware deactivation.

If someone can even point me in the direction of reading material for how to fix any of those issues, I'll be happy to provide what I come up with. Keep in mind that I am only a second-year physics major who is only picking up programming as a secondary skill, though, going over my head is a distinct possibility.

Wifi Catcher (KE_PROG2) needs to be held down for a second or so.

KEYMAP for XT2, including tablet bezel keys.

wifi Catcher = 157
Fn+F2 = ??? (KE_BATTERY, idk how to remap this, xev doesn't catch it)
Fn+F3 = not mappable yet
Fn+F7 = 156
Fn+F8 = 33 (useless for now)
Fn+F9 = 246

side button = 159
toggle left = 163
toggle right = 155
toggle press = 158
wrench = 143
rotate = 199

[markginter24]

Here's what I found:

Using kernel 2.6.36 rc2 - no difference - touch is still offset.

Replacing ids/N-trig entries in script with touch and stylus (as listed by xsetwacom list) causes the cursor to line up with touch, but not respond to any 'click' actions. The cursor will trail my finger properly, but will not click buttons or highlight things. Upon further review, it is the 'Relative' 'Absolute' value that makes the only difference. Set to 'Absolute' it properly trails but does not respond, set to relative it does not trail, but it can 'click' on things.

Interesting.

[Ayuthia]

Here's what I found:

Using kernel 2.6.36 rc2 - no difference - touch is still offset.

Replacing ids/N-trig entries in script with touch and stylus (as listed by xsetwacom list) causes the cursor to line up with touch, but not respond to any 'click' actions. The cursor will trail my finger properly, but will not click buttons or highlight things. Upon further review, it is the 'Relative' 'Absolute' value that makes the only difference. Set to 'Absolute' it properly trails but does not respond, set to relative it does not trail, but it can 'click' on things.

Interesting.

EDIT:You will want to try the next post first. It will be a little easier.

If you don't mind testing out a different hid-ntrig kernel version, can you try this one? You will need to replace the /usr/src/linux/drivers/hid/hid-ntrig.c with the one in that link and rebuild the kernel module. The kernel module source from this link has Rafi's most recent changes in May along with some additional changes that he submitted in the input mailing list (but it does not seem like it is going to remain in there). The more recent change provides a way to extract the firmware version which will help us identify yours.

If you are able to build the kernel module, you will need to check:

ls /sys/class/hidraw/hidraw0/device/
ls /sys/class/hidraw/hidraw1/device/

In one of those results, you should see something like:

activate_slack     deactivate_slack  hidraw      power                  sensor_physical_height  uevent
activation_height  driver            min_height  sensor_logical_height  sensor_physical_width
activation_width   firmware          min_width   sensor_logical_width   subsystem

If you find it, please report the values from:
If the files can be found in hidraw0

cat /sys/class/hidraw/hidraw0/device/*

If the files can be found in hidraw1

cat /sys/class/hidraw/hidraw1/device/*

[rafiyr]

Hi, I seem to be quite behind in the forum. For firmware version, switching modes (turn off touch or pen in hardware) and calibration, I have a few little tools at:
http://code.launchpad.net/~utouch-te...rig-calibrator

For people with apt:

apt-get install libusb-dev
bzr branch lp:~utouch-team/utouch/ntrig-calibrator
cd ntrig_calib
./firmware_version.sh

For suse users and others, get the headers for libusb one way or another

Note, older firmwares won't give a proper version with this code. But in those cases, you don't really have a shot at getting multi-touch working anyway. (If there are enough people out there, who really care, I do have a general idea how to get the version off the older firmwares, so speak up).

I'm working on pushing load time firmware id to the kernel driver, so hopefully 2.6.36 or 37 will just have the version in the boot log.

[rafiyr]

Here it is updated source for XT2 WMI driver.

Interesting I have different codes for my xt and a friend's xt2 for the hinge rotation. 0xe04[67] for the xt, I don't have the xt2 codes at the moment, but they were pretty close, not 0xff...