uclogic: Add support for Parblo A640 #393
Conversation
|
hey spbnick first of all greetings if you are developing form the #385 issue for parblo a640 thanks |
|
Hmm, could @DONofDESTROY specify which of the dumps different, and if possible to specify, in what respect? I wonder if the order of pressing buttons etc. was different. What I did
Results
|
|
wait a minute my usb-hid string changes to like yours I guess the dumps I make were from default with no drivers(specified by myself) and yours is digimend or wacom (looks like the last two hex bits ie 0x00 and 0x00 are tilt atleast me) and i wonder wether they plan to do that cause on one of the descriptors i saw tilt(from interface 0 without any drives specified by me) (idk it might be proximity bit or even some garbage values) |
|
issues and evtest is not showing any outputs on any events however the unwanted events are hided or not showing |
|
Could you please check that you're actually using the installed driver? You should get |
|
cat /sys/module/hid_uclogic/version my dmesg |
|
@DONofDESTROY, do you also get P.S. Please use triple back-ticks for quoting blocks of pre-formatted text, not single ones. |
|
@spbnick yes I am geting that |
|
OK, I'll have to find some more time to look into the diagnostics you collected and fix this. |
|
Thank you, @DONofDESTROY, and no problem about being new to GitHub :) |
|
and to mention that evest folder is kinda usless I think as the event listed by evtest sometimes changes even it is on the same port without any additonal devices connected after removal of the tablet the pen and pen button use interface 0 in which and finally I suspect that this tablet might have tilt but it is not advertised as that but I noticed something in the interface 0 xml file |
Yes. After an unplug and Ready for any more tests. Really appreciate your support in this. |
|
hey @spbnick today I played with the driver and found a thing I forked digimend and when give value to rc with tablets that have report id 0x09 my mouse cursor moves with the tablet to say short I think the cursor moves when the report id is 0x09 but I am not sure and anyways here is the changes I made |
|
From my experiments digging around the |
|
@spbnick finally I made a working prototype of the tablet with working pressure,I need your guidence in this and |
|
Tested out 0468d77 by |
|
@hyiltiz can you please unloaded any configuration i.e made by you (including your wacom configuration)and with only digimend and try again once because the tablet works fine for me and make sure your tablet dosen't use any wacom driver |
|
maybe try this and my dmesg |
|
@spbnick I find the reason that this PR is not working and the modified version in short the original descriptor has report id 7 which has everything needed had usage page as |
How is this .zip different from the previous one I tried? I'll unload my xorg and udev configs, reboot and try one out, but not sure which one is the most recent one at this point. |
In the previous one I deleted some of the things in the report descriptor that I thought that they were useless (To keep it neet ) but this .zip contains full form of the report descriptor with some minor changes and I have tried both in a lubuntu vm and pressure is working |
|
Tested out digimend-kernel-drivers.zip and it works! The pressure is working, and it still works after a replug. None of Krita, MyPaint or Konsole crashes when using the pen. However:
It would resolve #385 and #383. Question: it doesn't seem to work with the X11 libwacom driver, which is probably why it is not listed in My stepsBuilt and installed via |
|
@hyiltiz The tablet can use libwacom with out any issue atleast in my laptop and reboot followed by that edit successfully makes the tablet to use libwacom TO DO's |
|
@hyiltiz added patch for upper button (button far away from the tip ) |
|
Installed the above Once installed that snippet into I'd try simply adding a line in the |
|
digimend-kernel-drivers.zip anyways @spbnick need guidence here @spbnick watched your ugee ex07 video on yt |
|
With my custom udev rules #383, I was able to only export the pen and pad. I then specified the device using device path for libwacom in xorg.conf (also shown there). Pad was recognized as a keyboard device. If the pad is recognized as a keyboard device, we can map it using |
|
is your xsetwacom recognise pad as keyboard |
|
It did without digimend through usbhid hid_generic and libwacom. With this new digimend patch, it doesn't; instead it recognizes 2 stylus, 2 eraser (no hardware for it though) and 2 touch (no hardware). I really hope digimend only exports the stylus and the pad. |
|
@hyiltiz actually I finished the pen stuff(with both buttons remapable) and currently working on the pad and Now I realized why you are having different report string form mine actually the dumps you provided are after using uc-logic probe or somehow you initialized the tablet, you actually enabled full functionality of the tablet and just a thing did you able to map the pad buttons before you trying this PR ie #383 |
|
I did use How did you map the pen keys? Any progress on the pad buttons? |
|
i mapped the pen keys with using xsetwacom --set command to key a and opened kate pressed the mapped button yay I am kindoff confused with the pad buttons tho the pad interface is kinda like keyboard emulation by default but when using UC-Logic probe, it made pad and pen in same interface take a look at #403 within few minutes I'll upload full patch for pen and pen buttons and here is that patch |
|
I was able to enable the lower left pad button without digimend using custom udev rules. |
|
make an evtest on that event and send it here
|
|
@hyiltiz is that tablet return original tilt values or they are junk bytes (I don't know about tilt that much) currently there is no idea for me to make pad working i mean, it should work idk how to implement it, I am missing at somewhere maybe it will be done if @spbnick help us or i should make one from scratch for further things see issue #407 |
|
It doesn't respond to tilt in any meaningful way. Guess "junk bytes" in your terms. |
|
Is there any progress on this? I then ran Anything I can provide you with to solve the tilt jitter and getting tablet buttons config available in UI? |
If you read carefully, the "progress" stopped when we got stylus and pressure working, but not the buttons or tilt (afaik Parblo doesn't have tilt but feel free to show us otherwise). To help, I think it is better if you start looking into |
|
@hyiltiz Thanks for update. and as noted by @DONofDESTROY he actually found a tilt on Parblo A640 (although it seem either buggy or garbage?). Regarding the and in xorg.conf: This didn't help much with |
|
Ok, here goes the wireshark usbpcap dump for frame buttons pressing, dumped from Windows 10. I was pressing buttons in the same order as described in DIGImend tutorial, but went a bit further and tried setting up buttons mapping via DigitizerPanel utility that came with the driver. As far as I could tell the mapping software didn't communicate with the device to set it up. |
|
@eduard-sukharev after you Installed your driver the pad buttons becomes keyboard -> pad buttons which is similar to mouse so when press event is noted the driver virtually press the key assigned and release the button if the key is released xsetwacom remap keys which is like these buttons ie either on or off state not like keyboard with key combos It is an hacky way used by companies to make tablet work with basic functional like mouse and keyboard without driver |
|
Ok, so from what I've understood during my attempts and from what you've said: As far as I can undersrand, we have two possible solutions:
|
|
Digimend is not driver for tablets it makes the kernel not to throw garbage stuff Here is how this worksHow to do this yourself Pen (usbhid-dump before probing i.e. tablet mode without any drivers)Coordinates003:006:000 It is 0th interface
Pen Buttonsyou need to concentrate on the 2nd hex byte i.e. XX
The pen stuff is on the 0 interface TLDR The interface 1 is useless Pen (dump with drivers or after probing)This is why @hyiltiz and @eduard-sukharev getting pen pressure working because they should have used uclogic-probe command before taking diagnostic for the tablet i.e. the tablet is working like a tablet (pen and pad), before this command it works like a mouse and keyboard Pen cords003:006:000 It is 0th interface
note the tilt in this tablet seems to be legit on the Pen buttons
Tablet buttons (before probing or using probe)003:006:002 It is 2th interface
Tablet buttons (after probing)003:006:000 It is 0th interface
The problem is before probing both pen and pad are at seperate interface 0th and 2nd What can we do
I am writing this cause this might be helpful for other tablet users |
No description provided.