|

linux 32 bit drivers

clauritaclaurita Posts: 8
edited on May 2 2017 in Interactive Whiteboards
I read in several posts about a new release of the linux (kernel) driver, but none of the proposed links are active. Where can they be found?
My actual problem is that the old driver (that I successfully compiled for kernel 4.9 on Debian) doesn't support activboard touch 78 (S/N [[Deleted by Promethean]] and others). All other models work well. The board has VID 28e1 instead of the usual 0d48. The board can be somehow used with the standard hid kernel drivers, but obviously calibration doesn't work and there are serious problems with double click.
Any help will be much appreciated.
Regards
Claudio Laurita
Bologna - Italy

Answers

  • @claurita, we are e-mailing you those links in a separate case we have opened.  We will be using the address used when registering for this support community.
  • clauritaclaurita Posts: 8
    Thank you very much.
    But I need some more help.
    I noticed that there were no changes but the version number in the sources of the kernel module and very few modifications in all the scripts of the driver package. Nothing regarding this particular board (ABTouch, id 28e1:b001).
    In fact, the driver does not recognize the board and is not loaded at all. So I wish to know which is the driver that's supposed to manage this board? The standard kernel hid_multitouch? If so, there are serious problems with click. The board often sends a pen up event even if the finger remains fixed on the board. I haven't seen any proprietary xinput driver. Is there any xinput settings to do?
    Thanks in advance
    Claudio

  • While we understand that the Touch Series ActivBoards are not being recognized by the Linux-based computers, can you confirm for us whether a Windows or macOS-based computer is able to recognize a connection to them?  Even if a connection is not recognized by the ActivManager, does any input made at the ActivBoard result in activity on the computer (e.g. an icon being selected or any movement of the cursor on the computer)?

    Please also confirm for us which version of Debian or other Linux-based operating system is in use on the computers attempting to connect to these boards.
  • clauritaclaurita Posts: 8
    I can't try with other OS. We use a Debian 8 (Sid) based system, kernel 4.9 686-pae, X server 1.19.3.
    As I tried to explain, the board works, being recognized as a HID device (both mouse AND touchscreen, two separate input devices, maybe this is the problem).
    No Promethean driver is loaded, just the standard hid_multitouch module,
    But it has serious problems with double click as almost always sends a fake pen up event (got by xinput test) even if the finger remains fixed on the board, so the double click is not recognized. I tried on four boards of this model, and several PCs, so I can exclude an hardware problem. For the calibration we can use the standard xinput_calibrator and the precision is acceptable, but, obviously, is not permanent.

    I took a look at the new relay framework. If I understand, it's a layer that permits to use the activ tools with this board. Unfortunately, the new qt5 based tools don't work well. activcalibrate exits without any notice or error after the welcome screen. Anyway, also with relay loaded, the double click problem remains. Should relay interact with X11 too, or is limited to the tools? May the old tools work with relay (haven't tried yet)?


  • The standard kernel hid_multitouch is used for our Driver.

     

    If the ActivDriver or ActivManager is not loading after installing the packages sent previously, please look to the instructions listed in Article 1470 from our Support Page for instructions on compiling those installation files manually.

     

    Please do be advised that even with the provided files and article above, the release of Debian described is not supported with our updated software or Driver.

  • clauritaclaurita Posts: 8
    Well, so ABTouch boards use hid_multitouch kernel driver, NOT Promethean kernel driver, in this release too.
    This is a first useful information. Thank you.

    Only for your information, Article 1470 is about Promethean kernel driver compilation, so, basing on the above useful information, has nothing to do with this thread. Please, be more careful when answering, just in case you have, on the other side, somebody who doesn't exactly know what is doing. You may let him/her get very confused. Thanks.

    Step 2 (please give me another useful answer) X11 server part:
    Are ABTouch boards intended to use the standard X11 input driver (evdev) or is there a different preferred xinput driver?

    This is the relevant part of my Xorg log. It's from X Server 1.16.2.901. Nothing changes with 1.19 version. Please confirm that this is the supposed correct behaviour. Thanks.

    [  3432.287] (II) config/udev: Adding input device HID 28e1:b001 (/dev/input/event2)
    [  3432.287] (**) HID 28e1:b001: Applying InputClass "evdev pointer catchall"
    [  3432.287] (II) Using input driver 'evdev' for 'HID 28e1:b001'
    [  3432.287] (**) HID 28e1:b001: always reports core events
    [  3432.287] (**) evdev: HID 28e1:b001: Device: "/dev/input/event2"
    [  3432.288] (--) evdev: HID 28e1:b001: Vendor 0x28e1 Product 0xb001
    [  3432.288] (--) evdev: HID 28e1:b001: Found 3 mouse buttons
    [  3432.288] (--) evdev: HID 28e1:b001: Found absolute axes
    [  3432.288] (--) evdev: HID 28e1:b001: Found x and y absolute axes
    [  3432.288] (--) evdev: HID 28e1:b001: Found absolute touchscreen
    [  3432.288] (II) evdev: HID 28e1:b001: Configuring as touchscreen
    [  3432.288] (**) evdev: HID 28e1:b001: YAxisMapping: buttons 4 and 5
    [  3432.288] (**) evdev: HID 28e1:b001: EmulateWheelButton: 4, EmulateWheelInertia: 10, EmulateWheelTimeout: 200
    [  3432.288] (**) Option "config_info" "udev:/sys/devices/pci0000:00/0000:00:14.0/usb1/1-2/1-2:1.0/0003:28E1:B001.0002/input/input4/event2"
    [  3432.288] (II) XINPUT: Adding extended input device "HID 28e1:b001" (type: TOUCHSCREEN, id 11)
    [  3432.288] (II) evdev: HID 28e1:b001: initialized for absolute axes.
    [  3432.288] (**) HID 28e1:b001: (accel) keeping acceleration scheme 1
    [  3432.288] (**) HID 28e1:b001: (accel) acceleration profile 0
    [  3432.288] (**) HID 28e1:b001: (accel) acceleration factor: 2.000
    [  3432.288] (**) HID 28e1:b001: (accel) acceleration threshold: 4
    [  3432.288] (II) config/udev: Adding input device HID 28e1:b001 (/dev/input/js0)
    [  3432.288] (II) No input driver specified, ignoring this device.
    [  3432.288] (II) This device may have been added with another device file.
    [  3432.289] (II) config/udev: Adding input device HID 28e1:b001 (/dev/input/mouse1)
    [  3432.289] (II) No input driver specified, ignoring this device.
    [  3432.289] (II) This device may have been added with another device file.
    [  3432.290] (II) config/udev: Adding input device HID 28e1:b001 (/dev/input/event9)
    [  3432.290] (**) HID 28e1:b001: Applying InputClass "evdev touchscreen catchall"
    [  3432.290] (II) Using input driver 'evdev' for 'HID 28e1:b001'
    [  3432.290] (**) HID 28e1:b001: always reports core events
    [  3432.290] (**) evdev: HID 28e1:b001: Device: "/dev/input/event9"
    [  3432.290] (--) evdev: HID 28e1:b001: Vendor 0x28e1 Product 0xb001
    [  3432.290] (--) evdev: HID 28e1:b001: Found absolute axes
    [  3432.290] (--) evdev: HID 28e1:b001: Found absolute multitouch axes
    [  3432.290] (II) evdev: HID 28e1:b001: No buttons found, faking one.
    [  3432.290] (--) evdev: HID 28e1:b001: Found x and y absolute axes
    [  3432.290] (--) evdev: HID 28e1:b001: Found absolute touchscreen
    [  3432.290] (II) evdev: HID 28e1:b001: Configuring as touchscreen
    [  3432.290] (**) evdev: HID 28e1:b001: YAxisMapping: buttons 4 and 5
    [  3432.290] (**) evdev: HID 28e1:b001: EmulateWheelButton: 4, EmulateWheelInertia: 10, EmulateWheelTimeout: 200
    [  3432.290] (**) Option "config_info" "udev:/sys/devices/pci0000:00/0000:00:14.0/usb1/1-2/1-2:1.1/0003:28E1:B001.0003/input/input10/event9"
    [  3432.290] (II) XINPUT: Adding extended input device "HID 28e1:b001" (type: TOUCHSCREEN, id 12)
    [  3432.290] (II) evdev: HID 28e1:b001: initialized for absolute axes.
    [  3432.290] (**) HID 28e1:b001: (accel) keeping acceleration scheme 1
    [  3432.290] (**) HID 28e1:b001: (accel) acceleration profile 0
    [  3432.290] (**) HID 28e1:b001: (accel) acceleration factor: 2.000
    [  3432.290] (**) HID 28e1:b001: (accel) acceleration threshold: 4
    [  3432.291] (II) config/udev: Adding input device HID 28e1:b001 (/dev/input/mouse3)
    [  3432.291] (II) No input driver specified, ignoring this device.
    [  3432.291] (II) This device may have been added with another device file.

    Regards

  • Adam K PrometheanAdam K Promethean Posts: 1,464 admin
    Hi claurita,

    Thanks for your reply.

    I've reviewed the previous comments and believe this could be a firmware issue on the ActivBoard.

    Firstly, I want to ensure that you are using the latest ActivDrivers for Linux. The latest 5.16.5 has only just been released, this can be downloaded from the below link. I've compressed these in to a zip file for you:
    https://promethean.box.com/v/5-15-5-32bit-Ubuntu-16-04
    Please ensure that these files are installed in alphabetical order.

    Once this has been installed, connect your ActivBoard, please can you confirm if the firmware version is shown within the ActivManager Control Panel? 
    I suspect that if you do not see your ActivBoard or the firmware that a firmware upgrade for your ActivBoard can fix this issue however the firmware upgrade can only be done by connecting a Windows computer to the ActivBoard. I know that you have mentioned you do not have one of these, would you be able to obtain one if the firmware upgrade is required?

    If you can obtain a Windows computer, please see the attached firmware upgrade instructions and upgrade the firmware of the ActivBoard. 
    Once the firmware has been upgraded, please connect to your Linux computer and see if you experience the same touch interaction issues you had mentioned. 

    If you have any questions please let me know.

    Kind regards,
    Adam
    Promethean Technical Support
  • clauritaclaurita Posts: 8
    Hi Adam
    first of all, thank you for your attention.
    I strongly suspected a firmware issue. I will try ASAP.
    I'm the maintainer of an open source Debian based distribution called ICS, tailored for Italian schools. It's a totally diskless system based on NBD, in some way slightly similar to an LTSP (very) fat clients cluster.
    At the moment I follow about 40 schools, about 300 whiteboards, mostly Promethean and Smart. Only 4 of them are ABTouch, all of the same age, so probably with the same firmware.
    ALL the schools (more then 2000 PCs ranging from P3 to four cores last generation machines) share exactly the same root filesystem, so everything must be totally automatically configured at boot, whiteboards included, of course. The excellent calibration mechanism of your software makes it possible and I strongly hope that your new version may extend it to ABTouch too.
    I already tried a first brute-force attempt to install, as explained in my previous post. It was almost encouraging (even if useless). The board was seen, even if the qt5 part is really problematic (refer to my previous post).
    But the driver upgrade, as you can imagine, is not just a matter of deb install, in such an environment.
    I had to modify some things in your previous software and I have to replicate the process for the new one to be sure that everything is effectively working as expected.
    Just in case somebody could care about it:
    first of all, I made the kernel driver DKMS-aware, so I can upgrade the kernel without getting mad (why haven't you done it yet?).
    Then I had to modify the udev part, as we use systemd and your software is not systemd-friendly (udev rule launches directly promethean.sh, but this leads to timeout issues on slower machines, nobody noticed that?).
    Last I preferred to udev/systemd trigger the launch of the manager, only in response to the connection of the board (cleaner, don't you think?).
    If you think that your developers could be interested in an opinion exchange on these subjects, let me know.
    I'll come back when successfully updated the firmware (going out to look for a windows laptop, what a shame! I hope nobody will see me).
    I'm sure that I'll have something more to ask.
    Regards
    Claudio
  • clauritaclaurita Posts: 8
    Unfortunately, the firmware update didn't solve the problem.
    I didn't succeeded in flashing the firmware under W8 64 bit as the ST DFU driver is not signed. Anyway the utility to put the device in DFU mode worked and I was able to flash the dfu file directly from linux with dfu-util. If you could share the details of the enter-dfu-cmd, all this windows stuff could be avoided.
    I repeat my previous question: is it possible to talk with one of your linux developers?
    Don't you think that the bugs I described in my last post are worth being considered (if you really still care about linux, of course)?
    Regards
  • Hello claurita,

    We would be happy to pass along your suggestions as enhancement requests to our developers, though we do not have a direct line for customers to speak with them. 

    Thanks,
    Michael
    Promethean Support
  • Thank you for your feedback, @claurita; we have directed it to the appropriate teams here for review.

    In the meantime, please look to recently-released Article 1623 from Promethean Support and follow the instructions there for installing the ActivDriver on the system there and let us know whether installing the driver through either of the listed methods help with the connection to the ActivBoard Touch.
  • clauritaclaurita Posts: 8
    OK, it's enough for me.
    I believe that you guys are doing your best and wish to thank you all.
    But I understand that Promethean ltd is not interested in linux market any more.
    It was actually rather clear, as all the linux stuff is deeply well hidden in the whole site.
    Not sure it's a wise idea, but that's it and I'm alone on my way.
    Regards
    Claudio
Sign In or Register to comment.