• Gliss

  • Issue updating Gliss

@JosH can you try running ioreg -p IOUSB -w0 -l in a terminal on your computer both with and without Gliss connected and paste here what you get in the two cases?

Resolved the problem, see below.

See post below - resolved due to a bad cable.

BTW, you might like to see the short version too (with only ioreg -p IOUSB), which gives this with the Gliss plugged in.

joesh@Joeshs-MacBook-Pro-2 ~ % ioreg -p IOUSB
+-o Root <class IORegistryEntry, id 0x100000100, retain 17>
+-o AppleUSBXHCI Root Hub Simulation@14000000 <class AppleUSBRootHubDevice, $
+-o Bluetooth USB Host Controller@14300000 <class AppleUSBDevice, id 0x100$

Right, so it seems it doesn't get recognized at all. Any chance you could try a different cable and/or different port on your computer?

Okay, have done Giuliomoro, but just hunted down another cable and the GLiss now shows up (using - ioreg -p IOUSB). Thought I post this here, in cas other users had problems.

Here's the output now (looks healthier):

+-o Root <class IORegistryEntry, id 0x100000100, retain 17>
+-o AppleUSBXHCI Root Hub Simulation@14000000 <class AppleUSBRootHubDevice, $
+-o Bluetooth USB Host Controller@14300000 <class AppleUSBDevice, id 0x100$
+-o Gliss-v1.2@14100000 <class AppleUSBDevice, id 0x100000cf9, registered,$

Just went back to the Gliss update, and everything worked fine. I'm not sure what's wrong with the other cable as I use it all the time and never had a problem. However, I guess I need to change cables as maybe it's fine for Teensy, but not for MIDI application(s).

Anyhow, all's well that ends well.

...heading over from modwiggler:

my console:
T: Bus=01 Lev=01 Prnt=01 Port=02 Cnt=03 Dev#= 19 Spd=12 MxCh= 0
D: Ver= 2.00 Cls=02(commc) Sub=02 Prot=00 MxPS=64 #Cfgs= 1
P: Vendor=be1a ProdID=eee3 Rev=02.00
S: Manufacturer=AIL
S: Product=Gliss-v1.2
S: SerialNumber=204E307F4856
C: #Ifs= 2 Cfg#= 1 Atr=80 MxPwr=98mA
I: If#=0x0 Alt= 0 #EPs= 0 Cls=01(audio) Sub=01 Prot=00 Driver=snd-usb-audio
I: If#=0x1 Alt= 0 #EPs= 2 Cls=01(audio) Sub=03 Prot=00 Driver=snd-usb-audio

I meant the Chromium Javascript console. If you are running this as a snap there may be some extra step to give it access to MIDI devices?

This is all the content from chromium console while connecting the gliss and trying to proceed by cicking the continue- button

[Violation] 'setTimeout' handler took 62ms
sysex-flasher-gui-index.js?1728638112:7030 Laden von Abrufen beendet: GET „https://bela.io/files/list.json?r=0.4771937424900321“.
fetchFirmwareFileList @ sysex-flasher-gui-index.js?1728638112:7030
(anonym) @ sysex-flasher-gui-index.js?1728638112:8576
(anonym) @ sysex-flasher-gui-index.js?1728638112:1741
callWithErrorHandling @ sysex-flasher-gui-index.js?1728638112:1374
callWithAsyncErrorHandling @ sysex-flasher-gui-index.js?1728638112:1381
hook.__weh.hook.__weh @ sysex-flasher-gui-index.js?1728638112:1726
flushPostFlushCbs @ sysex-flasher-gui-index.js?1728638112:1526
render @ sysex-flasher-gui-index.js?1728638112:4081
mount @ sysex-flasher-gui-index.js?1728638112:2349
app.mount @ sysex-flasher-gui-index.js?1728638112:5793
(anonym) @ sysex-flasher-gui-index.js?1728638112:8644
js?id=G-9LZETSC8VZ:177 Abrufen konnte nicht geladen werden: POST „https://region1.google-analytics.com/g/collect?v=2&tid=G-9LZETSC8VZ&gtm=45je4a90v9128679380za200&_p=1728663440300&gcd=13l3lPl2l1l1&npa=1&dma_cps=syphamo&dma=1&tag_exp=101671035~101686685&cid=421240020.1728657011&ul=de-de&sr=1280x1024&ir=1&uaa=x86&uab=64&uafvl=Chromium%3B129.0.6668.58%7CNot%253DA%253FBrand%3B8.0.0.0&uamb=0&uam=&uap=Linux&uapv=5.4.0&uaw=0&are=1&frm=0&pscdl=noapi&_eu=EA&_s=1&sid=1728663034&sct=2&seg=1&dl=https%3A%2F%2Fbela.io%2Fupgrade-gliss&dt=UPGRADE%20GLISS%20-%20Bela.io&en=page_view&_ee=1&tfd=5858“.
Lc @ js?id=G-9LZETSC8VZ:177
EG @ js?id=G-9LZETSC8VZ:599
(anonym) @ js?id=G-9LZETSC8VZ:603
(anonym) @ js?id=G-9LZETSC8VZ:602
setTimeout
(anonym) @ js?id=G-9LZETSC8VZ:602
(anonym) @ js?id=G-9LZETSC8VZ:603
(anonym) @ js?id=G-9LZETSC8VZ:620
(anonym) @ js?id=G-9LZETSC8VZ:619
(anonym) @ js?id=G-9LZETSC8VZ:616
Uk @ js?id=G-9LZETSC8VZ:335
tl @ js?id=G-9LZETSC8VZ:341
(anonym) @ js?id=G-9LZETSC8VZ:616
c @ js?id=G-9LZETSC8VZ:471
ku @ js?id=G-9LZETSC8VZ:471
(anonym) @ js?id=G-9LZETSC8VZ:616
c @ js?id=G-9LZETSC8VZ:630
Nm @ js?id=G-9LZETSC8VZ:354
(anonym) @ js?id=G-9LZETSC8VZ:356
(anonym) @ js?id=G-9LZETSC8VZ:354
config @ js?id=G-9LZETSC8VZ:516
Uy @ js?id=G-9LZETSC8VZ:526
Wy @ js?id=G-9LZETSC8VZ:528
setTimeout
F @ js?id=G-9LZETSC8VZ:173
Yy @ js?id=G-9LZETSC8VZ:531
(anonym) @ js?id=G-9LZETSC8VZ:695
c @ js?id=G-9LZETSC8VZ:688
(anonym) @ js?id=G-9LZETSC8VZ:690
(anonym) @ js?id=G-9LZETSC8VZ:690
(anonym) @ js?id=G-9LZETSC8VZ:701

One problem flagged:
A form field has an id or name attribute that the browser's autofill recognizes. However, it doesn't have an autocomplete attribute assigned. This might prevent the browser from correctly autofilling the form.

To fix this issue, provide an autocomplete attribute.

How did you install Chromium? What's your OS version?

Pardon, forgot to mention: I'm pretty sure it's snap-based and OS is Ubuntu 20.04.
I was a bit puzzled about webMIDI not working in this case as it had worked flawlessly in numerous other cases before.

    Can you load https://arachsys.github.io/webmidi/sysex ? See if you can see the device lists and select Gliss as both input and output. Then send

    f0 00 21 3e 00 05 06 f7

    and verify that you get this response:

    f0 00 21 3e 00 05 07 02 47 6c 69 73 73 2d 76 32 2e 30 2e 31 00 f7

    As an alternative, I assume most Chromium-based browsers that are not sandboxed would work. We tested Brave,for instance, with success, see if you have something installed that works.

    PS: Also Firefox works on some platforms. On Linux it doesn't because it doesn't re-detect the device after it resets it, but if when it hangs the first time you close it and restart it without unplugging the device it may get you to the end of the process.

    Thanks, I will try the verify. EDIT: it does not work ("Web MIDI not supported or port access refused") although I basically allowed anything in the config menu of chromium. Brave browser install alas not possible at the moment (do not have the resources to meet its system requirements).
    Updating via firefox (having read your modwiggler-post) does start, but fails at "reboot to application". The whole module is now steadily lit red (apprently only the flasher has been written on it). EDIT: it still drops out at "reeboot to application", but having done the procedure overagain once more, it seems to have witten the 2.0 and is blinking green-red (button) and red-orange(strip). Can I disconnect it, or will it be bricked?

      When it blinks yellow-red it is correctly flashed, but it is in factory test mode.
      Enter the menu by pressing the button and placing two fingers on the touch strip, press the green selector and you'll be in control.mode
      Press the button to exit the menu.

      crowhurst EDIT: it does not work ("Web MIDI not supported or port access refused")

      Well this should mean that webmidi doesn't work on this browser - in its current configuration and with the current system permissions - so that's at odd with your earlier statement

      crowhurst I was a bit puzzled about webMIDI not working in this case as it had worked flawlessly in numerous other cases before.

      That may have happened with a different system/browser/config? If you go back to those flawless earlier cases, I am pretty sure they won't work with the current system/browser/config.

      although I basically allowed anything in the config menu of chromium.

      I know nothing about snap and almost nothing about sandboxing in general, but it makes sense that any permissions you may set in the software won't affect its system-wide capabilities: it's about the OS giving permissions to the software besides anything you may set in the software itself. It may be hell though and one may be better off using a browser with system access instead.

          @crowhurst I think we were able to reproduce your issue and that it is related to your browser being a sandboxed snap. We pushed an update that detects that and warns the user. The easiest solution is to get another browser that is not sandboxed. It may be possible to work around the limitation some other way but it's a system specific hell

          hey giuliomoro,
          thanks a lot for looking into this! I understand that snap-related chromium won't be sufficient.
          However, with firefox I managed to get through the procedure until button blinks red-green and strip blinks red-orange. That is the state it is in now, when powered up via USB.
          Twofingers-on-strip plus Button-pressing enters several states one can cycle through by pressing the button again and again. For example a white dot wandering upwards, a yellow-lit strip, a red-blinking lower part of the strip, a yellow upper part of the strip and some more cycles, until it goes back to that green-red (button) orange-red (strip) blinking-state.
          I wonder, what state this is.. has Gliss been updated successfully and if so, how can it be set to working order?
          Thanks again!

            crowhurst Twofingers-on-strip plus Button-pressing enters several states one can cycle through by pressing the button again and again.

            Once it's flashed in firefox the module is in factory test mode. A description for that is available in the manual. Here you have to enter the menu as usual by pressing the button and placing two fingers on the touch strip (start with the button then add the fingers, the order is important). The menu in this case only has the green selector for changing mode. Tap that and it will take you to control mode. Press the button to exit menu as usual. You may still want to perform the calibration, see the manual for instructions on calibration mode.

            The v2 manual is available here https://github.com/BelaPlatform/Gliss/blob/main/GLISS_user_manual-2.pdf

              Thanks - did as you said: leaving factory test mode manually after firmware upload via firefox and applying calibration afterwards.
              As Slider&Touch is now accessible option on both updated Gliss units, I think it's fair to say the 2.0 upload can be considered successful!
              Thank you for your support!. In hindsight the "work-around process" was solid and easy. Please understand my temporary anxiousness. Since Gliss is a most valuable module for me, I'd have been gutted having it bricked by user error.
              Thank you for 2.0! Super valuable upgrade whilst not cluttering up the UI with hundreds of additional features - great effort!

              If the firefox-bug only means that one has to do the web-udate twice leaving the module connected and manually exit factory test mode at the end, I think it would be fair to claim that it basically works as well. With help of a little demo-video I think any firefox user could be enabled to do this update. Maybe that would be less hassle than getting the mozilla-crew to finally fix that auto-reconnecting-bug of firefox.

              Can't shoot such a video at the moment but if it helps I can report back over at modwiggler that firefox did the job in my case, illustrating the procedure.
              Cheers and thanks again!