[Bug] Mastermind PBC erases saved changes on page

Can you send me this new file? I'd like to try it here.

It's true that all data for all buttons is always present. There's no concept of deleting anything. So yes, if you turn off the IA Cycle setting, that changes, but nothing else will change. The system is supposed to ignore any step related settings once IA Cycle is turned off, but the data is still there.
 
Here's the file! Upon examination, I can see that in the Editor, multiple buttons are activated on the HX One button page, which explains the strange errant PC 19 & 20 messages. The IA Cycle button in the "Mute" position was previously responsible for Off/Pre/Post switching before I removed that button during troubleshooting. So it appears that button's saved IA Cycle state (when it was handling loop bypass) was always in the background even when it was changed from an IA Cycle to a IA on/off button... but it wasn't sending its messages until I imported it (?). Or maybe that button, when it was changed to a preset button, was always sending its "Off" state messages unless it was toggled in the Editor? That would explain why button state would show in the Editor but not on the device, since the Editor is shown to be able to display errant button state when changes to those buttons are made (see the above issue with two buttons on in one button group).

On the device itself, none of the HX One buttons light up, though the PCs are sent from that button in the "Mute" position.

I'm somewhat confident now that somewhere I had made a change to an IA Cycle button that caused this issue in some way, it's the theory that makes the most sense to me. Especially since newly created presets don't exhibit the problem.
 

Attachments

  • empire-strikes-back.rjs
    1.7 MB · Views: 3
OK, can you give me an example of one preset that doesn't work in the current file - what is it doing, and what should it be doing?
 
Okay, I've been through so many steps trying to debug this and find the issue that I've confused myself, so I'm not surprised if anyone else following along is completely lost. So here's a concise order of operations to recreate and observe the issue:

1.) Flash the included file "consistent_issues." Observe that the "Edgy" preset on button 2 sends PC1 and has corresponding button state on Page 4.
2.) Power cycle, and observe that the button state is now missing from Page 4.

3.) Flash the included file "fixer"
4.) Observe that the button state is present, and that the only change I've made is to change the buttons in the Mute position and Switches position on Page 4 into IA Cycle buttons that belong to the same Button Group, and send unique PCs.
5.) Power cycle and observe that the Page 4 button state has survived!

BUT...

After a few power cycles, the problem reared its head again on my device while I was testing this process for repeatability. I'm now less certain than before that my IA Cycle cached state button theory holds any water. What's so strange now is that the state seems to survive multiple power cycles, which is inexplicable to me.

Apologies, this is a troubling and very difficult issue to trace.
 

Attachments

  • consistent-issues.rjs
    1.7 MB · Views: 1
  • fixer.rjs
    1.7 MB · Views: 0
rjmmusic said:
How often does it happen, do you figure? I tried power cycling maybe 20 times and I couldn't get the problem to appear.
Click to expand...

With the "consistent_issues" file, every single power cycle. With the "fixer" file, maybe once every 20 power cycles?

I'll point out that I started a completely new page on Page 9 to control my HX One. Didn't copy any buttons, started them all from scratch. That page didn't give me any issues yesterday or at rehearsal last night. I strongly suspect, based on this, that the problem relates to lingering state data under the surface of the buttons, though I have nothing to prove this outside a hunch.
 
ZachSelindh said:
With the "consistent_issues" file, every single power cycle. With the "fixer" file, maybe once every 20 power cycles?

I'll point out that I started a completely new page on Page 9 to control my HX One. Didn't copy any buttons, started them all from scratch. That page didn't give me any issues yesterday or at rehearsal last night. I strongly suspect, based on this, that the problem relates to lingering state data under the surface of the buttons, though I have nothing to prove this outside a hunch.
Click to expand...
So weird. It seems too specific to be a problem with your hardware. I'll try another PBC off the shelf just in case.
 
Incidentally, I've been using my completely new Page 9 (meaning I overwrote my Page 9 with an empty page using "Copy to Page", and recreated all my buttons from scratch) for a few days and at one gig, and I have not had a single issue with that new page. If my theory is correct, and that latent button state data was interrupting proper function of my IA Cycle buttons, then one possible solution seems to be to swap with an empty page and start over.
 
Interesting. Please let me know if the problem returns (hopefully it doesn't). A problem like that isn't something I've seen often, but I couldn't rule it out as a possiblity.
 
You must log in or register to reply here.
Back
Top