The mini has 2 buttons, labeled A and B, for switching between “layers”, or banks. And these buttons have LEDs to indicate which layer is selected. Changing layers changes which CCs or notes are sent when you press buttons or turn knobs. But that’s in “normal” mode.
In MCU mode layer switching doesn’t work, the A and B buttons don’t send anything, and their LEDs aren’t accessible.
Right now I plan to use the 2 layers and will also use 4 if the regular buttons to switch between “modes” - 2 layers of amp controls, another mode for EQs, third mode for FX on/off, fourth mode for FX tweaking, both in local and global racks.
I also want the fader to control whatever widget was touched last, for fine tuning if needed.
For visual feedback, I’ll have the A and B buttons, and also the buttons I dedicate to “modes”
The way I did it with the footswitch is different - it doesn’t have any layers, so there I have a gig script which detects long press and then rewrites the channel from which CCs are sent, emulating another controller, to switch between either changing song parts (which I use as presets/patches) or looper control. There’s no visual feedback on the footswitch to indicate which mode it is in except I blink all LEDs when switching between modes happens, and the LEDs behave differently.
So yeah, lack of feedback is a bit of a problem, but maybe not too big of a deal. Lack of banks, on the other hand, is.
I don’t know how controllers interact with extensions, maybe it would be possible for a gig script to intercept the comms and fake channels to emulate banks, too?