EEPROM needs a magic sigil #559
Description
I am describing a potential solution to my understanding of an issue customers are running into. My solution may be flawed. As might my understanding of the customer issue.
Customers are flashing Chrysalis 0.33+4x or Kaleidoscope HEAD firmware onto their keyboards and the keyboards are appearing to be "bricked" - What's actually going on is that the EEPROM keymaps are empty, but the bits we're reading out of EEPROM to decide whether to ignore the builtin layers tell Kaleidoscope to ignore the builtin layers. This results in the apparently-bricked firmware.
We're being too trusting of the data we get out of EEPROM. I think we need to have some kind of flag in firmware to say "Yes, I have real data here, not random garbage" -- It probably only needs to be a single byte. But it needs to not be all 0 or all 1 ;)