Closed xudongzheng closed 2 years ago
Will this have an impact on circuitpython UF2 builds? I suspect they use the same family. There are no Bluemicro833 in the wild but I am not sure about the pca10100. Is this related to ZMK or other firmwares?
If this is merged in, changes to the CircuitPython builds will be needed. See makefile here @tannewt comments on how to proceed?
I think there would be a case for keeping 0xADA52840 if there are lots of nRF52833 boards with the existing bootloader out there.
I have some plans to run ZMK on nRF52833 but other stuff as well.
Wouldn't CircuitPython already be building different firmware for nRF52833 compared to nRF52840? I'm not too familiar with it. Is the Python script also deployed via UF2?
yeah, it should use different uf2 family to prevent incompatible updatring, before merging this PR. you need to submit an PR to uf2 specs to update https://github.com/microsoft/uf2/blob/master/utils/uf2families.json .
Note: I am not sure if uf2 maintainer would agree to use ada52833
, if not you could just generate a random number.
I agree we should get a new number. Now is the time to make it right and get the changes done here and at other places (CP). There aren't too many of these nrf52833 out there yet. And those people who have them would more likely have a dev kit where they can update the bootloader easily.
I've changed the UF2_FAMILY to 0x621e937a
.
Per my understanding, the logic in
src/usb/uf2/ghostfat.c
will prevent nRF52840 UF2 firmware from being flashed to nRF52833 and vice versa (so long asuf2conv.py
is invoked with the correct UF2_FAMILY token).