@sympatron OK. I do not know where I found that, and cannot see it in the sources either. The only thing I found in the comments to 1.17.3.b1 was to switch of the RGB led early. So maybe it lights up for other reasons. A transition on P2 may bring the LED on. But at least the fact, that it switched off indicates the normal boot being started.

Hello, Thanks for reporting this issue. We will investigate it. Cheers, Daniel

@chriseb Which firmware version are you using? The latest release 1.17.3.b1 seems pretty stable.

@bucknall I am able to corrupt the file system pretty easily. Running 1.7.6FW ATM. It seems to be when the program has a file open and an error occurs. What is the file system running on these boards? It doesn't appear to be a journaling file system. It is way too easy to corrupt the file system and have it erase the whole flash. I am developing a BT application that can update any file on the device and really need protection against this type of failure. I'm already doing a double buffer approach with the file writes so I have a good copy of the target file before deleting the original and renaming the file. The issue is resetting the board in the middle of a file write causes any manor of flash corruption. Before someone says "don't do that", I try very hard to design my system to recover from any type of errors. The cardinal rule is that there should be no way to brick a unit doing an update. To that end, I test every kind of failure during updates to make sure the device remains in a operating state.

Thanks for the details on boot mode @jmarcelino, PyCom could put a bit more detail in the docs regarding this - the pin out diagram suggests it should work regardless and that it is working now, not a to be added feature. Regarding the firmware update issue I have resolved it :-) Turns out I was needing a pull-up resistor on the Reset pin - my guess is that noise from the serial pins was enough to cause a reset during the update or maybe I was just lucky the the time I added the resistor. I was also able to recover my second board with the resistor in place adding weight to the pull-up being required theory. Thanks for everyone's help. Cheers Dave