Mike

@lawrence-jeff For now you'll need to pull in the develop branch manually, but once 4.2.0 is released this step won't be required.

But you're still in the administrator command prompt. If you open a new (non-admin) command prompt, SMING_HOME should be correct. Can you check?

That's an old version! Can you dump the entire choco log? Might shed some light on what's happening...

OK, I see where I've gone wrong. I assumed that the latest package versions would be used, but they need to be explicit. I'll do some updates and drop you...

I've been chewing on this one a little. My current thinking is to get rid of `sming.core` entirely, just add that as a manual step. That way, the installation only...

It's also a dependency for ESP8266 lwip2, pulled in as a sub-submodule from https://github.com/d-a-v/esp82xx-nonos-linklayer.git

Though I've not had the same issues on my dev. machines. Not sure why it's happening in CI.

@kmihaylov You've highlighted the problem very accurately! The logic in `spiffs_mount_internal()` (`spiffs_sming.c` line 114) is too simplistic - ``` uint32_t dat; bool writeFirst = false; flashmem_read(&dat, cfg->phys_addr, 4); //debugf("%X", dat);...

Happy that's working for you! The `DISABLE_SPIFFS` only controls whether an image is built and uploaded during flash. >SPIFFS magically inits the FS on the fly Pretty much! It doesn't...

This is probably an unrealistic test, for the flash to get in that state would probably indicate other problems so perhaps not realistic. For more robust operation, try setting `SPIFFS_USE_MAGIC=1`...