Mike Dunston

I put one comment on #19, but it won't impact my testing of the fix. I'll do some testing in the full setup and confirm there are no issues.

This appears to have fixed the caching issue I was seeing, more work will be needed in the dependency which is opening the same file multiple times so that it...

> With the current solution, each individual file descriptor doesn't have its own position in the file. Thinking of a way around this... Yeah, I was thinking of the same...

I think overall it sounds good but do add some locking of some sort to ensure no two threads are directly modifying the `lfs_file_t` concurrently.

Note that this is a perfectly valid use case, multiple file handles on the same file from different threads. It works on other FS types (including SPIFFS). But you are...

> I never had the need for this. Yeah, this would definitely be considered a corner case. > I don't think the HAL should be the right layer to implement...

The issue of atomicity is not a problem in the code in using and is working correctly on SPIFFS and FatFS (SD) as well as on other platforms (Linux mainly)....

@geky it's a small bit of a bug in a library I'm using but it works on all FS types except LittleFS from what I've tested so far. It would...

Yeah, it is an interesting issue for sure. FatFS on esp32 works though when using an SD for storage as long as you flush after writes (something we are fixing...

I'd suggest a plugin approach for the CAN interface as there may not be a consistent protocol between various inverters and/or charge controllers. As an example EPEver (or EPSolar, whichever...