Raphael Coeffic

@breadoven This is probably something that needs to discussed together with ELRS. It would be much easier if the output is not supposed to be some PPM, but digital as...

Maybe something to consider: we do use oversampling & decimation, which relies on noise to enhance the precision. Before disabling flash prefetch, we should have extensive tests showing the tradeoffs...

> is via the stm...hal.h file the right way to do this `PREFETCH_ENABLE` is used only in `HAL_Init()`, which we don't use at all. For reference: ```c HAL_StatusTypeDef HAL_Init(void) {...

It is *kind-of-the-plan* 😅 The basic idea is to split the code into real-time and UI parts. The real-time would still need to run on an MCU, which would be...

Possibly the same as #6083. Fix has been cherry-picked on `2.11` branch (see 4ddafc1fdfccd74529399ebc1072cc575c335389 and #6283) and will be part of [v2.11.2](https://github.com/EdgeTX/edgetx/milestone/31).

> Great, glad to see I didn't dream it up! :P Did you test the fix? Does it work for you?

Also relevant: https://github.com/EdgeTX/edgetx/issues/6404 @robustini did you re-test? In case the detection is failing for you as well, you might wanna try to firmware I posted in https://github.com/EdgeTX/edgetx/issues/6404#issuecomment-3219672970.

@rickard-von-essen it seems the code coverage for web worker breaks with vitest v3. Can you please have a look? You can see that this PR changes the code coverage numbers,...

> Does not run on MacOS: > > * package name is incorrect 'companion.app', should be 'EdgeTX Companion 3.0.app' > * Missing 'Frameworks', and 'Plugins' folders from bundle (QT, SDL...

> Another case of dynamic ui build not auto sizing like it should. Likely needs an explicit kick which I will look at. I'd suggest tackling this in a follow-up...