Dario Nieuwenhuis

It'd be nicer to have a single defmt-decoder version support multiple wire format versions. If it's one-to-one, bins can hack it by depending on multiple defmt-decoder versions at once, yes,...

+1 for removing peripheral singletons and making all register access unsafe. I've found they have a few severe downsides, I've written about them [here](https://github.com/embassy-rs/chiptool/#no-owned-structs). Embassy's HALs use this approach and...

there's a WIP PR https://github.com/embassy-rs/embassy/pull/2113

https://github.com/embassy-rs/nrf-softdevice

I was actually thinking porting the book to embassy-nrf some months ago (I ordered a microbit v2, then the plans fell apart because the store decided to take my money...

we should be careful about how this interacts with the embedded-hal traits. There's 2 SPI traits, [SpiBus and SpiDevice](https://docs.rs/embedded-hal/1.0.0-rc.2/embedded_hal/spi/index.html). SpiBus must be implemented only when no CS pin is being...

This should now be fixed for all timers/chips after the last timer refactor and fixes in the above linked PRs. If you find a timer that doesn't work due to...

> If I'm not mistaken, the executor would treat time spend during an interrupt (e.g. all the tasks in an InterruptExecutor) as idle time, because the interrupt handler is executed...

i'm marking as draft, please mask as ready to review when it's no longer WIP.

yep! https://docs.rs/embedded-io-async/latest/embedded_io_async/trait.Write.html