ivmarkov

> I left some comments about the implementation, overall it looks sensible. Thank you. > In Rust for Linux, we have the need for [`InPlaceWrite`](https://elixir.bootlin.com/linux/v6.12-rc1/source/rust/kernel/init.rs#L1287) I haven't yet added it...

> > > I took a quick look at the use case of `Apply::apply` and what I find a bit odd, is how the `FromTLV::update_from_tlv` function is implemented. The whole...

> Besides the small c2 thing, i am a bit skeptical about the correctness of the magic numbers we set on the attenuation matcher. E.g the return values we define...

To add a bit more, to me (IMHO) this whole `configure_for_async` pattern also feels out of place in code that embraces the notion of a `Peripheral`, in that it takes...

> I think you should be calling `configure`/`configure_for_async` before `new_foo`, rather than inside it. Though I do see it becoming a trait in the future (with an associated type that...

> I suppose what I'm trying to say there is Channel should be the thing you borrow, not the creator. (You can't do this now but that's what it should...

> > * What if I want to construct a **different** Channel from the original creator, which is no longer "async"? This is currently impossible. > > This will be...

Sounds complex, but I think that would work, at least for my case. Also, the peripherals of the Esp are not simple either...

How much more complex is the explicit variant I'm suggesting: ```rust static RESOURCES: StaticCell = StaticCell::new(); ... void main() { let server = Server::new_with_config(..., RESOURCES::init(Resources::new())); ``` ?

@Volkalex28 Would you like to contribute a PR that addresses this issue?