esp-hal 1.0.0 beta announcement · Espressif Developer Portal

[giscus[bot]]

esp-hal 1.0.0 beta announcement · Espressif Developer Portal

The developer resources in just one place!

https://developer.espressif.com/blog/2025/02/rust-esp-hal-beta/

[maciek134]

Here is a diff of type information of our old pin types, versus the new ones. Everything is now hidden within the type itself, seven generic paramters to zero. This makes storing a array of pins trivial, as they are all the same type!

And so you took out one of the better Rust features out of your hal. How often do you need to store pins in an array? It doesn't sound like maintainable code.

but in our opinion the less generics the better, for reasons beyond just usability;

Then your opinion goes against the majority of the community. If you don't like looking at long types you can use type aliases.

Common cases of useless type info is storing pin information - this is usually not required after configuring the pins and will bloat the complexity of the type massively.

This quote is from your developer guidelines. I don't understand how someone can be working on a hal and don't see value in strongly typed pins. It's not just used for configuration, it provides strong guarantees about pins for the rest of the code. There is a reason the rust-embedded book recommends that.

If my task needs two inputs and one pulled-up output I don't want to have to remember about that - the information can be encoded in the type system and then it won't compile if it's misconfigured.

there are also some nice code size improvements which are really important for small constrained resources devices.

The solution to that is type erasure for release builds. You get strong typing for development and small binaries when you go to flash the device.

I'm not really sure why you went with Rust if you don't see value in strong typing.

[maciek134]

This response on the issue about erasing GPIOs is the final straw for me - I don't have faith in the architecture of this hal if runtime checks are being actually considered for things solvable at compile time.

esp-rs/esp-hal#2572 (comment)