What's everyone working on this week (41/2019)?

New week, new Rust! What are you folks up to?

I'm working on improving the GPIO pin ergonomics for stm32f1xx-hal.


At work, I'm working on improvements to the ADC API in stm32l0xx-hal.

In my free time, I've started seriously looking at embedding my programming language into a host application (a CAD program) again. The host application needs some types though that the programming language doesn't have at the moment. Now I'm thinking about whether to add a facility to define new builtin types, or if I bite the bullet and finally add the capability to define structs (or something along those lines).

1 Like

I am working on a driver for the Si4703 FM radio receivers.

1 Like

On the home stretch with Tetra 0.3!

I realized over the weekend that I was trying way too hard to fix everything forever in a single breaking release, and it was starting to burn me out. My goal now is to just get 0.3 out ASAP, with:

  • Experimental WASM support (no audio yet, need to implement async asset loading for that to be viable)
  • Some API tweaks/cleanup
  • A simpler/faster renderer

If I have to release a 0.4 in a few months, so be it :stuck_out_tongue:

1 Like

Just released proc-macro-error - the crate that's meant to make error reporting in proc-macros fluent and encourage people to keep span info while awaiting for proc_macro::Diagnostics.

It was built for structopt's sake originally, but I've encountered a lot of crates that do their error reporting via panic! since then. I've asked some acquaintances of mine about reasons why they don't bother with span info and generally the answer was: "What's point to bother about span info if the error reporting is really complicated? I'll think about it when my crate is popular enough, until then a panic will suffice.".

So it's an attempt to make error reporting in proc-macro almost as simple as with panic!


Still working on encrypted DNS server. This week I'll be implementing anonymized DNS that prevents DNS servers from learning anything about client IP addresses.

I may also migrate from tokio to async-std if I ever manage to make it play well with hyper.