Rust GSoC and internship project brainstorming


Hi there. Mozilla will be applying for GSoC this year. There’s a wiki page up for brainstorming project ideas here. I’ve seeded it with some of my own ideas but I’m interested in hearing others.

What would make a good internship project?

GSoC is short and pretty hands off. A good project needs to be deliverable by a motivated student in 2 months. Projects with minimal dependencies and multiple measures of success are great. Internships are great for wild experiments that don’t need to pay off.

If you are interested in Rust cross-compilation, verification, tooling, or performance and are looking for an internship next year, there may be fun projects on Rust at Mozilla. Don’t hesitate to apply, either for GSoC or a Mozilla internship. Mozilla internship interviews are happening now. Apply here.


I think a good project for people interested in verification would be prototyping Rust VC(verification condition) generator, like Jessie (for C) and Krakatoa (for Java) in Frama-C. One could start with either VC for lack of out of bound indexing, or VC for lack of arithmetic overflow.


Glad to see you already have these:

  • reproducible builds
  • libstd::sys refactoring


What happens if an accepted GSoC project happens to get implemented by the community before summer?


I’ve run the GSoC mentoring org for Ruby for the past few years, but I’ll be stepping back from that this year. I’d love to help mentor a Rust project!


Not sure! I guess they would have a lot of free time.

Cool! I wonder if there are any 2-month crypto projects that would be worthwhile.

@sanxiyn thanks for the neat idea.

I’ve added a wasm project too.


Shall those projects also be extended to RGSoC, Outreachy and similiar or is that a different topic?


Yes, I think it’s worth thinking about all internships at once.


My understanding is that other languages like Python have “umbrella” GSoC projects because there are many small projects written in that language that can’t justify their own org admin for GSoC, but that could desire to mentor a few students. Do you think it makes sense to make possible the same for Rust projects? This could greatly broaden the scope of available project ideas, and the benefits of the program across the wider Rust ecosystem.