speaker notes

bullet points

welcome
- im nils
- talk about how to contribute to the rust project
- esteban rustconf 3 years ago
- you in 3 years!
whoami
- nils, compiler contributors team
- helping people, triaging, shitposting on discord
- 1.72 cfg
others
- reviewed the talk
- important path of why I contribute
- friends reviewing your PRs
start
- some contributed before
- wanted to but never have
- many ways, no matter experience
- some domain-expecific expertise
- some no knowledge of rust
- will learn something
- requirement: time and interest
what to contribute?
- where to get started?
not code
- usual focus on code
- rust also not good, thanks.rust-lang.org
triage
- easy to get started
- variable amounts of time
- labelling, minimization, bisection, adding tests
labelling
- helps with
  - organizing issues
  - finding related issues
- least time
- std document mmeory ordering thread::spawn
- feature request, documentation, threads, atomics, libs api
- -needs-triage +A-docs +T-libs-api +C-feature-request +A-thread +A-atomics
- A- , T-, C-
- use rustbot
- see link
bisect
- helps with
  - finding the root cause of the problem
  - figure out when regression happened -> priorization
- E-needs-bisect (not only)
- ICE easiest
- cargo-bisect-rustc (--access github)
- start date, often last year or last few months
- issue comment
- potential problems
  - feature
minimization
- helps with
  - debugging
  - coming up with a test
- E-needs-mcve
- me in 2 hours to 60 lines, on github
- felix (compiler team) blog post (outdated)
- cargo-minimize
- helps you learn more about rust
- starting point for fixing
adding tests
- E-needs-test
- fixed, no test
- commit test
existing unstable features
- a lot of unstable features
- used some or many of them, a favourite
- goal: stabilize or close
tracking issues
- C-tracking-issue
- over 700
- dont new
- really old -> problems
- middle
summarize
- most important: summarize discussion
- aggregate open and resolved questions
- propose next steps, suggest answers
stabilization report
- no open questions
- proportional to amont of discussion
- team takes a look
close
- use cases haven't proven themselves
- subseeded by other feature
- also summarize discussion
fcps
- team decisions
- majority approves
writing code
- want to write code
- lots of code
official tooling
- very good, but needs help
- different, choose
- fix bugs, implement new features (like lints)
standard library
compiler
- me
- diagnostics (many, many)
- internal compiler error fixes (variable difficulty)
- E-mentor
diagnostics
- my first
- search for issue, fix it
examples
rustc dev guide
looking at PRs
- if unsure
- know what's happening
- learn from others
- do leave comments
to remember
- many volunteers, no guaranteed review time
ask for help on zulip
end
- i hope i was able to encourage some of you to get started
- for others, learn how rust works
- questions

3.3 KiB Raw Blame History

speaker notes

bullet points

3.3 KiB

Raw Blame History