Adversarial Testnet is coming!
Sign up →
Code review guidelines
A good pull request:
Does about one thing (new feature, bug fix, etc)
Adds tests and documentation for any new functionality
When fixing a bug, adds or fixes test that would have caught said bug
Do the signatures make sense? Are they minimal and reusable?
Does anything need to be functored over?
Are there any error cases that aren't handled correctly?
Are calls to
functions justified? Are their preconditions for not throwing an exception met? Is the exception it throws useful?
There shouldn't be commented out code.
No stray debug code lying around.
Any logging is appropriate. All
logs should be inessential, because they won't be shown to anyone by default.
Should this code live in its library? Should it live in a different library?
Does the code confuse you? Maybe there should be a comment, or it should be structured differently.
Does a behavior change break assumptions other code makes?