You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
### What does this PR try to resolve?
The fact that we don't check for `derive_ord_xor_partial_ord ` almost
bit us in #14663. If we used clippy's default lint levels, this would
have been caught automatically.
Instead of just enabling this one lint, I figured it'd be good to audit
all of the `deny` by default lints.
That is long enough that I was concerned about maintaining it (or
bikeshedding which to enable or disable).
All `deny` by default lints are `correctness` lints and I figure we
could just enable the group.
We normally opt-in to individual clippy lints.
From what I remember of that conversation, it mostly stems from how
liberal clippy is with making a lint `warn` by default.
It also makes an unpinned CI more brittle.
I figured clippy is more conservative about `deny` by default lints
and slower to add them that this is unlikely to be a problem.
As for what existing problems this found,
- Some missing serde functions that would be useful for formats besides
toml (since `toml` never uses borrowed strings)
- Code that behaves differently than the syntax says (a 0-1 iteration
loop)
- A redundant assignment (wasn't even removing `mut`ness
### How should we test and review this PR?
### Additional information
0 commit comments