0866b739d0
## Proposed Changes Clippy 1.67.0 put us on blast for the size of some of our errors, most of them written by me ( 👀 ). This PR shrinks the size of `BeaconChainError` by dropping some extraneous info and boxing an inner error which should only occur infrequently anyway. For the `AttestationSlashInfo` and `BlockSlashInfo` I opted to ignore the lint as they are always used in a `Result<A, Info>` where `A` is a similar size. This means they don't bloat the size of the `Result`, so it's a bit annoying for Clippy to report this as an issue. I also chose to ignore `clippy::uninlined-format-args` because I think the benefit-to-churn ratio is too low. E.g. sometimes we have long identifiers in `format!` args and IMO the non-inlined form is easier to read: ```rust // I prefer this... format!( "{} did {} to {}", REALLY_LONG_CONSTANT_NAME, ANOTHER_REALLY_LONG_CONSTANT_NAME, regular_long_identifier_name ); // To this format!("{REALLY_LONG_CONSTANT_NAME} did {ANOTHER_REALLY_LONG_CONSTANT_NAME} to {regular_long_identifier_name}"); ``` I tried generating an automatic diff with `cargo clippy --fix` but it came out at: ``` 250 files changed, 1209 insertions(+), 1469 deletions(-) ``` Which seems like a bad idea when we'd have to back-merge it to `capella` and `eip4844` 😱 |
||
---|---|---|
.. | ||
cached_tree_hash | ||
fork_choice | ||
int_to_bytes | ||
merkle_proof | ||
proto_array | ||
safe_arith | ||
serde_utils | ||
ssz | ||
ssz_derive | ||
ssz_types | ||
state_processing | ||
swap_or_not_shuffle | ||
tree_hash | ||
tree_hash_derive | ||
types |