Philippe Cholet

**Update on the lint idea:** After discussion on zulip, [`missing_trait_methods`](https://rust-lang.github.io/rust-clippy/master/#/missing_trait_methods) _might_ accept a configuration (in `clippy.toml`) instead of a new lint that would be too specific to a single trait...

@phimuemue Oh... I find embarrassing I was aware about `--message-format json` for "check" (never used) but not for "clippy". After some tinkering on [jq playground](https://jqplay.org/), I think I have a...

@luander Can I help you more on `Combinations::fold`? Alternatively, I suggested `ZipEq::fold` but if you want to work on `Combinations`, specialize `nth` would be appreciated (it would solve #301 and...

@jswrenn A lint `missing_iterator_fold` was not particularly well received: too specific to iterators basically. The main alternative: `clippy::missing_trait_methods` could be configurable to limit warnings (6538 right now): # clippy.toml missing_trait_methods...

> I tried this out and, although it works, it is markedly slower than the unspecialized implementation @kinto-b We won't specialize it if it's slower (but that alone is a...

I know it's different but it's quite similar to `it.tuple_windows::()` which can provide more than one previous item.

I don't see a situation where I would prefer `with_prev` over `tuple_windows`. Do you have an example in mind? Following my previous comment, considering how `tuple_windows` is implemented, it might...

I'm currently going through all issues, this is similar to #319.

> Who calls the shots 'round 'ere? I would say, in that order: bluss, jswrenn, then phimuemue and (since a week) myself. bluss is not around anymore, jswrenn and phimuemue...

I guess `with_next` would be valuable too. About types, I guess our options are (N meaning might change between types) - `(Option, T)` Not generalized. - `([Option; N], T)` No...