Shane F. Carr

As in https://github.com/tc39/ecma402/pull/958#issuecomment-3085694374, I'm seeking agreement on the following proposed conclusion: > We feel that the proposed behavior is better, but we are skeptical that making this change would not...

TG2 agreed to the above conclusion. https://github.com/tc39/ecma402/blob/main/meetings/notes-2025-08-14.md#normative-dont-add-default-formatting-to-lone-era-957

I filed this issue on test262 because we should add tests, including `"und"` but also things like `"und-Latn"` and `"und-u-hc-h12"` which should all return `"und"` from `.language`

TG2 discussion and approval: https://github.com/tc39/ecma402/blob/main/meetings/notes-2025-04-03.md#add-test-for-intllocale-from-und-4444

I don't think my position is changed relative to the OP - I see no substantive difference between warnings and errors that can be suppressed - I disagree that `#[derive(ULE)]`...

This is not a 1.0 issue though since this affects a util crate. It is 1.0 for utils only.

I think we decided not to stabilize utils for ICU4X 1.0; only component APIs and data file format. I do want to get a decision on this issue for ZeroVec...

Is there something like a Clippy lint plugin framework where `zerovec` adds zerovec-specific lints to Clippy that can be controlled as Clippy lints; then we could add this as a...

Hmm. `#[derive(ULE)]` could add `#[clippy::warn_if_private]` to the struct. It would make Clippy generate a warning `clippy::exposed_private_fields` if there are any private fields, and then `#[allow(clippy::exposed_private_fields)]` would suppress it.

Note: related to this, I have an old draft PR to add `validate_with`: #2733