# Winit meeting
Date: 2025-03-28T06:00:00Z (UTC)
Attended: @madsmtm, @daxpedda, @kchibisov
Last meeting notes: <https://hackmd.io/rf0BtP-9QrGOrA-oMnCelg>
## Next meeting time?
2025-04-25T10:00:00Z (UTC)
(Note: Past daylight savings time)
## Since last time
- `keyboard-types` got closer to integration into Winit.
- @madsmtm: Write and publish call for maintainers.
- Not done.
- @kchibisov: Cursor -> `dyn` and `EventLoop` -> `dyn`, to move forwards with API rework.
- @madsmtm: Review the above.
- Cursor done, Icon mostly done.
- @daxpedda: `wasm-bindgen` status?
- Was on vacation, but starting today.
- @madsmtm: [`RedrawRequested` iOS](https://github.com/rust-windowing/winit/issues/3406)
- Not done.
## John's maintainer status
https://hachyderm.io/@notgull/114152193300248947
## [Nominated](https://github.com/rust-windowing/winit/labels/C%20-%20nominated) issues and PRs
### [DPI `no_std`](https://github.com/rust-windowing/winit/pull/4175)
Options:
1. Breaking release.
2. "optimal" feature setup, break users who did `default-features = false`
3. suboptimal feature setup, don't break ^
4. Copy `libm` implementation
Decision: Copy `libm` implementation (without `intrinsic` optimization).
Links by @daxpedda:
https://github.com/rust-lang/libm/blob/libm-v0.2.11/src/math/trunc.rs#L8-L12
https://github.com/daxpedda/web-time/blob/main/.github/workflows/upstream.yaml
https://github.com/daxpedda/web-time/blob/main/.github/upstream-sources.json
## Food for thought about `ratatui`
https://old.reddit.com/r/rust/comments/1j5lhfd/request_for_community_feedback_adding_ratatuirun/
## [Return immediately from `run_app` on web](https://github.com/rust-windowing/winit/pull/4165)
@daxpedda's desired solution: `async`.
And his preferences:
1. Revert back to not returning, requiring a solution for Android (see also [#2709](https://github.com/rust-windowing/winit/issues/2709)).
2. No cross-platform API.
3. Cross-platform API with different behavior.
@kchibisov: You need some extensions for web anyhow.
@madsmtm: Downstream libraries (Bevy, Xilem/Masonry, ...) strongly prefer to provide a single entry-point. Providing it in Winit allows us to clearly document this difference there.
Conclusion: @daxpedda will talk to @MarijnS95 about Android and `process::exit`, otherwise we'll go @madsmtm's PR.
## `pointer-events`
WIP: https://github.com/waywardmonkeys/pointer-events
Like `keyboard-types`, but for the W3C pointer events.
Motivation: Masonry depends on Winit for event types only.
@kchibisov: `winit-core` will be POD + traits.
Conclusion: Let's wait and see, generally in favour of splitting into smaller parts where it makes sense for downstream consumers.