Add key repetition for the wayland backend #628
Conversation
|
This is a simpler implementation of key repetition, based on a thread. It has been developed in #371 that no background thread should be needed. I do have some plans to introduce some real, proper event loop capability to my wayland libs (in a way making it possible to have good support for secondary event sources, including timers). But I don't know when I'll have the bandwidth to do this (this is related to Smithay/wayland-rs#185 for those interested). In the meantime, this works, and leaves room to easily implement stuff like disabling or discriminating key repetition events, depending on what is decided wrt #310. |
Cargo.toml
Outdated
| @@ -55,7 +55,7 @@ features = [ | |||
|
|
|||
| [target.'cfg(any(target_os = "linux", target_os = "dragonfly", target_os = "freebsd", target_os = "openbsd", target_os = "netbsd"))'.dependencies] | |||
| wayland-client = { version = "0.20.10", features = [ "dlopen", "egl", "cursor"] } | |||
| smithay-client-toolkit = "0.2.6" | |||
| smithay-client-toolkit = { git = "https://github.com/Smithay/client-toolkit" } | |||
There was a problem hiding this comment.
I'm going to change this pull request to 'Upgrade to smithay-client-toolkit 0.3.0' as 0.3.0 contains few breaking changes from Smithay/client-toolkit#25
There was a problem hiding this comment.
Sorry my bad, it seems none of those breaking changes apply
trimental
changed the title
trimental
changed the title
|
@francesca64, are there any places where you would like these changes to be reflected in the documentation, if not can I remove 'Updated documentation...' off the checklist? |
|
Oh, that would be fine. The checklist system isn't perfect, though I don't really know if there's a better way... anyway, is this PR ready for merge? |
| // { variables to be captured by the closure | ||
| let mut target = None; | ||
| // { variables to be captured by the closures | ||
| let target = Arc::new(Mutex::new(None)); |
There was a problem hiding this comment.
Default::default() should be equivalent to this.
There was a problem hiding this comment.
let target: Arc<Mutex<Option<_>>> = Default::default();
or
let target = Arc::new(Mutex::new(Default::default()));
There was a problem hiding this comment.
Oh, since the type isn't inferred in this case, then it's preferable the way it already is. Sorry!
|
I believe it should be in a state to merge after implementing your review point. I just want to be sure though that my understanding is correct that there is no difference between key press and key repeat events in winit and that they should be indistinguishable? |
|
That's correct; our API presently has no indication of a key press being a repeat. Anyway, thanks for the PR! |
|
Thanks! I'm sure you guys have your own release cycle but I wanted to mention that I'm looking forward to seeing this tagged for alacritty's sake. |
|
This is going to be included in a new release as part of #630, which will merge as soon as the CI passes! |
|
🎉 |
CHANGELOG.mdif knowledge of this change could be valuable to usersThis PR adds the feature of key repetition to the wayland backend. This PR is requested prematurely so that it may be reviewed before the release of smithay-client-toolkit 0.3.0, currently it points to the git version as a dependency.
I have tested the changes and added a small update to the changelog (might need more explanation and better wording?), however I cannot find any documentation of key repetition (I assume this is just standard behavior).
Please look to Smithay/client-toolkit#16 for further information about the new
map_keyboard_with_repeat()function