bevy-quic-networking

A plugin for Bevy which implements components for the use of sending data over the QUIC protocol using Amazon's s2n-quic.

Getting Started

First start by selecting the relevant version of the crate based on the version of Aeronet and Bevy you are using. If you are not using Aeronet you can ignore the Aeronet version.

Bevy Version	Aeronet Version (optional)	Crate Version
0.18	0.19	0.18

You can add the crate to your project with the following addition to your cargo.toml

[dependencies]
bevy-s2n-quic = "0.18"

And by adding the default plugins

let app = App::new()
    .add_plugins(QuicDefaultPlugins)

And for Aeronet functionality you will need the Aeronet plugin

let app = App::new()
    .add_plugins(QuicAeronetPlugins)

You can open a server in the following way:

let server_comp = QuicServer::bind(&runtime, ip, cert_path, key_path)
        .expect("Unable to bind to server address");

    commands.spawn(server_comp);

We can then spawn a client component:

// Spawn a new client with our cert
let mut client_comp = QuicClient::new_with_tls(&runtime, cert_path).expect("Invalid cert");
let connect = Connect::new(ip).with_server_name("localhost");
let conn_bundle = client_comp.open_connection(connect);

// Spawn client with connection attempt as child
commands.spawn(client_comp).with_children(|parent| {
    parent.spawn(conn_bundle);
});

You'll have to query for completed connections. Automatically the default plugins will convert successful attempts into full connection components:

fn client_open_stream(
    mut commands: Commands,
    connection_query: Query<(Entity, &mut QuicClientConnection), Without<Children>>,
) {
    // As soon as we see a connection without any children,
    // attempt to spawn a bidirectional stream on the client end.
    // Both servers and clients may open streams.
    for (entity, mut connection) in connection_query {
        let stream_bundle = connection.open_bidrectional_stream();
        commands.spawn((stream_bundle, DebugCount(0), ChildOf(entity)));
    }
}

AI Usage

AI has been used for the purposes of indexing documentation, generating examples and helping identify under documented public APIs, as well as in assisting in refactors or code snippets. All code is human authored with a human being responsible for maintaining any code which does not behave as intended.

License

Licensed under either of:

• MIT License (LICENSE-MIT)
• Apache License, Version 2.0 (LICENSE-APACHE)

at your option.