iroh 0.27.0 - Squashing Bugs and Taking Names

Welcome to a new release of iroh, a library for building on direct connections between devices, putting more control in the hands of your users.

We’ve fixed some bugs, including a yucky one that only showed up on long running nodes (thank you fishfolk for making us aware!), adjusted some config options that made it too easy to point production code to our staging relays, and streamlined adding Discovery services to the iroh_net::Endpoint.

☎️ Stay connected plz

An integral part of our hole punching work includes maintaining a connection to one of our relay servers. If our connection to the relay server ever fails, the next time we go to send information to the relay server, we run the connect process over again to ensure we have a connection.

This is great if you are on the sending side of a connection, since if anything fails, you immediately attempt to re-connect to the server the next time you send bytes, but if you are on the receiving end, and something fails…you never attempt to re-connect!

We’ve refactored this code path to attempt to reconnect when there is a connection failure, and to not wait for the node to need to send data before attempting to reconnect.

💡 Sensible config options can go a looooong way

We had a bug that did not select the proper relay server (production vs staging) when building our binaries. So we decided to simplify the logic that does the selection!

This resulted in some minor but breaking changes to the existing API.

iroh-net now exposes a new function named force_staging_infra in iroh-net/src/relay.rs which abstracts the selection logic.

We’ve also removed the TEST_DNS_NODE_ORIGIN and refactored any tests that require a DNS Discovery service to either spin up their own node to publish to or run against already existing infrastructure.

We no longer rely on the test-utils feature or the #[cfg(test)] annotations for determining whether code runs against production or staging infrastructure, but only on the IROH_FORCE_STAGING_RELAYS environment variable being set to a non empty value

For end users this should have no real effect, but if you do rely on any of these, setting the IROH_FORCE_STAGING_RELAYS environment variable when running in CI or running tests will result in the same behaviour as before!

🪄 Easier Discovery

We’ve streamlined adding the “default” Discovery services, when you use the iroh-net crate directly.

Before, in order to enable the different default Discovery services on the iroh_net::Endpoint, you needed to do the following:

let secret_key = SecretKey::generate();
let discovery = ConcurrentDiscovery::from_services(vec![
    Box::new(PkarrPublisher::n0_dns(secret_key.clone())),
    Box::new(DnsDiscovery::n0_dns()),
    Box::new(LocalSwarmDiscovery::new(secret_key.public())?),
]);
let ep = iroh_net::Endpoint::builder()
    .secret_key(secret_key)
    .discovery(Box::new(discovery))
    .alpns(vec![EXAMPLE_ALPN.to_vec()])
    .bind()
     .await

Now, that’s simplified to:

let ep = iroh_net::Endpoint::builder()
    .discovery_n0()
    .alpns(vec![EXAMPLE_ALPN.to_vec()])
    .bind()
     .await

But wait, there's more!

Many bugs were squashed, and smaller features were added. For all those details, check out the full changelog: https://github.com/n0-computer/iroh/releases/tag/v0.27.0.

If you want to know what is coming up, check out the 0.27.0 , and if you have any wishes, let us know about the issues! If you need help using iroh or just want to chat, please join us on discord! And to keep up with all things iroh, check out our Twitter.