Paul Colomiets

[WIP] Initial implementation of async resolver

2

This is a very dumb implementation, i.e. it doesn't implement timeouts. But it works in simple cases. It can't be merged yet because it relies on tokio-rs/tokio-core#35 being merged in,...

Any plans for async DNS

1

Hi, Any plans to integrate asynchronous DNS resolver into this crate? I'm looking forward to port `rotor-dns` to futures, and I'm wondering should I contribute to this library or create...

WIP: Implements `owning_iter` method

1

This method might be used to iterate over the owned value producing new OwningRef'erences. I don't see any other way to produce multiple OwningRefs by iterating over something (except calling...

The new version of rotor

1

I've just pushed first working version of rotor-http working on top of the new rotor. It now provisions all of the gory details of HTTP (1.x) protocol (except perhaps websockets)....

Things left after #882 1. [ ] Monitor `edgedb.toml` at least for edgedb version (probably just warn) 2. [x] Use new error API (edgedb/edgedb#4625)

Is the service trait still needed?

4

As `FramedIo` now `Stream + Sink` the server `Service` is also basically `Stream + Sink`. Here is how I can make a server example (also note I don't need any...

Resolve NN_CONFIG_UPDATES address from config service itself rather than from environment

1

Makes less things to configure, and makes configuration service more future-proof (in case better protocol invented for updates instead of non-reliable pubsub)

We has always wanted to expose bindings under `edgedb` crate instead of the `edgedb-tokio/whatever`. But the structure of that is not very clear. Here is a sketch of my current...