Lucas Pardue
Lucas Pardue
Done using `cargo clippy --fix` against Rust 1.63.0
The qlog specs recently landed a change to separate out StatelessResetToken into its own type, that is distinct from https://github.com/quicwg/qlog/commit/c2a302d9bf9b97c36e264ff1db42c14aa6e6705e. This updates the data and quiche, which simplifies some of...
h3: add HTTP/3 datagram RFC version With this change, we add support for advertising the HTTP/3 datagram SETTING value (0x33) defined in draft 11 and RFC-to-be when an h3 Connection...
HTTP/3 request and response message exchanges allow multiple HEADERS frames to be sent, which allows for trailer sections or non-final responses. Previously, quiche has only supported generating a single HEADERS...
Previously, quiche kept all details about the relationship between stream capacity and writable DATA internal. This could lead to awkward situations where quiche would report a stream as writable, `quiche::conn::stream_capacity()`...
The draft focuses on the conventional interactions for client-sent requests and server-sent responses. But there seems to be no consideration about server push. I'm not strongly advocating that signatures should...
In draft 00 it says: >The layout and semantics of the frame payload are identical to those of the HTTP/2 frame defined in [[ORIGIN](https://www.ietf.org/archive/id/draft-ietf-httpbis-origin-h3-00.html#ORIGIN)]. The ORIGIN frame type is 0xc...
I'm sure if this is work the effort but about a year ago [I tried to add HTTP/3](https://github.com/hyperium/http/pull/294) to the `Version` enum and we hit the problem with exhaustive matching....
Fixes #203. Add an HTTP/3 column to server and CDN/PaaS tables. In the interest of simplicity, this column reflects support for the final RFC version of HTTP/3. Since the IETF...
While preparing the WIP #204, I noticed that the answer to the question "TLS still adds an extra RTT; can we fix that?" was written in a pre-[RFC8446](https://tools.ietf.org/html/rfc8446) world and...