Peter Johnston

Name server sorting is problematic

9

**Describe the bug** The Trust-DNS resolver [sorts](https://github.com/bluejekyll/trust-dns/blob/main/crates/resolver/src/name_server/name_server_pool.rs#L185) the name servers it's been configured with, by (1) ["connection state"](https://github.com/bluejekyll/trust-dns/blob/main/crates/resolver/src/name_server/name_server.rs#L204), so that name servers that have been successfully reached are prioritized above...

In order to let the `loom::lazy_static` macro work in contexts that are `#[no_std]`, use `marker::PhantomData` from `core` instead of `std`.

extended version of `std::net::parser` does not include CVE fix

4

It looks like the [copied and extended version](https://github.com/krisprice/ipnet/blob/master/src/parser.rs#L158) of the `std::net::parser` module in the ipnet parser doesn't include this recent CVE fix to the standard library that disallows the use...

How to update parking_lot's dependency on parking_lot_core

7

`parking_lot_core`'s dependency on `windows-sys` was [recently updated](https://github.com/Amanieu/parking_lot/pull/331) from v0.33.0 to v0.34.0. What's the process for updating `parking_lot`'s dependency on `parking_lot_core`? I suppose it would require tagging a new release of...

Always retain the route in the tcp and fake transport endpoints on connect

4

**Description** Currently the tcp endpoint and [fakeTransportEndpoint](https://cs.opensource.google/gvisor/gvisor/+/master:pkg/tcpip/stack/transport_test.go;l=149?ss=gvisor) release the route on connect rather than retaining it. Since multiple connect calls are permitted, it might make sense to retain the route...