Description

So rough description of what was happening:

• the errors got swallowed but nuked the loop so it failed silently - fixed that
• the 0rtt tries connecting on the new port/addr but still prefers the old path so needs to wait out for the path invalidation to do it's job before it continues to work

Being slow enough made it sometimes pass. Locally now tests pass ~instantly vs 15 seconds or so after the fix to gracefully handle errors.

I'm unsure if this is safe at all, like handling path updates from other angles or partial updates etc.
If it turns out too risky, we can drop the path invalidation changes and just keep the graceful error handling. Tests will pass.

Breaking Changes

Notes & open questions

Change checklist

• Self-review.
• Documentation updates following the style guide, if relevant.
• Tests if relevant.
• All breaking changes documented.
  • List all breaking changes in the above "Breaking Changes" section.
  • Open an issue or PR on any number0 repos that are affected by this breaking change. Give guidance on how the updates should be handled or do the actual updates themselves. The major ones are:
    • quic-rpc
    • iroh-gossip
    • iroh-blobs
    • dumbpipe
    • sendme