There are two major issues I think we need to discuss in a principled way for 1.0, whether that means coming up with a solution or coming up with a justification for why not to support these use cases.
I am excluding specialization, bignums, object safety and SerializeTo because all of those have been explored at length already, existing proposals come with significant unresolved downsides, and in my opinion Serde 1.0 is more valuable than solving any one of those issues. That said, I would love to see progress on any of them despite not planning to dedicate much of my own time.
And of course a ton of docs work: https://github.com/serde-rs/serde-rs.github.io/issues
:confetti_ball:
Most helpful comment
:confetti_ball: