Jack Grigg

I believe @nuttycom was working on this before going on PTO, and will be picking it up when he gets back.

The amount of mined funds show is also incorrect: it shows the total block reward, not the reward that went to the local miner. This is because the metrics screen...

See #5863 for how I think this could be fixed.

My concern with @daira's approach is that we then have an individual lock per span, and risk deadlocks if the order in which we process the nodes is ever non-deterministic....

#6172 makes significant reductions to memory consumption of wallet scanning. It shouldn't in theory affect this issue because (ACAICT from the above comments) no shielded addresses had been created for...

Benchmarked `codegen-units` on Ryzen 9 5950X. Note that the `-j30` parameter affects the C++ side, not the Rust side (which just uses all 32 threads). ## `./zcutil/clean.sh && ./zcutil/build.sh -j30`...

Ran the same scripts on the same CPU again, but on current `master`: ## `./zcutil/clean.sh && time ./zcutil/build.sh -j30` Note that this does not include the build time of vendored...

Hmm, I had `zcash-inspect` disabled because of a local linker issue. I re-enabled it because parallelising the library and binaries also fixed my linker issue, and it revealed a significantly...

Have you tried using [`MockProver::assert_satisfied`](https://docs.rs/halo2_proofs/0.2.0/halo2_proofs/dev/struct.MockProver.html#method.assert_satisfied) instead of `MockProver::verify`? `assert_satisfied` pretty-prints the errors (except for permutation errors IIRC which haven't had a pretty printer implemented yet) before panicking. `verify` returns the...

@defuse confirmed that they are happy with the progress on this, and the remaining work doesn't need to block 5.0.0.