Kirill Bulatov

I like the way it's implemented in the pull request (using `%` instead of `debug_assert`), allowing the javascript code to receive the index for any row and column pair specified...

Makes sense, here is the PR: https://github.com/rustwasm/wasm_game_of_life/pull/38

Looks great, thanks for coming with this up. I have nothing to add to the issue message, but the [first comment](https://github.com/launchbadge/sqlx/issues/942#issuecomment-752231796) with the query format proposal looks worth discussing. *...

> So the slow query log at `WARN` should probably be the same format as `DEBUG`. I presume, you have to do the same for `Error` level too then, keeping...

> what arbitrary number is acceptable? What about for a slow-query log? I think you can either make it configurable, or just pick some arbitrary number (100?) and recommend to...

> With this PR, we make a full Postgres build on macOS too. That's great, also nice that we share the same "stricter compiler" flags. Yet, we also run our...

If I understood the proposal right, we're adding a new file? Could name it `timeline.json`, for better clarity. Might also extend the name with version number, or similar metadata. We...

> My theory: when the pageserver receives SIGTERM, it starts the shutdown sequence, but it doesn't immediately kill GC and/or compaction. This it not a theory and might be triggered...

Hard to say for sure what happened without seeing pageserver and safekeeper logs, ideally etcd logs either (you can find those in `.neon/` directory ) and the workload details. I...

I was thinking on making a few benchmarks and maybe a few more path-handling improvements before even starting to think about this myself. Should we also close https://github.com/rust-analyzer/rust-analyzer/issues/4544 ?