refi64

Sounds fine! For a future idea, maybe make just the bottom bar swipe-able for switching instances?

You should put anything using the at symbol inside backticks (``` `@foo` ```), since you ended up pinging two users named `row` and `col`.

As an aside, we've been running this on our production infra for a bit now, and nothing has exploded, which is probably a good thing.

 beautiful. (not sure why codecov is failing, this PR did not touch any Go code files.)

@neolynx whoops didn't notice those, fixed now!

ah I believe those are because the source apt repos themselves changed, so now the mirrored packages also changed. I think I've seen that test fixed before via manual updates...not...

@neolynx I fixed the mirror tests separately in #1244. I did also discover a small bug buried in the other mirror tests (!) and fixed it here, but I left...

CI failures seem to be unrelated? Looks like `r-project` updated their packages: ```diff -Success downloading https://cloud.r-project.org/bin/linux/debian/bullseye-cran40/python3-rpy2-dbgsym_3.5.10-1~bullseyecran.0_amd64.deb -Success downloading https://cloud.r-project.org/bin/linux/debian/bullseye-cran40/python3-rpy2-dbgsym_3.5.10-1~bullseyecran.0_i386.deb +Success downloading https://cloud.r-project.org/bin/linux/debian/bullseye-cran40/littler_0.3.18-2~bullseyecran.0_all.deb +Success downloading https://cloud.r-project.org/bin/linux/debian/bullseye-cran40/python3-rpy2-dbgsym_3.5.12-1~bullseyecran.0_amd64.deb ```

Ohh, that makes sense, when I first stumbled upon this splat behavior I guess I didn't realize it and thought it was just a cool trick for joining arrays inside...

I'm not sure what the progress of this was, but I [merged `master` back on and (successfully) enabled `concurrent_spec.cr`](https://github.com/refi64/crystal/tree/feat/wasm32-fibers) (most of the channel specs also pass, but I'd have to...