James Ide

Will need to think about this a bit. I think the API is good and often how you want to use the lock. Some pros: - Can't mess up by...

What I mean is that there is a more direct path from ```js await lock.acquireAsync(); try { ... } finally { lock.release(); } ``` to ```diff await lock.acquireAsync(); try {...

A simple solution is to upgrade to Babel 6 and rely on the fact that RN depends on Babel anyway and let npm 3 dedupe the dependencies. Another idea is...

Sure, my SN is chocoflyer.

> it's slightly different for my setup, e.g. volta: /Users/kudo/.volta/bin/volta -> /Users/kudo/.volta/tools/image/node/16.16.0/bin/node where it was originally pointed to volta wrapper path and now to final node executable. that should be...

> the first path that is not in $TMPDIR is exactly the same with `node --print ...` That's too bad. Let's just use what you have in this PR then.

Thanks for reporting this issue. From your report it looks like this affects the Classic Updates service (`expo publish`) which is in maintenance mode now the next-generation Expo Updates protocol...

The `loadScript` approach sounds directionally correct to me. One question for the Facebook team that comes to mind is whether to deprecate RAM bundles and focus just on loading strings...

This looks good to me. I believe it's fine to bump the target Android version, which is separate from the minimum Android version, so we'll still maintain compatibility with 4.x....

@gengjiawen The work was happening in this repo: https://github.com/SoftwareMansion/jsc-android-buildscripts. The one thing that comes to mind is testing the new JSC in a lot of conditions (different types of phones,...