Andrew Kelley

> Why not store the deps in the local zig cache? It is ephemeral anyway, and can be viewed as a cache (because they can be recomputed/re-downloaded). because of this:...

The zig-cache directory is intended to be excluded from source control.

Closing in favor of #20180. I think this use case is solved with a combination of that, plus some follow-up tooling, plus the `--system` flag that is already implemented.

Not only am I reopening this, I'm accepting it. Over time I've decided that the following scenario will be better: * Global cache stores recompressed, immutable packages * Local cache...

I don't think this is a particularly contributor-friendly issue.

I'm not suggesting a tree structure. For starters I expect this to have the same structure as master branch global cache.

What makes it "hacky" and what would a non "hacky" fix do differently?

@klkblake thank you for this breakdown and analysis. This is maddening, and I will need to go through the 5 stages of grief before suggesting a course of action.

I want to point out that format-on-save breaks redo, which is pretty frustrating. Sometimes I do `:let g:zig_fmt_autosave = 0` when working on tricky code just so that I can...

> Please let the users -choose- when they want to format their own code. Don't make decisions for us. Thank you. It's by definition impossible to let users choose the...