Michael Chirico

Not really -- intended logic is to make sure `{` is attached to `function()`, `while ()`, `repeat`, `if ()`, `else`, and a few other cases. Constraining to the top level...

In particular, the top-level `{ ... }` usage is not a violation of the style guide as I read it: https://style.tidyverse.org/syntax.html#control-flow > `{` should be the last character on the...

I could imagine this as being part of the `infix_spaces_linter` -- the linter checks that low-precedence linters _do_ have spaces, and high-precedence linters _do not_. Adding 'help wanted' for now...

Should we just consider "infix-then-linebreak" OK in all cases? Would be easier to implement. Isolating `!` and the others feels like a bit of a judgment call -- I agree...

I don't think `x[\n` is a violation of the style guide -- it's not **surrounded** by spaces, it just has a space on one side. So I am leaning towards:...

The problem IMO is every `mutate()` copies the table again which is quite wasteful in general. For `dbplyr`, it's not as wasteful since those backends are typically lazy. We have...

OK, looks like an allowlist is the way to go then. not tagging for 3.0.0 release either.

I think this can be a more general `translation_linter()` that flags usages likely to cause issues for translations. For example any `stop()` call that uses implicit concatenation to build strings...

Agree the purposes are pretty close, but think they should be kept separate. Users that don't plan on translating their package would have no need for `translation_linter()` (other names could...

right -- the more important point is that `identical(1:10, 1L:10L)`. the performance bit is just demonstrating that there's no performance reason to prefer one or the other either.