Michael Chirico

Impressive work here! Thanks so much. A lot of my comments can probably be punted to future CLs, but OTOH, this will be a new default, so we should balance...

sorry, I don't follow why `length(data.frame()):1` is ok. the reasoning is the same as for `1:length(data.frame())`. The point is the is usually used to index columns, but `1` is not...

Sorry, still not following why the test is wrong

Note some discussion here BTW: https://github.com/r-lib/lintr/pull/1032#issuecomment-1129325440 I think at root we want to detect the value of `perl=` and tweak the logic for what's a fixed regex or not accordingly....

This is what `available_linters(tags = NULL)` is for IINM. We don't include all linters in `default_linters` because `default_linters` is targeted for the tidyverse style guide (this is not strictly true,...

Oh, and to use these, you'd use `linters_with_tags(tags = NULL)` to activate all linters

marked it as a documentation issue -- if a regular contributor is only now learning, we need to advertise it better!

It doesn't look broken to me: ```r remote_users_res = list(list(username = 'a'), list(username = 'b')) message(glue::glue("Users found: {glue::glue_collapse(purrr::map_chr(remote_users_res, ~ .x[['username']]), sep = \", \")}")) # Users found: a, b ```

Oh sorry, I missed it on L91; thought I'd checked both. But L47 you mentioned doesn't have the error.

OK filed https://github.com/rstudio/connectapi/pull/174; re-closing this as invalid, though maybe we could improve the debugging experience here.