I've updated this comment because the check results changed, even though I didn't change anything.

Originally, the R-CMD-check / ubuntu-20.04 (release) (pull_request) and Autocomment atime-based performance regression analysis on PRs / comment (pull_request) checks failed and the R-CMD-check / windows-latest (devel) (pull_request) check was cancelled.

I was surprised that the R-CMD-check failed, because before doing the PR I had built the package and run R CMD check data.table_1.15.99.tar.gz and it finished with Status: OK.

However, when I looked at the PR again the next day, it said both of the R-CMD-check checks were successful. So now the only failure was the Autocomment check, which tdhock explained is expected to fail for PRs that come from forks.

it is normal that Autocomment atime-based performance regression analysis on PRs / comment (pull_request) fails for PRs that come from forks.

Thanks for the explanation @tdhock.

LGTM! Two small tweaks & we can merge.

@MichaelChirico Thanks for reviewing. I've deleted the test, deleted the file issue_6082_tests.Rraw, and moved the News item from Bug Fixes to Notes.

Thanks @markseeto! You should have received an invite to become a project Member. The main thing this does is give you rights to create branches on this repo directly to facilitate future collaboration -- please do so as it makes the review process a bit easier. Welcome!

@MichaelChirico Thanks for approving and thanks for the invitation.