[R-pkg-devel] Debugging M1mac problems

Ivan Krylov |kry|ov @end|ng |rom |@@er@chem@m@u@ru
Sat Mar 27 17:11:55 CET 2021

Dirk, thank you a prompt reply!

I see the issue you took part in solving:

Wondering if a combination of environment variables for R CMD check
could have helped reproduce this on another machine. Unfortunately
<https://www.stats.ox.ac.uk/pub/bdr/M1mac/README.txt> doesn't list any
environment variables.

On Sat, 27 Mar 2021 17:55:07 +0300
Dirk Eddelbuettel <edd using debian.org> wrote:

> I think we are now being asked to no longer leave [the temp] files
> behind (whereas in the past we were asked to create them _exactly
> this way_).

I will make sure to address this deficiency of ?write.openfluor in the
next version and avoid leaving files in tempdir() in the future, but
I'm afraid this won't help me fix tests/feemlist.R on M1mac. Or is it
not enough that I clean up all temporary files created in
tests/feemlist.R at the very end by means of unlink(..., recursive =


More information about the R-package-devel mailing list