On 16 January 2024 at 10:28, Josiah Parry wrote: | Oddly making the change has made CI happy. | https://github.com/R-ArcGIS/arcgisutils/actions/runs/7543315551/job/20534063601 | | It may be that the issue was OS related but I'm unsure since only oldrel for | windows and macos check results are published https://cran.r-project.org/web/ | checks/check_results_arcgisutils.html
Seb solved the puzzle (in direct email to me). It has to do with the fact that _the container_ defaults to UTC. If I add '-e TZ=America/Chicago' to the invocation we do indeed see a difference between r-release and r-oldrel (and I also brought the version string display inside R): edd@rob:~$ for v in 4.3.2 4.2.2; do docker run --rm -ti -e TZ=America/Chicago r-base:${v} Rscript -e 'cat(format(getRversion()), format(as.POSIXct(Sys.Date(), tz = "UTC")), Sys.getenv("TZ"), "\n")'; done 4.3.2 2024-01-16 America/Chicago 4.2.2 2024-01-15 18:00:00 America/Chicago edd@rob:~$ Thanks to Seb for the cluebat wave. Dirk | | | On Tue, Jan 16, 2024 at 9:59 AM Dirk Eddelbuettel <e...@debian.org> wrote: | | | Doesn't seem to be the case as it moderately easy to check (especially when | you happen to have local images of r-base around anyway): | | edd@rob:~$ for v in 4.3.2 4.2.2 4.1.3 4.0.5 3.6.3 3.5.3 3.4.4 3.3.3; do | echo -n "R ${v}: "; docker run --rm -ti r-base:${v} Rscript -e 'as.POSIXct | (Sys.Date(), tz = "UTC")'; done | R 4.3.2: [1] "2024-01-16 UTC" | R 4.2.2: [1] "2024-01-16 UTC" | R 4.1.3: [1] "2024-01-16 UTC" | R 4.0.5: [1] "2024-01-16 UTC" | R 3.6.3: [1] "2024-01-16 UTC" | R 3.5.3: [1] "2024-01-16 UTC" | R 3.4.4: [1] "2024-01-16 UTC" | R 3.3.3: [1] "2024-01-16 UTC" | edd@rob:~$ | | Dirk | | -- | dirk.eddelbuettel.com | @eddelbuettel | e...@debian.org | -- dirk.eddelbuettel.com | @eddelbuettel | e...@debian.org ______________________________________________ R-package-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-package-devel