Skip to content

[R-pkg-devel] False Error in CRAN checks

5 messages · Ivan Krylov, Simon Urbanek, Josiah Parry +1 more

Original
Josiah Parry
# 
My package arcgisutils has an "ERROR Installation failed" recorded for arm64 macOS (1)

However, if you go to the log (2) it shows ?Status: OK? and the installation log (3) also shows a clean install.

This is affecting my downstream packages arcgisgeocode (4),  arcgislayers (5), and arcpbf (6), which are showing Status OK and clean installs as well.

Is this a ?blip? in the check system? Or am I missing something and am at risk of archival? 

Thanks all! 



1) https://cran.r-project.org/web/checks/check_results_arcgisutils.html
2) https://www.r-project.org/nosvn/R.check/r-devel-macos-arm64/arcgisutils-00check.html
3) https://www.r-project.org/nosvn/R.check/r-devel-macos-arm64/arcgisutils-00install.html
4) https://cran.r-project.org/web/checks/check_results_arcgisgeocode.html
5) https://cran.r-project.org/web/checks/check_results_arcgislayers.html
6) https://www.r-project.org/nosvn/R.check/r-devel-macos-arm64/arcpbf-00check.html
Ivan Krylov
# 
? Wed, 11 Feb 2026 10:17:31 -0800
Josiah Parry <josiah.parry at gmail.com> ?????:
Something's not right with r-devel-macos-arm64. For many packages,
<https://www.r-project.org/nosvn/R.check/r-devel-macos-arm64/PACKAGE-00check.html>
points to outdated results (sometimes a previous version of the
package) or a 404 error, and there's no corresponding binary for
R-devel/Sonoma. Examples include 'bapred', 'aroma.affymetrix', 'b32',
'RPostgreSQL' (approximately 197 packages in total). Try asking Simon
Urbanek about this problem.

At least there's no deadline set right now:

tools::CRAN_package_db() |>
subset(grepl('^arcgis|^arcpbf$', Package)) |>
_$Deadline |> is.na() |> all()
# [1] TRUE
Simon Urbanek
# 
Josiah,

we just enabled the r-devel logs and the websites are not entirely correct yet, because the logs on the website show from the old big-sur build instead of the new sonoma build - which is where the result status is from.

That said, In your particular case it failed in the sonoma build yesterday since it was a manual catch-up run and didn't have the rust compiler (hence the short test time - I have noticed and fixed it at the time), but the subsequent regular run succeeded. Give it another day or so before it all settles.

Cheers,
Simon
Josiah Parry
# 
Thank you so much for the update, Simon! Brilliant <3
Kurt Hornik
# 
Sorry, my fault, I had added x86_64 where I should have added arm64 ...

That's what happens when doing these things along with running a
department etc.

Fixed now.

-k