Skip to content

Latest commit

 

History

History
75 lines (55 loc) · 4.25 KB

cran-comments.md

File metadata and controls

75 lines (55 loc) · 4.25 KB
Raw

Patch for policy violation

  • The package used rappdirs::user_cache_dir for determining the correct cache development. This has now been replaced with tools::R_user_dir('comtradr', which = 'cache').

  • The package left a file that resulted from a faulty test in the cache directory. This is not happening anymore.

Further comments

  • I understand that the previous version created some additional files that were annoying to the CRAN team. Sorry about that. However, I would like to make a few comments that I think might enhance the process for future submitters. Please take these comments in the light of my relative ignorance concerning the underlying processes.
  1. The documentation of tools is wrong. In the help page for the package it specifies rappdirs as a valid tool to specify cache directories. See: https://stat.ethz.ch/R-manual/R-patched/library/tools/html/userdir.html This almost certainly will lead others astray.

--> Maybe you could use your authority as CRAN to let the R-Core Developers know that this is in fact wrong. I have already done the same for the people maintaining rappdirs and they have changed their documentation.

  1. I do not exactly understand why the package was accepted previously. It adds to the mystery of CRAN that seemingly not all tests that could lead to the removal of a package are done all the time, but just sometimes. Now people have already adopted the package in a previous iteration (version 0.4.0) which also includes the "wrong" cache directory.

--> Is there a possibility to run tests that would trigger a removal consistently for all submissions?

  1. Could it be advisable to restrict write-permissions for users on the CRAN servers? If packages are not allowed to write to the cache directory specified by rappdirs, maybe write-permissions could be restricted so that packages error out instead of creating clutter that the CRAN team has to sweep up after.

As per usual: Thanks for your work!

Test environments

  • local OS X x86_64-apple-darwin23.2.0(64-bit), R 4.3.3
  • 1 🖥 linux R-* (any version) ubuntu-latest on GitHub
  • 2 🖥 macos R-* (any version) macos-13 on GitHub
  • 3 🖥 macos-arm64 R-* (any version) macos-latest on GitHub
  • 4 🖥 windows R-* (any version) windows-latest on GitHub
  • 5 🐋 atlas R-devel (2024-05-12 r86534) Fedora Linux 38 (Container Image)
  • 6 🐋 clang-asan R-devel (2024-05-12 r86534) Ubuntu 22.04.4 LTS
  • 7 🐋 clang16 R-devel (2024-05-11 r86532) Ubuntu 22.04.4 LTS
  • 8 🐋 clang17 R-devel (2024-05-11 r86532) Ubuntu 22.04.4 LTS
  • 9 🐋 clang18 R-devel (2024-05-11 r86532) Ubuntu 22.04.4 LTS
  • 10 🐋 donttest R-devel (2024-05-11 r86532) Ubuntu 22.04.4 LTS
  • 11 🐋 gcc13 R-devel (2024-05-12 r86534) Fedora Linux 38 (Container Image)
  • 12 🐋 intel R-devel (2024-05-12 r86534) Fedora Linux 38 (Container Image)
  • 13 🐋 mkl R-devel (2024-05-12 r86534) Fedora Linux 38 (Container Image)
  • 14 🐋 nold R-devel (2024-05-12 r86534) Ubuntu 22.04.4 LTS
  • 16 🐋 ubuntu-clang R-devel (2024-05-12 r86534) Ubuntu 22.04.4 LTS
  • 17 🐋 ubuntu-gcc12 R-devel (2024-05-12 r86534) Ubuntu 22.04.4 LTS
  • 18 🐋 ubuntu-next R-4.4.0 (patched) (2024-05-12 r86534) Ubuntu 22.04.4 LTS
  • 19 🐋 ubuntu-release R-4.4.0 (2024-04-24) Ubuntu 22.04.4 LTS
  • Windows Server 2022, R-devel, 64 bit

R CMD check results

There were no ERRORs or WARNINGs.

There is 1 NOTE.

  1. One related to URLs:
Found the following (possibly) invalid URLs:
  URL: https://comtradeapi.un.org/files/v1/app/reference/Reporters.json
    From: man/country_codes.Rd
    Status: 404
    Message: Not Found
  URL: https://comtradeapi.un.org/files/v1/app/reference/partnerAreas.json
    From: man/country_codes.Rd
    Status: 404
    Message: Not Found

These URLs are online and working, I am not sure what causes this issue. I can not see any redirects that may cause this. This is the URL provided by the UN hence I cannot change it.

Thanks and looking forward to the review! Paul Bochtler