[R-pkg-devel] Cascade effect of non-available packages?

[Dirk Eddelbuettel]

On 3 November 2024 at 11:02, Tiago Olivoto wrote:
| Today, I noticed that several stable packages, such as Rcpp and sf, were
| unavailable during the check process for this submission. This wasn’t an
| issue in the previously published version of pliman. Could this be a
| temporary problem with the package availability on CRAN?

I could be. (Auto-)build systems for 21k packages and 3 architectures can be
fragile, and we are in time of transition (R 4.4.2 just came about) so this
could have been spurious.  If in doubt, check the status of a given package
at its CRAN page. But I looked at the log links yu provided, and I saw no
build failure over 'missing Rcpp and/or sf' there.  Can you point us to an
error?

Moreover, reading
  https://win-builder.r-project.org/incoming_pretest/pliman_3.0.0_20241102_181842/specialChecks/noSuggests/summary.txt
I do not see your aforementioned 'check_ebi()' function at work. The
suggested package is not present, you are asked to not fail and just skip
but do not seem to do so. I think you need to fix that, and that the CRAN
messaging is fairly clear and standard here. I may of course have missed
something in which case 'my bad'.

Dirk

-- 
dirk.eddelbuettel.com | @eddelbuettel | edd using debian.org