[R] Problem with updating packages

[Duncan Murdoch]

On 2026-05-27 12:01 p.m., Karl Schilling wrote:
> I am running R under Windows 11. recently updated R to the 4.6patchhed
> version. I then copied my packages from my previous version (4.5) to the
> library of 4.6. Then I updated all packages.

You don't say exactly what you mean by "my packages".  If you copy *all* 
R 4.5 packages to the default R 4.6 library directory, you will probably 
damage R beyond repair.  You need to remember that packages with 
priority "base" contain code that implements R internals.  Trying to use 
R 4.5 internals in R 4.6 will cause a lot of problems.

If you only copy contributed packages (i.e. ones that you installed 
which get priority NA, or recommended packages) then what you did might 
work, but it's safest to just save the names of all contributed 
packages, and reinstall those.>
> Since then, I see the following behavior:
> 
> Each time I run "update.packages(ask='graphics',checkBuilt=TRUE)" I am
> asked TWICE
> 
> "--- Please select a CRAN mirror for use in this session ---".
> 
> And, more embarrassingly, all my Cran and Bioconductor are said to
> require an update. That also happens when I run "update.packages", say,
> one hour after my last update.

I think others have pointed out that every contributed package *does* 
need an update when going from R 4.5.* to R 4.6.*.  You can only get 
away with keeping installed copies of packages when you stay within the 
same minor version number, e.g. 4.6.1 to 4.6.2.

Duncan Murdoch