[R-pkg-devel] What is a "retired"package?

[Lionel Henry]

Hello,

We renamed "retired" to "superseded" some time ago to avoid any
confusion. Superseded functions and packages continue to be maintained
on CRAN for the foreseeable future and it is safe to depend on them.
See the "superseded" definition in https://tidyverse.org/lifecycle/.

plyr probably still mentions "retired" because it is rarely updated
(only when a CRAN failure requires attention).

Best,
Lionel



On 9/21/21, Lenth, Russell V <russell-lenth using uiowa.edu> wrote:
> I received a request that I remove the 'plyr' package from the Imports for
> my package, because plyr is retired. Indeed, the README file for plyr
> states:
>
>     > plyr is retired: this means only changes necessary to keep it on CRAN
> will be made.
>     > We recommend using dplyr (for data frames) or purrr (for lists)
> instead.
>
> This says "retired" but it also suggests that plyr will continue to be
> maintained. And that is a good thing because plyr has over 700 direct
> reverse-dependents, and almost 2000 if we include indirect reverse
> dependencies.
>
> So it seems to me that it isn't a problem at all to have my package import
> plyr. I use its 'aaply' and 'alply' functions, which are like 'apply' but
> work for any dimensional array. There are no obvious replacements in purrr
> or dplyr, and if there were and I used them instead, it would increase my
> indirect dependencies to several packages that are not actually needed.
>
> The user requesting that I drop plyr states that this is needed to satisfy
> regulatory needs, that it is problematic to qualify my package since it
> imports a retired package.
>
> So my question is: Is there a specific meaning in CRAN for "retired," or is
> that just loose language from the plyr developers? I did not find the term
> in "CRAN Repository Policy" or "Writing R Extensions." It appears that my
> user or their regulatory agency thinks it means that it could be deprecated
> in the near future. (If that is indeed what it means, there are 700+
> packages in trouble!) Otherwise, perhaps the appropriate request may be to
> the plyr maintainers to modify how they describe its status, so as to avoid
> confusion.
>
> Russ Lenth
> University of Iowa
>
> ______________________________________________
> R-package-devel using r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-package-devel
>