[R-pkg-devel] Installation took CPU time elapsed time for rust-based package

[Josiah Parry]

Thanks, Ivan! Yes, I would use the NOT_CRAN flag. But in the case of that
package, the CRAN_FLAGS are not set dynamically.
It would be a welcome PR to rextendr if anyone more comfortable with Make
would like to. The current
template Makevars is at
https://github.com/extendr/rextendr/blob/main/inst/templates/cran/Makevars.

David, it looks like you went about crafting your CRAN-compliance by
hand—hats off! If you want to be a little more
hands off in the future, you can check out the development version of
{rextendr} https://extendr.github.io/rextendr/dev/reference/cran.html
which provides Makevars and configure scripts that are CRAN compatible.

The general workflow for a new package would be something like:

usethis::create_package()
rextendr::use_extendr()
rextendr::use_cran_defaults()
rextendr::use_msrv("1.69") # or whatever MSRV you have
rextendr::vendor_pkgs()

If you, or anyone else for that matter, have additional questions, I
recommend joining the
extendr discord at https://discord.gg/Mt4Kauus.

For more "formal" questions a GitHub discussion or issue is also welcome
https://github.com/extendr/extendr/discussions


On Wed, Oct 16, 2024 at 11:50 AM Ivan Krylov via R-package-devel <
r-package-devel using r-project.org> wrote:

> В Wed, 16 Oct 2024 17:38:27 +0000
> D Z <david_j_zimmermann using hotmail.com> пишет:
>
> > if [ -n "$_R_CHECK_LIMIT_CORES" ]; then \
> >  export BUILDFLAGS="-j 2 --offline"; \
>
> This environment variable is spelled with an extra underscore:
>
> https://cran.r-project.org/doc/manuals/r-devel/R-ints.html#index-_005fR_005fCHECK_005fLIMIT_005fCORES_005f
>
> At least one of Josiah's packages instead relies on the user setting
> the NOT_CRAN environment variable:
>
> https://github.com/extendr/b64/blob/67ff753f71676870e3b04657909050ae10ec2145/src/Makevars#L10-L28
>
> > Now I am wondering if there is anything that I can do about this NOTE
> > or if I should ignore it.
> > I did not find any reference to this in the "Using Rust in CRAN
> > Packages"
>
> "Not taking more than two parallel processes or threads during the CRAN
> check" is a more general requirement for all packages, regardless of
> whether they use Rust. A package that uses cmake -GNinja without
> limiting the number of compiler processes or gives the -j$(nproc)
> argument to a Make subprocess would have received this NOTE as well.
>
> --
> Best regards,
> Ivan
>
> ______________________________________________
> R-package-devel using r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-package-devel
>

	[[alternative HTML version deleted]]