Michael,
this is really a CRAN question rather than a package question so you may have
more luck asking CRAN, but generally it is not a a good idea to have strong
dependencies (Depends/Imports) on Bioconductor packages. If you want to use
non-CRAN packages I would recommend weak depends like Suggests since your
package won't work otherwise. For example, Bioconductor does not provide any
binaries for the big-sur CRAN release of R, so your package cannot be installed
by users if it depends on any BioC packages with native code. In fact, the
check report you linked already warns you about your dependencies in general:
Version: 4.0.2
Check: package dependencies
Result: NOTE
Imports includes 22 non-default packages.
Importing from so many packages makes the package vulnerable to any of
them becoming unavailable. Move as many as possible to Suggests and
use conditionally.
FWIW the reason you got the notification later is that macOS checks are not
done immediately on submission otherwise the package would not have been
accepted in the first place. That said, if you have issues with dependencies on
the macOS build machines just contact me and I can install them if they are
available (I have now kick-started a script that attempts to automatically
detect and install missing dependencies if available so you may see it in the
next check run), but as discussed above you should still re-think your
dependencies.
Cheers,
Simon
> On Nov 16, 2021, at 3:33 PM, Michael Hellstern <[email protected]> wrote:
>
> Hi all,
>
> I maintain a package (netgsa) that has been on CRAN since september 2021
> and recently got an email about CRAN check failures. There are 2 errors and
> 1 warning. The failures are listed here:
> https://cran.r-project.org/web/checks/check_results_netgsa.html.
>
> The failures are a result of Bioconductor packages. Both errors are about
> package dependencies not being available (org.Hs.eg.db and/or ndexr) while
> the warning comes from the vignette building failing which is caused by an
> error in the graphite package. I checked and these packages were and still
> are available on Bioconductor. The errors occurred on macOS 10.13.6. While
> not exactly the same OS, I ran R CMD check --as-cran on my macOS 10.15.7
> and did not get any errors or warnings. It seems odd that I would get a
> package dependency availability error for only 2 flavors and that I would
> not be able to reproduce this on a similar OS.
>
> I searched online but have not found anything useful. I built the package
> under R 4.0.2 which is now a little dated so am wondering if it is related
> to this? Do I need to rebuild using the most recent version of R? Any
> advice on how to resolve this would be greatly appreciated!
>
> Best,
>
> Mike
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> [email protected] mailing list
> https://stat.ethz.ch/mailman/listinfo/r-package-devel
>
______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-package-devel
