[Bioc-devel] Bioc_check on devel - Checking R Version dependency warning

[Ahmed Mohamed]

Thanks Lori for the clarification. If the purpose of the check is to
capture the R version upon submission to Bioconductor, then it makes sense.
I would suggest a bit of clarification in the docs though. Also, a warning
is not produced when "Depends: R (xxx)" is completely missing in
DESCRIPTION. If my understanding is correct, a warning should also be given
in that case.

For my package, I plan to keep every thing as is in a pre-release version
(0.98.0), and then pump R version upon submission to tracker (v0.99.0),
skipping CI tests for bioc-release. Hope this will be acceptable.

Cheers.
Ahmed.

On Wed, 27 Mar 2019 at 21:23, Shepherd, Lori <Lori.Shepherd using roswellpark.org>
wrote:

> Bioconductor has release cycles to ensure that packages are building
> cleaning and working with dependency at any given time.  The version would
> not have to be updated each release, it is an indication of the version of
> R used at the time of submission and to give a base line of what versions
> were minimally working.
>
>
> Bioconductor builds packages nightly on release and devel branches. we
> assume any development will be pushed to the devel branch to ensure if a
> package functionality changes and packages that depend on it may break that
> there is some indication of that to give package maintainers an environment
> to update and adapt, and that there is still the release stable environment
> of working dependencies.
>
>
> If your package is already in Bioconductor it is safe to ignore this
> warning when running the testing environment (or if you feel like disabling
> it you could use the BiocCheck option --no-check-R-ver)  However,  if
> your package is in the submission tracker we will generally insist on
> updating this R version dependency for the reasons stated.  For example,
> there have been many significant changes to R from 3.5 to 3.6 that effected
> a lot of packages and required some major changes, we would have no way to
> test or guarantee that a package can work on a previous version and be able
> to grab all the correct versions of packages for a past point in time
> (beside current release/devel).  If you really feel strongly about this you
> could still submit your package and argue your case to the reviewer - we
> sometimes make an exception but it is not the norm.
>
>
> Hope this helps with clarification
>
> Cheers,
>
>
> Lori Shepherd
>
> Bioconductor Core Team
>
> Roswell Park Cancer Institute
>
> Department of Biostatistics & Bioinformatics
>
> Elm & Carlton Streets
>
> Buffalo, New York 14263
> ------------------------------
> *From:* Bioc-devel <bioc-devel-bounces using r-project.org> on behalf of Ahmed
> Mohamed <mohamed using kuicr.kyoto-u.ac.jp>
> *Sent:* Wednesday, March 27, 2019 12:54:14 AM
> *To:* bioc-devel using r-project.org
> *Subject:* [Bioc-devel] Bioc_check on devel - Checking R Version
> dependency warning
>
> Hi all,
>
> I noticed that bioc_check on devel checks for R version dependency, and
> produces a warning to change R dependency to 3.6 (currently set as R >=
> 3.5.0). Just quickly looking at Bioconductor packages, it seems that at
> least half of the packages will have this warning, if tested. It also means
> that packages will need to change R dependency each release cycle (and
> different R versions for release and devel).
>
> To be honest, I am not fully convinced also on the rationale behind it.
> Quoting
>
> http://master.bioconductor.org/packages/devel/bioc/vignettes/BiocCheck/inst/doc/BiocCheck.html#version-checks
>
> "BiocCheck checks to make sure that the R version specified matches the
> version currently used by *Bioconductor*. This prevents the package from
> being used in earlier versions of R, which is not recommended and is a
> frequent cause of user confusion (WARNING)."
>
> Currently, I am using TravisCI to Bioc_check my package on both release and
> devel modes, so it is not also possible to change the R version as
> suggested in the warning. I wonder if there is a possible work around,
> other than looking the other way.
>
> Link to build here:
> https://travis-ci.org/ahmohamed/lipidr/jobs/511789232#L5673
>
> Cheers.
> Ahmed.
>
>         [[alternative HTML version deleted]]
>
> _______________________________________________
> Bioc-devel using r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/bioc-devel
>
> This email message may contain legally privileged and/or confidential
> information. If you are not the intended recipient(s), or the employee or
> agent responsible for the delivery of this message to the intended
> recipient(s), you are hereby notified that any disclosure, copying,
> distribution, or use of this email message is prohibited. If you have
> received this message in error, please notify the sender immediately by
> e-mail and delete this email message from your computer. Thank you.

	[[alternative HTML version deleted]]