[R-pkg-devel] Additional CRAN Checks

[Uwe Ligges]

On 07.09.2021 16:39, Colin Gillespie wrote:
> ))> Exactly what errors do you think you found? It is not an error for
> a package to be compatible with a range of versions of R.
> 
> I suspect that the author didn't mean to specify 7 versions of R
> (https://cran.r-project.org/web/packages/NGSSEML/) under Depends.

I suspect so, too, but it works.


> When a package specifies R >= 2.01 did they mean R 2.0.1 or R 2.1.0
> (https://cran.r-project.org/web/packages/diagram/)

Yes, this is not obvious. But then it will be interpreted as R >= 2.1 
which is not wrong in any case (even if 2.0.1 was meant).


> Or specifying R >= 2.16 (which doesn't exist) -
> https://cran.r-project.org/web/packages/seqDesign/

Actually R-2.16.0 did exist as a development version that was later on 
renumbered to 3.0.0, so it is right.


> I realise that the above aren't "errors", but they do strongly
> indicate that the package author is making a mistake.

The only unclear statement is the 2.01, but compareVersion() does a 
sensible thing. So I do not worry too much.

Best,
Uwe Ligges


> Thanks
> 
> Colin
> 
> 
> 
>>
>> On September 6, 2021 4:02:08 AM PDT, Colin Gillespie <csgillespie using gmail.com> wrote:
>>> Dear All,
>>>
>>> Sorry if this is the wrong mailing address.
>>>
>>> I was doing a little investigation of R versions in packages and
>>> discovered a number of errors:
>>>
>>> https://cran.r-project.org/web/packages/bartCause/index.html  - R (≥ 3.1-0)
>>> https://cran.r-project.org/web/packages/activityGCMM/index.html  - R (3.00)
>>> https://cran.r-project.org/web/packages/deTestSet/index.html - R (≥ 2.01)
>>>
>>> and also
>>>
>>> https://cran.r-project.org/web/packages/NGSSEML/ - R (≥ 1.9.0), R (≥
>>> 3.5.0), R (≥ 3.5.0), R (≥ 3.5.0), R (≥ 3.5.0), R (≥ 3.5.0), R (≥
>>> 3.5.0)
>>>
>>> Overall I detected ~40 packages with issues:
>>>
>>> c("activityGCMM", "behavr", "bioassays", "bvpSolve", "celestial",
>>> "chest", "CRTSize", "csppData", "damr", "deTestSet", "diagram",
>>> "directPA", "DiscreteFDR", "ecolMod", "ecp", "epibasix", "europepmc",
>>> "fabisearch", "FDX", "ggetho", "hyper.fit", "MedSurvey", "NBPSeq",
>>> "NetIndices", "NFWdist", "OmegaG", "primer", "RATest", "refund",
>>> "RmarineHeatWaves", "rmcfs", "rootSolve", "RPPASPACE", "scopr",
>>> "scoringRules", "seqDesign", "shape", "sleepr", "SMPracticals",
>>> "text", "tggd", "zeitgebr")
>>>
>>> It seems like this would be a useful check to implement on the R CMD check.
>>>
>>> Thanks
>>>
>>> Colin
>>>
>>>
>>> Dr Colin Gillespie
>>> https://twitter.com/csgillespie
>>>
>>> ______________________________________________
>>> R-package-devel using r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-package-devel
>>
>> --
>> Sent from my phone. Please excuse my brevity.
> 
> ______________________________________________
> R-package-devel using r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-package-devel
>