[R-pkg-devel] CRAN upload and prechecking confusion

Johannes Graumann nonsense at graumannschaft.org
Thu Aug 24 13:43:39 CEST 2017


On Thu, 2017-08-24 at 02:59 -0500, Brian G. Peterson wrote:
> On 08/24/2017 02:28 AM, Johannes Graumann wrote:
> > Hello,
> > 
> > I'm trying to submit a new version (0.10.2) of my package
> > (multipanelfigure) to CRAN and have received the following email:
> > 
> > > Dear maintainer,
> > >   
> > > package multipanelfigure_0.10.2.tar.gz does not pass the incoming
> > > checks automatically, please see the pre-test at:
> > > <https://win-builder.r-project.org/incoming_pretest/170824_085724
> > > _mul
> > > tipanelfigure_0102/00check.log>
> > > Status: 2 NOTEs
> > >   
> > > Current CRAN status: ERROR: 1, OK: 10
> > > See: <https://CRAN.R-project.org/web/checks/check_results_multipa
> > > nelf
> > > igure.html>
> > 
> > This is confusing. What's the deal breaker?
> > The 2 Notes (which are timing issues related to drawing grid
> > graphics)
> > or
> > the Error on x86_64-apple-darwin15.6.0 which is based on the fact
> > that
> > gridGraphics is unavailable for that platform and even more
> > confungsingly seems to be based on the former version (0.9.0) of
> > the
> > package ...
> 
> The URL's you've referenced have both expired.
> 
> I would suggest re-running your package checks using R-devel on RHub
> and 
> Winbuilder first, and copying any output you receive.
> 
> Then it will be easier for others to help you.
Have a look hier:
https://builder.r-hub.io/status/multipanelfigure_0.10.2.tar.gz-7184a1dc
6654eaea5463f8cc9cac3f73
https://builder.r-hub.io/status/multipanelfigure_0.10.2.tar.gz-033531ed
74ed5a90ce66bac077293e3c
https://builder.r-hub.io/status/multipanelfigure_0.10.2.tar.gz-1c5377c9
369c4936e0c5e1a81c863275

Linux & windows work. OSX fails because gridGraphics (currently) does
not exist for the platform.

In the meantime CRAN (Uwe Ligges) got back telling me the problem is
long timing in an example ... this is based on grid speed and as the
package heavily uses that I feel uncomfortable moving the example to
'dontrun' ... thoughts.

Sincerely, Joh



More information about the R-package-devel mailing list