[R] problem with building package on CRAN

[Romain Francois]

Hello,

(ccing Rcpp-devel too because this is relevant)

This comes up every now and again on packages that are completely 
unrelated to Rcpp. We don't know yet why or what to do to fix the issue. 
I believe (but this might not be the case) that this is due to packages 
that do use Rcpp and fail to follow our guidelines and documentation and 
emails about using LinkingTo to pull in header files.
See 
http://lists.r-forge.r-project.org/pipermail/rcpp-devel/2010-July/000898.html

I'm really sorry you are caught in the middle of this, there is probably 
nothing wrong with your package, it just happens sort of randomly. I 
believe the chances of this happening would be lower if package 
developers (of package using Rcpp) would be so kind and follow our 
guidelines, but we can only offer to write the guidelines, we cannot 
force them to read or apply them.

Romain

Le 26/07/10 22:43, William Revelle a écrit :
>
> Dear friends,
> I have just gotten a strange error message back from Uwe saying that the
> most recent version of psych failed to pass R CMD check for Windows.
>
> The error message was less than helpful, in that it seems to have failed
> when trying to include the Rcpp library, which I do not directly call.
> (see below)
>
>
> "* using log directory 'd:/Rcompile/CRANpkg/local/2.11/psych.Rcheck'
> * using R version 2.11.1 (2010-05-31)
> * using session charset: ISO8859-1
> * checking for file 'psych/DESCRIPTION' ... OK
> * this is package 'psych' version '1.0-90'
> * checking package name space information ... OK
> * checking package dependencies ... OK
> * checking if this is a source package ... OK
> * checking whether package 'psych' can be installed ... ERROR
> Installation failed.
> The installation logfile:
> -Id:/Rcompile/CRANpkg/lib/2.11/Rcpp/include
>
>
>
> I do have several suggested packages (polycor, GPArotation, MASS, graph,
> Rgraphviz, mvtnorm, Rcsdp), but none of these are actually required. My
> examples all ask if the suggested packages are available and then do not
> call them if they are not.
>
> Any suggestions on what to do would be appreciated.
>
> Thanks.
>
> Bill



-- 
Romain Francois
Professional R Enthusiast
+33(0) 6 28 91 30 30
http://romainfrancois.blog.free.fr
|- http://bit.ly/bc8jNi : Rcpp 0.8.4
|- http://bit.ly/dz0RlX : bibtex 0.2-1
`- http://bit.ly/a5CK2h : Les estivales 2010