[Rd] Strange issues with new R-based 'R CMD build' and 'R CMD check' scripts on Windows

[Hervé Pagès]

Hi,

Kind of related to the new R-based 'R CMD build' script being
sometimes confused on Windows (as reported here a couple of weeks
ago https://stat.ethz.ch/pipermail/r-devel/2010-June/057759.html),
I also started to observe that the new R-based 'R CMD check' script
gets also confused sometimes on Windows in a random way.

It seems that sometimes (maybe 1% of the times) running 'R CMD check'
is actually running another script, not the 'check' script, e.g.,
here
 
http://bioconductor.org/checkResults/2.7/bioc-LATEST/geneplotter/gewurz-checksrc.html
the output of the command looks like if 'R CMD build' was run instead
of 'R CMD check'.

Here
 
http://bioconductor.org/checkResults/2.7/bioc-LATEST/lumi/gewurz-checksrc.html
it looks like if 'R CMD Rd2txt' was run.

Same here
 
http://bioconductor.org/checkResults/2.7/bioc-LATEST/HTqPCR/liverpool-checksrc.html
and here
 
http://bioconductor.org/checkResults/2.7/bioc-LATEST/iterativeBMAsurv/liverpool-checksrc.html

Like with the new R-based 'R CMD build' script, the problem seems
to happen randomly for a couple of packages every day. It seems to
affect our 2 Windows build machines, 32-bit and 64-bit. Both are
using R-2.12.0 (2010-07-06 r52464).

Cheers,
H.


-- 
Hervé Pagès

Program in Computational Biology
Division of Public Health Sciences
Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N, M2-B876
P.O. Box 19024
Seattle, WA 98109-1024

E-mail: hpages at fhcrc.org
Phone:  (206) 667-5791
Fax:    (206) 667-1319