[Rd] No output/no source tarball produced by 'R CMD build' on Windows (but ret code is 0)

Hervé Pagès hpages at fhcrc.org
Fri Jun 25 00:11:18 CEST 2010


Hi,

Just to note that, after updating to R-2.12.0 2010-06-22 r52364, the
problem is still here. Today's victim are altcdfenvs and CALIB:

http://bioconductor.org/checkResults/2.7/bioc-LATEST/altcdfenvs/liverpool-buildsrc.html
http://bioconductor.org/checkResults/2.7/bioc-LATEST/CALIB/liverpool-buildsrc.html

Cheers,
H.

On 06/21/2010 12:20 PM, Hervé Pagès wrote:
> Hi,
>
> After updating R-2.12 on the Bioconductor build system about 1 week ago
> (we update R every 4-6 weeks for our devel builds), we've started to see
> something never seen before on Windows: for some BioC packages,
> 'R CMD build' would produce no output at all (0 byte) and, of course,
> no source tarball. Even though, the code returned by the command is 0!
>
> See an example of this is here:
>
>
> http://bioconductor.org/checkResults/2.7/bioc-LATEST/MotIV/liverpool-buildsrc.html 
>
>   (the content of this page is updated every day)
>
> Note the:
>   RetCode: 0
>   PackageFile: None
> and the lack of output!
>
> This seems to happen randomly for a couple of packages every day. Today
> it's for MotIV only. Last time I checked, it was charm and genefilter.
> Before that, they were 2 other packages.
>
> Note that the only thing that changed on our Windows build machine
> liverpool (Windows Server 2003 R2 (32-bit)) is that R-2.12.0 was
> updated to 2010-06-11 r52254.
>
> Let me know if I can help in any way.
>
> Thanks,
> H.
>



More information about the R-devel mailing list