[Rd] R CMD build question

Martin Maechler maechler at stat.math.ethz.ch
Fri Sep 22 18:33:35 CEST 2006


>>>>> "JimMcD" == James W MacDonald <jmacdon at med.umich.edu>
>>>>>     on Fri, 22 Sep 2006 11:28:48 -0400 writes:

    JimMcD> On Solaris when my package is built, I get the following result:

I don't think this is bound to Solaris at all.

    JimMcD> creating vignettes ...Segmentation Fault - core dumped
    JimMcD> OK

    JimMcD> My question isn't why I get a segfault, but why does
    JimMcD> build return an OK after such an inauspicious event?
    JimMcD> Is build only supposed to error out if something
    JimMcD> more central to the package is off?

    JimMcD> R CMD check does issue a warning that the vignette
    JimMcD> is missing, so maybe that is the intended result.

I have been puzzled recently by similar behavior, and I think
that current behavior is not satisfactory.
IIRC, I even got
--------------------------      
creating vignettes ... OK
--------------------------      
But from the 3 vignettes, the 2nd one gave an error (in R code),
and consequently only the pdf of the 1st vignette had been built  --- all
completely silently until you "R CMD check" the built tar file
get the warning about missing pdf.

I only did not "complain" because people could have pointed out to
me that I could fix the problem myself .. :-\

Martin




More information about the R-devel mailing list