Vignettes can fail to build for reasons unrelated to code. In that case it seems useful to the developer to know whether the the code is failing (indicating a likely problem in the package itself) or just the TeX in the vignette.
Also, I could be wrong about this, but I thought the "run the vignette code" test happened *before* vignette building. ~G On Sat, May 31, 2014 at 3:52 PM, Yihui Xie <x...@yihui.name> wrote: > Note the test has been done once in weave, since R CMD check will try > to rebuild vignettes. The problem is whether the related tools in R > should change their tangle utilities so we can **repeat** the test, > and it seems the answer is "no" in my eyes. > > Regards, > Yihui > -- > Yihui Xie <xieyi...@gmail.com> > Web: http://yihui.name > > > On Sat, May 31, 2014 at 4:54 PM, Gabriel Becker <gmbec...@ucdavis.edu> > wrote: > > > > > > > > On Fri, May 30, 2014 at 9:22 PM, Yihui Xie <x...@yihui.name> wrote: > >> > >> Hi Kevin, > >> > >> > >> I tend to adopt Henrik's idea, i.e., to provide vignette > >> engines that just ignore tangle. At the moment, it seems R CMD check > >> is comfortable with vignettes that do not have corresponding R > >> scripts, and I hope these R scripts will not become mandatory in the > >> future. > > > > > > I'm not sure this is the right approach. This would essentially make the > > test optional based on decisions by the package author. I'm not arguing > in > > favor if this particular test, but if package authors are able to turn a > > test off then the test loses quite a bit of it's value. > > > > I think that R CMD check has done a great deal for the R community by > > presenting a uniform, minimum "barrier to entry" for R packages. Allowing > > package developers to alter the tests it does (other than the obvious > case > > of their own unit tests) would remove that. > > > > That having been said, it seems to me that tangle-like utilities should > have > > the option of extracting inline code, and that during R CMD check that > > option should *always* be turned on. That would solve the problem in > > question while retaining the test would it not? > > > > ~G > -- Gabriel Becker Graduate Student Statistics Department University of California, Davis [[alternative HTML version deleted]] ______________________________________________ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
