[R-pkg-devel] Handling Not-Always-Needed Dependencies?
It is really simple: Having _both_ Suggests: foo _and_ an unconditonal call to foo::bar() in the code. Whereas Josh and I argue that it needs to be conditional on requireNames() coming back TRUE.
I am feeling particularly dense today. What about "Having _both_ Suggests: foo _and_ an unconditonal call to foo::bar() in the code." did you fail to convince CRAN about? That it is a good thing, or a bad thing?
The key question is: should R CMD check work if you the suggested packages are installed? This generally affects tests, examples, and vignettes. If the answer is yes, R CMD check should work, then it would be helpful for CRAN to enforce this, so that those of us who regularly run revdep checks on large numbers of packages don't have to install suggested packages. Personally, I just install all the suggested dependencies of reverse dependencies - it's not a huge burden for me since I can install binaries from CRAN. Hadley