On 2019/01/02 17:08:58, Wes McKinney <w...@gmail.com> wrote: > hi folks,> > > With 0.12 around the corner and significant progress on the R bindings> > project (sufficient for Spark integration [1]), I am wondering how> > everyday R users are going to be able to install the software> > respectively on Linux, macOS, and Windows. Thoughts about the strategy> > for this?>
The R packaging is a bit different than python. For Windows and macOS, we can statically link external libs into the R package, to ship a standalone binary R package without any runtime dependencies. On Linux, R requires the system package manager (apt/yum) to provide external libs. The R package manager doesn't work well with libs from Conda. The easiest way to build R packages on macOS is using binaries from Homebrew. Because arrow is already in homebrew this should be straightforward: https://github.com/Homebrew/homebrew-core/blob/master/Formula/apache-arrow.rb For Windows, we need to build libarrow using the R mingw-w64 toolchain (called Rtools). We are currently working to automate this process, but for now I manually build and maintain these binaries here: https://github.com/rwinlib/arrow For both Windows and MacOS holds that the process creating the statically linked R binary package gets more complex for every extra system dependency. Currently both the homebrew and rwinlib binaries only depend on Boost and disable all extra arrow features. So we can make the R package to work on Windows and Mac, however, for the R package to be eligible for publication on CRAN, the required Linux libs (i.e. libarrow-dev, arrow-devel) need to be available from the official Debian and/or Fedora repository. The CRAN maintainers want to build and test the R package at least on Debian using only libs from official repositories, they won't build or install other software on the build servers. Hence a prerequisite for getting the R package on CRAN is getting libarrow-dev and arrow-devel into the official Debian/Fedora repo's. In the mean time we could setup a custom R repository providing the R package in source and binary form on e.g. Bintray. That way users can at least do e.g: install.packages("arrow", repos = "https://dl.bintray.com/apache/arrow/r";) which is much easier than installing manually. However as long as the package is not on CRAN, other R packages cannot formally depend on it.
