I requested an academic copy of the Revolution R Enterprise today. Since they only have Windows or Redhat version, I installed it on an old spare machine dual-booting Linux and Windows. I like what they did with the IDE (the debugger is nice) and some of their tailored made packages dealing with large data and parallel processing. It should be a good choice to introduce R to students. I guess I am not surprised to see the Windows version is made using MS Visual Studio, but having to install a load of VS-related stuff just to install this relatively simple software (with respect to the GUI, of course) make me feel a little, …unhappy.
I think this is a serious alternative (for the first time) to the “official” R on Windows platform. And Revolution is generous to let faculty and students to use their software for free. If I decide to teach my students R, this software is absolutely my first choice.
By the way, I was not able to get the package “MCMCglmm” running because the required package “tensorA” is not in their official repository, and the version I pulled off CRAN is no good.
============== UPDATES (SEPT. 11, 2010) ========================
The problem with MCMCglmm seems to be fixed for the W32 system: the package “tensorA” was added to the official repository so I can run my favorite package for mixed effect modeling now. I still cannot use it on Win64 system.