R is an object oriented programming for statistics and graphics. R is GNU S, i.e. an open source implementation of the language. R has been developed for many years and is now clearly a mature project and has eclipsed S, even some of the original developers of S contribute now to the development of R. R is now fast become a de facto standard for publishing statistical algorithms. A huge community of researchers contribute packages (components for the R library) in every imaginable "corner" of statistics. Core R itself is actively developed; there are scheduled new releases twice a year (with occasional intermediate releases).
As both R and S are based on the same language, for most basic statistical and graphical tasks there are no differences at all, although in more specialized areas they differ, e.g. S provides many additions for commercial customers e.g. for finance. There are some fundamental differences between R and S, especially the graphics engine, but for ordinary uses these differences do not matter, and usually if there are differences, equivalent functions using a different name are provided, e.g. in R lattice provides the same functionality as trellis in S. [Differences between S and R]