2  General Feelings

My overall feelings about R are tough to quantify. As I mentioned near the start, its ultimate problem is the sum of its little problems. However, if I must speak generally, then I think that the problem with R is that it’s always some mix of the following:

1. A statistics language with countless useful libraries and an excellent collection of mathematical tools.
2. A Scheme-inspired language that tries to be functional while maintaining a C-like syntax.
3. Decades of haphazard patches for S.
4. A collection of semantic semtex that is powerful in the hands of a master and crippling in the hands of a novice.

When it’s anything but #3, R is great. Statisticians and mathematicians love it for #1 and programmers love it for #2 and #4. If it weren’t for #3, R would be an amazing – albeit, domain-specific – language, but #3 is such a big factor that it makes the language unpredictable, inconsistent, and infuriating. Mixed with #4, it makes being an R novice hellish. It gives me little doubt that R is not the ideal tool for many of the jobs that it wants to do, but #1 and #2 leave me with equally little doubt that R can be a very good tool.