Ratfor, R, RUNOFF, RPG and Ruby
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.
R is for Ratfor, R, RUNOFF, RPG and Ruby
Ratfor is a structured form of Fortran from the days when structured programming was the in-thing and Fortran did not have much of it (lots got added in later revisions). I think its success came from allowing users to claim a degree of respectability that Fortran did not have, and which Fortran did not appear to gain when structure constructs were added to it (but then all successful languages are treated with suspicion in some circles).
The maintainers of R provide a valuable lesson on issues that are not important for a language to be widely used, such as the following (which I’m sure many of those involved with languages incorrectly think are important):
- Technical language details are not important (e.g., functional, imperative, declarative, object-oriented, etc); as far as I can tell the language has hardly changed over the years, yet users are not bothered. What is important is how easily the language can be used to solve its users’ problems. There are umpteen different ways a language can be designed around R’s very useful ability to operate on vectors as a single unit or to dynamically create data-frames, it does not make much difference how things are done as long as they work.
- Runtime efficiency is often not important; a look at the source of the R runtime system suggests that there are lots of savings to be had (maybe a factor of two). Users are usually a lot more willing to wait for a running program to complete than struggle with getting the program running in the first place; the R maintainers have concentrated on the tuning the usability of the ecosystem (intentionally or not, I don’t know). Also, R appears to be like Cobol in that the libraries are the best place to invest in performance optimization. I say ‘appears’ because I have noticed a growing number of R libraries completely written in R, rather than being a wrapper to code in C or Fortran; perhaps the efficiency of the runtime system is becoming an important issue.
Most programs don’t use a lot of cpu resources, this was true back when I was using 8-bit cpus running at 4MHz and is even more true today. I used to sell add-on tools to make code run faster and it was surprising how often developers had no idea how long their code took to run, they just felt it was not fast enough; I was happy to go along with these feelings (if the developers could recite timing data a sale was virtually guaranteed).
plot
is an unsung hero in R’s success, taking whatever is thrown at it and often producing something useful (if somewhat workman-like visually).
RUNOFF is the granddaddy of text processing systems such as *roff and LaTeX. RUNOFF will do what you tell it to do (groff is a modern descendant), while LaTeX will listen to what you have to say before deciding what to do. A child of RUNOFF shows that visual elegance can be achieved using simple means (maintainers of R’s plot
function take note). Businesses used to buy computers and expensive printers so they could use this language, or one of its immediate descendants.
RPG must be the most widely used proprietary programming language ever.
Is Ruby’s current success all down to the success of one web application framework written in it? In its early years C claim to fame was as the language used to write Unix (I know people who gave this as the reason for choosing to use C). We will have to wait and see if Ruby has a life outside of one framework.
R-bloggers.com offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.