Archive for the ‘programming’ Category
Programming Language Popularity – Part Eleven
I made a number of Google searches of the forms below and summed the results:
"implemented in " "written in " "developed in " "programmed in "
See Part Ten for prior results.
I’ve divided the table into sections based on percentage increases of more than 50% from one language to the next.
|------+-------------+------------+-------+-------| | Rank | Language | # Search | Prev. | Rank | | | | Results | Rank | Delta | |------+-------------+------------+-------+-------| | 1 | C | 22,403,000 | 1 | | |------+-------------+------------+-------+-------| | 2 | R | 2,089,000 | 2 | | | 3 | Java | 1,684,000 | 4 | 1 | | 4 | C++ | 1,404,500 | 3 | -1 | | 5 | Python | 1,154,100 | 5 | | | 6 | JavaScript | 850,500 | 7 | 1 | | 7 | C# | 829,600 | 6 | -1 | | 8 | Scheme | 744,600 | 11 | 3 | | 9 | Go | 645,700 | 12 | 3 | | 10 | FORTRAN | 618,700 | 8 | -2 | | 11 | Perl | 480,000 | 9 | -2 | | 12 | Ruby | 461,660 | 10 | -2 | |------+-------------+------------+-------+-------| | 13 | Lisp | 228,370 | 15 | 2 | | 14 | Rust | 216,440 | 21 | 7 | | 15 | Haskell | 210,600 | 14 | -1 | | 16 | Julia | 204,370 | 26 | 10 | | 17 | Scala | 188,500 | 17 | | | 18 | COBOL | 181,100 | 19 | 1 | | 19 | Common Lisp | 177,700 | 18 | -1 | | 20 | Prolog | 166,100 | 16 | -4 | | 21 | Smalltalk | 130,000 | 22 | 1 | | 22 | Forth | 114,100 | 24 | 2 | | 23 | Lua | 109,180 | 13 | -10 | | 24 | Erlang | 78,050 | 20 | -4 | | 25 | Clojure | 62,900 | 25 | | | 26 | OCaml | 60,690 | 23 | -3 | |------+-------------+------------+-------+-------| | 27 | Pony | 41,129 | 27 | | | 28 | Racket | 30,160 | 28 | | | 29 | Elixir | 24,466 | 29 | | | 30 | Kotlin | 19,077 | N/A | | |------+-------------+------------+-------+-------|
Programming Language Popularity – Part Ten
I made a number of Google searches of the forms below and summed the results:
"implemented in <language>" "written in <language>" "developed in <language>" "programmed in <language>"
See Part Nine for prior results 4 months ago.
I’ve divided the table into sections based on percentage increases of more than 50% from one language to the next. To compute the rank delta, I excluded the newly added languages.
|------+-------------+------------+-------+-------| | Rank | Language | # Search | Prev. | Rank | | | | Results | Rank | Delta | |------+-------------+------------+-------+-------| | 1 | C | 30,370,000 | 1 | | |------+-------------+------------+-------+-------| | 2 | R | 4,101,600 | 2 | | |------+-------------+------------+-------+-------| | 3 | C++ | 1,406,100 | 4 | 1 | | 4 | Java | 1,394,000 | 3 | -1 | |------+-------------+------------+-------+-------| | 5 | Python | 868,000 | 6 | 1 | | 6 | C# | 844,800 | 5 | -1 | | 7 | Javascript | 632,600 | 7 | | | 8 | FORTRAN | 548,500 | 8 | | | 9 | Perl | 489,600 | 9 | | | 10 | Ruby | 399,200 | 10 | | | 11 | Scheme | 362,500 | 11 | | | 12 | Go | 333,800 | 12 | | |------+-------------+------------+-------+-------| | 13 | Lua | 170,940 | 17 | 4 | | 14 | Haskell | 151,190 | 15 | 1 | | 15 | Lisp | 141,880 | 13 | -2 | | 16 | Prolog | 139,890 | 16 | | | 17 | Scala | 124,450 | 14 | -3 | | 18 | Common Lisp | 113,510 | 20 | 2 | | 19 | COBOL | 113,000 | 18 | -1 | | 20 | Erlang | 80,060 | 19 | -1 | | 21 | Rust | 79,650 | 21 | | | 22 | Smalltalk | 73,120 | 23 | 1 | | 23 | OCaml | 59,970 | 24 | 1 | | 24 | Forth | 58,300 | 26 | 2 | | 25 | Clojure | 51,690 | 22 | -3 | | 26 | Julia | 45,210 | 25 | -1 | |------+-------------+------------+-------+-------| | 27 | Pony | 16,927 | N/A | N/A | | 28 | Racket | 16,150 | 28 | | | 29 | Elixir | 12,723 | 29 | | |------+-------------+------------+-------+-------|
The raw data is here:
Path=/search?q=%22written%20in%20c%22 Num=23200000 Path=/search?q=%22programmed%20in%20c%22 Num=180000 Path=/search?q=%22developed%20in%20c%22 Num=3320000 Path=/search?q=%22implemented%20in%20c%22 Num=3670000 (c 30370000) Path=/search?q=%22written%20in%20c%23%22 Num=396000 Path=/search?q=%22programmed%20in%20c%23%22 Num=40800 Path=/search?q=%22developed%20in%20c%23%22 Num=215000 Path=/search?q=%22implemented%20in%20c%23%22 Num=193000 (c# 844800) Path=/search?q=%22written%20in%20c%2B%2B%22 Num=477000 Path=/search?q=%22programmed%20in%20c%2B%2B%22 Num=80100 Path=/search?q=%22developed%20in%20c%2B%2B%22 Num=270000 Path=/search?q=%22implemented%20in%20c%2B%2B%22 Num=579000 (c++ 1406100) Path=/search?q=%22written%20in%20clojure%22 Num=26300 Path=/search?q=%22programmed%20in%20clojure%22 Num=4730 Path=/search?q=%22developed%20in%20clojure%22 Num=11800 Path=/search?q=%22implemented%20in%20clojure%22 Num=8860 (clojure 51690) Path=/search?q=%22written%20in%20cobol%22 Num=57600 Path=/search?q=%22programmed%20in%20cobol%22 Num=20900 Path=/search?q=%22developed%20in%20cobol%22 Num=18200 Path=/search?q=%22implemented%20in%20cobol%22 Num=16300 (cobol 113000) Path=/search?q=%22written%20in%20coffeescript%22 Num=26900 Path=/search?q=%22programmed%20in%20coffeescript%22 Num=1250 Path=/search?q=%22developed%20in%20coffeescript%22 Num=4330 Path=/search?q=%22implemented%20in%20coffeescript%22 Num=2510 (coffeescript 34990) Path=/search?q=%22written%20in%20common%20lisp%22 Num=49600 Path=/search?q=%22programmed%20in%20common%20lisp%22 Num=8110 Path=/search?q=%22developed%20in%20common%20lisp%22 Num=31500 Path=/search?q=%22implemented%20in%20common%20lisp%22 Num=24300 (common lisp 113510) Path=/search?q=%22written%20in%20elixir%22 Num=6000 Path=/search?q=%22programmed%20in%20elixir%22 Num=3 Path=/search?q=%22developed%20in%20elixir%22 Num=2480 Path=/search?q=%22implemented%20in%20elixir%22 Num=4240 (elixir 12723) Path=/search?q=%22written%20in%20erlang%22 Num=49300 Path=/search?q=%22programmed%20in%20erlang%22 Num=7460 Path=/search?q=%22developed%20in%20erlang%22 Num=6400 Path=/search?q=%22implemented%20in%20erlang%22 Num=16900 (erlang 80060) Path=/search?q=%22written%20in%20forth%22 Num=21800 Path=/search?q=%22programmed%20in%20forth%22 Num=2650 Path=/search?q=%22developed%20in%20forth%22 Num=25600 Path=/search?q=%22implemented%20in%20forth%22 Num=8250 (forth 58300) Path=/search?q=%22written%20in%20fortran%22 Num=289000 Path=/search?q=%22programmed%20in%20fortran%22 Num=29500 Path=/search?q=%22developed%20in%20fortran%22 Num=107000 Path=/search?q=%22implemented%20in%20fortran%22 Num=123000 (fortran 548500) Path=/search?q=%22written%20in%20go%22 Num=196000 Path=/search?q=%22programmed%20in%20go%22 Num=14900 Path=/search?q=%22developed%20in%20go%22 Num=67700 Path=/search?q=%22implemented%20in%20go%22 Num=55200 (go 333800) Path=/search?q=%22written%20in%20haskell%22 Num=85800 Path=/search?q=%22programmed%20in%20haskell%22 Num=4690 Path=/search?q=%22developed%20in%20haskell%22 Num=19500 Path=/search?q=%22implemented%20in%20haskell%22 Num=41200 (haskell 151190) Path=/search?q=%22written%20in%20java%22 Num=535000 Path=/search?q=%22programmed%20in%20java%22 Num=112000 Path=/search?q=%22developed%20in%20java%22 Num=352000 Path=/search?q=%22implemented%20in%20java%22 Num=395000 (java 1394000) Path=/search?q=%22written%20in%20javascript%22 Num=370000 Path=/search?q=%22programmed%20in%20javascript%22 Num=19900 Path=/search?q=%22developed%20in%20javascript%22 Num=75700 Path=/search?q=%22implemented%20in%20javascript%22 Num=167000 (javascript 632600) Path=/search?q=%22written%20in%20julia%22 Num=23000 Path=/search?q=%22programmed%20in%20julia%22 Num=1090 Path=/search?q=%22developed%20in%20julia%22 Num=15900 Path=/search?q=%22implemented%20in%20julia%22 Num=5220 (julia 45210) Path=/search?q=%22written%20in%20lisp%22 Num=71800 Path=/search?q=%22programmed%20in%20lisp%22 Num=7980 Path=/search?q=%22developed%20in%20lisp%22 Num=20600 Path=/search?q=%22implemented%20in%20lisp%22 Num=41500 (lisp 141880) Path=/search?q=%22written%20in%20lua%22 Num=54800 Path=/search?q=%22programmed%20in%20lua%22 Num=3720 Path=/search?q=%22developed%20in%20lua%22 Num=5420 Path=/search?q=%22implemented%20in%20lua%22 Num=107000 (lua 170940) Path=/search?q=%22written%20in%20ml%22 Num=77200 Path=/search?q=%22programmed%20in%20ml%22 Num=10700 Path=/search?q=%22developed%20in%20ml%22 Num=154000 Path=/search?q=%22implemented%20in%20ml%22 Num=61000 (ml 302900) Path=/search?q=%22written%20in%20ocaml%22 Num=28300 Path=/search?q=%22programmed%20in%20ocaml%22 Num=5070 Path=/search?q=%22developed%20in%20ocaml%22 Num=12800 Path=/search?q=%22implemented%20in%20ocaml%22 Num=13800 (ocaml 59970) Path=/search?q=%22written%20in%20perl%22 Num=342000 Path=/search?q=%22programmed%20in%20perl%22 Num=13400 Path=/search?q=%22developed%20in%20perl%22 Num=42100 Path=/search?q=%22implemented%20in%20perl%22 Num=92100 (perl 489600) Path=/search?q=%22written%20in%20php%22 Num=461000 Path=/search?q=%22programmed%20in%20php%22 Num=36000 Path=/search?q=%22developed%20in%20php%22 Num=202000 Path=/search?q=%22implemented%20in%20php%22 Num=143000 (php 842000) Path=/search?q=%22written%20in%20pony%22 Num=15200 Path=/search?q=%22programmed%20in%20pony%22 Num=2 Path=/search?q=%22developed%20in%20pony%22 Num=5 Path=/search?q=%22implemented%20in%20pony%22 Num=1720 (pony 16927) Path=/search?q=%22written%20in%20prolog%22 Num=57200 Path=/search?q=%22programmed%20in%20prolog%22 Num=5090 Path=/search?q=%22developed%20in%20prolog%22 Num=27400 Path=/search?q=%22implemented%20in%20prolog%22 Num=50200 (prolog 139890) Path=/search?q=%22written%20in%20python%22 Num=449000 Path=/search?q=%22programmed%20in%20python%22 Num=36000 Path=/search?q=%22developed%20in%20python%22 Num=131000 Path=/search?q=%22implemented%20in%20python%22 Num=252000 (python 868000) Path=/search?q=%22written%20in%20r%22 Num=1740000 Path=/search?q=%22programmed%20in%20r%22 Num=43600 Path=/search?q=%22developed%20in%20r%22 Num=1740000 Path=/search?q=%22implemented%20in%20r%22 Num=578000 (r 4101600) Path=/search?q=%22written%20in%20racket%22 Num=6210 Path=/search?q=%22programmed%20in%20racket%22 Num=1520 Path=/search?q=%22developed%20in%20racket%22 Num=5090 Path=/search?q=%22implemented%20in%20racket%22 Num=3330 (racket 16150) Path=/search?q=%22written%20in%20ruby%22 Num=248000 Path=/search?q=%22programmed%20in%20ruby%22 Num=49800 Path=/search?q=%22developed%20in%20ruby%22 Num=37100 Path=/search?q=%22implemented%20in%20ruby%22 Num=64300 (ruby 399200) Path=/search?q=%22written%20in%20rust%22 Num=39200 Path=/search?q=%22programmed%20in%20rust%22 Num=2120 Path=/search?q=%22developed%20in%20rust%22 Num=28800 Path=/search?q=%22implemented%20in%20rust%22 Num=9530 (rust 79650) Path=/search?q=%22written%20in%20scala%22 Num=77300 Path=/search?q=%22programmed%20in%20scala%22 Num=2550 Path=/search?q=%22developed%20in%20scala%22 Num=15500 Path=/search?q=%22implemented%20in%20scala%22 Num=29100 (scala 124450) Path=/search?q=%22written%20in%20scheme%22 Num=58000 Path=/search?q=%22programmed%20in%20scheme%22 Num=11200 Path=/search?q=%22developed%20in%20scheme%22 Num=213000 Path=/search?q=%22implemented%20in%20scheme%22 Num=80300 (scheme 362500) Path=/search?q=%22written%20in%20smalltalk%22 Num=27200 Path=/search?q=%22programmed%20in%20smalltalk%22 Num=9920 Path=/search?q=%22developed%20in%20smalltalk%22 Num=13200 Path=/search?q=%22implemented%20in%20smalltalk%22 Num=22800 (smalltalk 73120) Path=/search?q=%22written%20in%20sml%22 Num=7430 Path=/search?q=%22programmed%20in%20sml%22 Num=2200 Path=/search?q=%22developed%20in%20sml%22 Num=6370 Path=/search?q=%22implemented%20in%20sml%22 Num=15300 (sml 31300) Path=/search?q=%22written%20in%20standard%20ml%22 Num=9820 Path=/search?q=%22programmed%20in%20standard%20ml%22 Num=2060 Path=/search?q=%22developed%20in%20standard%20ml%22 Num=9 Path=/search?q=%22implemented%20in%20standard%20ml%22 Num=15700 (standard ml 27589)
Upgrading to Emacs 24.5
I previously wrote a post about switching from Carbon Emacs to Gnu Emacs. Upgrading is pretty simple now, so I thought I’d record the procedure for installing Emacs from source on OSX for future reference:
- Go here and click on a mirror link
- Download the appropriate file, e.g. emacs-24.5.tar.gz
- tar xzf emacs-24.5.tar.gz
- cd emacs-24.5
- ./configure –with-ns
- make install
- drag nextstep/Emacs.app to the Applications directory
Emacs is still awesome!
How to Write a Spelling Corrector in Racket
In September, 2008, I translated Peter Norvig’s spelling corrector into Ruby. My current favorite language is Racket, so I thought it would be a good exercise to port it to Racket. After some helpful tips by Vincent St-Amour and Sam Tobin-Hochstadt in the #racket IRC channel, I came up with the following. I’ll show it two different ways, the first minimizes the line count (without sacrificing too much stylistically) to 27 lines, and the second is closer to how I’d normally format it:
#lang racket
(define (words text) (regexp-match* #rx"[a-z]+" (string-downcase text)))
(define (train features)
(define model (make-hash))
(for ([f features]) (hash-update! model f add1 1)) model)
(define nwords (train (words (file->string "big.txt"))))
(define (edits1 word)
(let* ([alphabet "abcdefghijklmnopqrstuvwxyz"]
[splits (for/list ([i (in-range (+ (string-length word) 1))])
(cons (substring word 0 i) (substring word i)))]
[deletes (for/set ([(a b) (in-dict splits)] #:when (> (string-length b) 0))
(string-append a (substring b 1)))]
[transposes (for/set ([(a b) (in-dict splits)] #:when (> (string-length b) 1))
(string-append a (substring b 1 2) (substring b 0 1) (substring b 2)))]
[replaces (for/set ([(a b) (in-dict splits)] #:when (> (string-length b) 0) [c alphabet])
(string-append a (string c) (substring b 1)))]
[inserts (for*/set ([(a b) (in-dict splits)] [c alphabet])
(string-append a (string c) b))])
(set-union deletes transposes replaces inserts)))
(define (known-edits2 word)
(for*/set ([e1 (edits1 word)] [e2 (edits1 e1)] #:when (hash-has-key? nwords e2)) e2))
(define (known words) (for/set ([w words] #:when (hash-has-key? nwords w)) w))
(define (nes set) (if (set-empty? set) #f set))
(define (correct word)
(let ([candidates (or (nes (known (list word))) (nes (known (edits1 word)))
(nes (known-edits2 word)) (set word))])
(argmax (λ (x) (hash-ref nwords x 1)) (set->list candidates))))
And here is a more aesthetically pleasing format:
#lang racket
(define (words text)
(regexp-match* #rx"[a-z]+" (string-downcase text)))
(define (train features)
(define model (make-hash))
(for ([f features])
(hash-update! model f add1 1))
model)
(define nwords
(train (words (file->string "big.txt"))))
(define (edits1 word)
(let* ([alphabet "abcdefghijklmnopqrstuvwxyz"]
[splits (for/list ([i (in-range (+ (string-length word) 1))])
(cons (substring word 0 i) (substring word i)))]
[deletes (for/set ([(a b) (in-dict splits)]
#:when (> (string-length b) 0))
(string-append a (substring b 1)))]
[transposes (for/set ([(a b) (in-dict splits)]
#:when (> (string-length b) 1))
(string-append a
(substring b 1 2)
(substring b 0 1)
(substring b 2)))]
[replaces (for/set ([(a b) (in-dict splits)]
#:when (> (string-length b) 0)
[c alphabet])
(string-append a (string c) (substring b 1)))]
[inserts (for*/set ([(a b) (in-dict splits)]
[c alphabet])
(string-append a (string c) b))])
(set-union deletes transposes replaces inserts)))
(define (known-edits2 word)
(for*/set ([e1 (edits1 word)]
[e2 (edits1 e1)]
#:when (hash-has-key? nwords e2))
e2))
(define (known words)
(for/set ([w words] #:when (hash-has-key? nwords w))
w))
(define (nes set)
(if (set-empty? set)
#f
set))
(define (correct word)
(let ([candidates (or (nes (known (list word)))
(nes (known (edits1 word)))
(nes (known-edits2 word))
(set word))])
(argmax (λ (x) (hash-ref nwords x 1)) (set->list candidates))))
DSL Embedding in Racket
Here’s a great, two part, video by Matthew Flatt about embedding DSLs in Racket. Being able to hack the language is one of Racket’s/Lisp’s killer features:
Programming Language Popularity – Part Nine
I made a number of Google searches of the forms below and summed the results:
"implemented in <language>" "written in <language>" "developed in <language>" "programmed in <language>"
See Part Eight for prior results 17 months ago. I finally got around to writing a program (in Racket) to automate the collection of the search results, so it was much easier!
I’ve divided the table into sections based on percentage increases of more than 50% from one language to the next. To compute the rank delta, I excluded the newly added languages.
|------+-----------------+------------+-------+-------| | Rank | Language | # Search | Prev. | Rank | | | | Results | Rank | Delta | |------+-----------------+------------+-------+-------| | 1 | C | 54,696,000 | 2 | 1 | |------+-----------------+------------+-------+-------| | 2 | R | 22,784,600 | 12 | 10 | |------+-----------------+------------+-------+-------| | 3 | Java | 1,490,000 | 6 | 3 | | 4 | C++ | 1,256,600 | 3 | -1 | | 5 | PHP | 1,082,000 | 1 | -4 | | 6 | C# | 1,065,900 | 5 | -1 | | 7 | Python | 1,016,500 | 4 | -3 | | 8 | Lisp Family (1) | 951,300 | 11 | 3 | | 9 | JavaScript | 764,100 | 8 | -1 | | 10 | FORTRAN | 565,000 | 9 | -1 | | 11 | Perl | 541,700 | 7 | -4 | | 12 | Ruby | 509,900 | 10 | -2 | | 13 | ML Family (2) | 445,716 | 13 | | | 14 | Scheme | 381,200 | 19 | 5 | | 15 | Go | 378,200 | 16 | 1 | | 16 | (S)ML (3) | 376,116 | 21 | 5 | | 17 | Lisp | 357,900 | 15 | -2 | |------+-----------------+------------+-------+-------| | 18 | Scala | 236,770 | 23 | 5 | | 19 | Haskell | 220,580 | 17 | -2 | | 20 | Prolog | 152,450 | 20 | | | 21 | Lua | 133,360 | 22 | 1 | | 22 | COBOL | 113,700 | 14 | -8 | | 23 | Erlang | 101,840 | 18 | -5 | | 24 | Common Lisp | 100,580 | 25 | 1 | | 25 | Rust | 90,860 | N/A | N/A | | 26 | Clojure | 85,930 | 27 | 2 | | 27 | Smalltalk | 75,000 | 24 | -2 | | 28 | OCaml | 69,600 | 26 | -1 | | 29 | Coffeescript | 60,420 | 29 | 1 | | 30 | Julia | 51,315 | N/A | N/A | | 31 | Forth | 50,920 | 28 | -1 | |------+-----------------+------------+-------+-------| | 32 | Racket | 25,660 | 30 | | |------+-----------------+------------+-------+-------| | 33 | Elixir | 8,112 | N/A | N/A | |------+-----------------+------------+-------+-------|
Update Oct. 6: I added Bing search results, and they’re significantly different. For a while Elixir was #2 because one of the four phrases resulted in no results, so Bing automatically switched to an unquoted search which skewed the results. I caught that and looked over the raw data and didn’t see anything similar for other languages. It’s interesting that C# came out #1 on Bing 🙂
|------+-----------------+----------| | Rank | Language | # Search | | | | Results | |------+-----------------+----------| | 1 | C# | 32100000 | | 2 | C++ | 31900000 | |------+-----------------+----------| | 3 | C | 4845000 | | 4 | Java | 3319000 | |------+-----------------+----------| | 5 | R | 1219310 | | 6 | Python | 1123700 | | 7 | Php | 1093800 | | 8 | Javascript | 945200 | |------+-----------------+----------| | 9 | Lisp Family (1) | 525530 | | 10 | Perl | 460360 | | 11 | Common Lisp | 398039 | | 12 | Ruby | 301740 | | 13 | Go | 270593 | | 14 | FORTRAN | 245500 | | 15 | Lua | 214200 | | 16 | Scala | 188570 | | 17 | Haskell | 188520 | | 18 | Lisp | 160930 | | 19 | COBOL | 140910 | | 20 | Scheme | 98012 | | 21 | ML Family (2) | 97425 | | 22 | Prolog | 75740 | | 23 | Erlang | 63270 | | 24 | (S)ML (3) | 56375 | | 25 | OCaml | 41050 | | 26 | Smalltalk | 27960 | | 27 | Julia | 25319 | | 28 | Clojure | 23091 | | 29 | Coffeescript | 20322 | | 30 | Forth | 17529 | | 31 | Rust | 17086 | |------+-----------------+----------| | 32 | Racket | 6388 | |------+-----------------+----------| | 33 | Elixir | 2542 | |------+-----------------+----------|
Here is the delta difference between Google and Bing:
|--------+------+-------+-----------------| | Google | Bing | Delta | Language | | Rank | Rank | G - B | | |--------+------+-------+-----------------| | 1 | 3 | -2 | C | | 2 | 5 | -3 | R | | 3 | 4 | -1 | Java | | 4 | 2 | 2 | C++ | | 5 | 7 | -2 | Php | | 6 | 1 | 5 | C# | | 7 | 6 | 1 | Python | | 8 | 9 | -1 | Lisp Family (1) | | 9 | 8 | 1 | Javascript | | 10 | 14 | -4 | FORTRAN | | 11 | 10 | 1 | Perl | | 12 | 12 | | Ruby | | 13 | 21 | -8 | ML Family (2) | | 14 | 20 | -6 | Scheme | | 15 | 13 | 2 | Go | | 16 | 24 | -8 | (S)ML (3) | | 17 | 18 | -1 | Lisp | | 18 | 16 | 2 | Scala | | 19 | 17 | 2 | Haskell | | 20 | 22 | -2 | Prolog | | 21 | 15 | 6 | Lua | | 22 | 19 | 3 | COBOL | | 23 | 23 | | Erlang | | 24 | 11 | 13 | Common Lisp | | 25 | 31 | -6 | Rust | | 26 | 28 | -2 | Clojure | | 27 | 26 | 1 | Smalltalk | | 28 | 25 | 3 | OCaml | | 29 | 29 | | Coffeescript | | 30 | 27 | 3 | Julia | | 31 | 30 | 1 | Forth | | 32 | 32 | | Racket | | 33 | 33 | | Elixir | |--------+------+-------+-----------------|
(1) combines Lisp, Common Lisp, Scheme, Clojure & Racket
(2) combines (S)ML & OCaml
(3) summed separate searches for standard ml, sml & ml
The Three R’s
I’ve read that ten years is a common “lifetime” for a technology. Whether that’s true, or not, it seems to be the case with me. After beginning with BASIC & S/360 Assembler, I programmed in C/C++ from ~ 1985 until 1995, then in Java until 2006, and since then I’ve been programming primarily in Ruby/Rails.
For a number of years, I’ve felt the need to consider what my next primary language should be, and after experimenting with various lisps, Haskell and OCaml, I found that Racket is the best fit for me. I’m still very much in learning mode now, but I hope to be using Racket professionally in 2016.
Despite the fact that Racket is significantly faster than Ruby, I do occasionally have compute-intensive tasks that require more raw speed, so for a high performance language to maximize all the cores at my disposal, I’ve recently settled on Rust. I expect I will mostly be developing in Racket with the occasional performance hungry lib/program in Rust.
A current interest of mine is statistical programming, big data, data science, etc. Both R and Python are on the short list here with Julia trailing behind. I really like what I’ve seen in Julia, and I wouldn’t at all be surprised if I end up there; however, as a newbie learning the field, I think I’ll find much better support (educational info, existing libraries, community support, etc.) with R, so I’m starting with that until I’m convinced of switching.
At my stage, the language isn’t as important as becoming more educated in probability & statistics, machine learning, etc.
I didn’t plan it this way, but it appears that I’m going to be spending a lot of time with Racket, Rust and R.
