Languages, not paradigms
I offer a brief peek at recent and older developments on programming languages and how they relate to some topics outside the field of computer programming. And building on that somewhat unconventional base I’ll explain why Clojure is totally awesome.
That’s all. No fight over paradigms or technology.
The secret of great design
In 1896 Form follows function becomes a law.
Lois Sullivan stated about everything:
“that form ever follows function. This is the law.”
It really is that simple! There is of course much to say about how this can be achieved, but that’s another post.
A short and somewhat biased history lesson
1929 People think about the concept of linguistic relativity
A cunning linguist named Sapir states that natural language constrains and shapes our thought processes. Later this is refined and known as Sapir-Whorf hypothesis.
1956 Chomsky hierarchy of languages published
Chomsky hierarchy is published. Essentially all major programming languages of later ages are defined with a “context free grammar” since there are practical ways to implement a parser (compiler) for that category of languages.
1960 A new “programming language” for so called “computers” is born
The world does not notice because most people have never seen any computers, but John McCarthy defines Lisp anyway. For fun and profit.
In it’s core it is small, elegant and extremely powerful. AI researchers go wild, people doing practical things yawn. Neither camp understands.
Most striking perhaps is the design where form and function are interconvertible. This is based on so called Lisp macros and REPL which neatly addresses Sapir-Whorf hypothesis by making the language fully extensible and unconstrained.
In fact, Lisp is “unconstrained” in a very real sense. Lisp macros are actually a full blown Turing complete metaprogramming language, which sidesteps the limitations of context free grammar required to define the core Lisp and it’s evaluator (compiler).
1972 programmers find logic
Mycroft Holmes is very excited. He finally gets to resign from his human computer post since Prolog takes his place with it’s fact-based data storage and logic programming.
1998 Java Architect denounces Java, praises Lisp
In a great talk titled Growing a Language Guy Steele, one of the main architects behind Java, admits that Lisp got it right when they made the language extensible with the macros. He states it in no uncertain terms:
“I should not design a small language, and I should not design a large one. I need to design a language that can grow. I need to plan ways in which it might grow—but I need, too, to leave some choices so that other persons can make those choices at a later time”
2000’s known programmers admit using Lisp
In 2003 Paul Graham comes out of the closet and talks openly about his addiction to Lisp.
In 2006 Steve Yegge talks about his experiences and why Java was not enough in the end though the first samples were free.
2007 Lisp is revived from the graveyard
Secretly Rich Hickey has listened to all of this. After reading a remarkable collection of extraordinarily fine literature he has created a new form of Lisp.
The time is now ripe. The voice of Rich Hickey booms:
“repent now thou Java programmers, I give thee a practical Lisp thou and all thy kin can use.”
Nobody listens.
2010 (let [datomic (loose :on :the :world)]
Rich Hickey’s hammer and anvil are steaming hot. With some friends he forges a new database which combines Prolog’s fact based view of data with a practical implementation and powerful query language. Putting time and revision history in is a nice plus.
Nobody understands Datomic but the bang of the hammer is now heard far away.
2012 and 2013 Public sightings of Clojure programmers
Clojure programmers are proven to exist outside controlled environments. Even here at Tampere we have witnessed this in 2012 (also 2013) and our capital Helsinki follows suit this year.
There is a great disturbance in the Force
What we now have therefore is a practical JVM stack where data, form and function are all interconvertible and expressible in the same very powerful, yet elegant language. There is no “impedance mismatch” in this stack.
I stand in awe.
The “functional train” is leaving the platform!
Last week we got a greenish light to go ahead with Clojure in our project. Solita has been running Clojure code in production for some time now, but this is the first real world Clojure project for me. I have been sitting on the metaphorical “functional train” for some years already. Patience is a virtue and now we are finally moving!
While Clojure is currently tied to JVM platform - a burning one one might argue - it is now reaching out to others. With Datomic and Clojurescript the stage for world domination is set.
I look forward to enjoyable, speedy and luxurious ride on this functional train.
