The promises of functional programming

The article starts exposing in certain way how are related the advances in hardware and the advances in software, explaining that the development of more powerful hardware permits the development of more complex software. Regardless of the easier way to implement and run imperative code, the advantages of functional code and his capability of concurrent and parallel programming, keep this coding technique’s fame. Also, the easier way to test this kind of code help too. The fundamental principle of functional programming is explained as “the mapping from input values to output values”, and is important to understand this, to understand the next explanation, the functional programming doesn’t work with variables, it works with functions, functions that aren’t supposed to change variables, and doesn’t have side effects.

The functional programming considers functions as data, this means that we have functions that call functions or return functions as parameters. The promise of the functional programming is basically, the concurrency and the parallelism. The first one refers to the capability to have in existence several execution threads operating on the same data, this means, that we could be capable to work at the same time with the same data in different environments. In other hand, the parallelism means to the division of computational task into multiple communicating processes running in parallel, which means that the work that the computer needs to do, in this case can be divided and distributed into different processors to reduce the time needed to produce a result.

These two characteristics are powerful tools that may be someday the key to archive more computational power and speed, and maybe the structural and object-oriented programming will be totally replaced with this new kind of languages. And these techniques will allow us to maximize the power and use of the multiple cores architectures.

Comentarios

Publicar un comentario

Entradas populares de este blog

Triumph of the Nerds Part II

The second part of the film is also interesting, this part explores at first how IBM realized that there was a new market rising, the market, of the personal computers. This part explains how IBM realized that they may lost all this rising and valuable market if they didn’t quickly develop a product to compete with Apple and the other little personal computer manufacturers. So, they decided to take advantage of their resources and bought as many as possible components to construct a computer, but they still had another problem, the Operating System. IBM decided to make a kind of outsourcing hiring Microsoft to develop this OS, but they weren’t their first option, at first, they tried to hire and bought another OS, but for some inconvenient they ended hiring Microsoft and earning the rights for the programming language that Microsoft (Bill Gates) developed, and, they obtained their OS.  It was a matter of time for Microsoft to realize that the business relationship with IBM wa...
Leer más

Dick Gabriel on Lisp

At the beginning, Dick told us his background, and it surprised to me, that he said he have some mathematical degrees and one in artificial intelligence, because last semester at my algorithm’s analysis class our professor Salvador E. Venegas said us that something beneficial for an engineer was, to “get the full package”, because there was always engineers that need to study mathematics in order to advance in their jobs, because the understanding of mathematics reduced the complexity of developing algorithms. I didn’t know Lisp has an interpreter instead of a compiler, something interesting because as I know it reads and traduce the program line per line, instead like the hole program first, as the compiler do. I didn’t understand well how the meta part of “eval” work, what I understood was that “eval” work in certain way as recursion. The “common Lisp” is interesting, how Lisp started to get “subdivisions” in different ways to get closer to user’s needs gave Lisp variety and trying ...
Leer más

Rich Hickey on Clojure

Clojure’s design is to be hosted, this caught my attention, because as it was explained at the previous podcast, the Yahoo Store was designed in Lisp, so, this makes of Clojure a good option to develop embed systems as I see. Lisp’s compiler doesn’t compile text, it compiles data structures, I still didn’t understand how Clojure can compile, read or execute code at the “same time” but it seems that the work with data structures is the key to make this posible. In Clojure, programming and metaprogramming are the same thing.  Macros in Clojure and Lisp are simple data structure processing code, is interesting this way to explain how to “write programs that later will write another program”. “Programming in large is a collective activity, it involves people, libraries, and it’s important to be able to interact with what other people have do in other languages”, this idea is especially interesting, due to the fact that is true, nowadays there are no just one programming language, ...
Leer más