Programming Languages

How I Started

Assembly

• Represents low level processor operations directly
• With minor syntactic sugar
• Non-portable (must be heavily rewritten for every new system)
• Hard to debug
• Can produce very fast or very slow code depending on the expertise of the writer

Compiled

• Code is turned into a binary executable
• This binary is then run

Interpreted

• Code is passed to an interpreter
• The code is run inside the interpreter

Statically typed

• The type of data that variables hold is static
• i.e. it must not change for the life of the variable
• most of the time it implies that the variable type must be explicitly provided

Dynamically typed

• The type of variables could change at any moment
• The type is checked whenever the variable is used
• Flexible but slower (and

int a = 54;
float b = 12.3;
float c = a + b;

a = 54
b = 12.3
c = a + b

c = "Hello World"

d = c + a # Error !

C or C++

Python

Interpreted vs. compiled

• Interpreted languages are often dynamically typed
• Interpreted languages offer faster development as code can be executed and tested in a faster workflow
• But interpreted languages offer this flexibility often at the cost of execution speed

Compiled

• C
• C++
• Objective C
• Java
• SmallTalk
• Processing
• "Arduino"

Interpreted

• Python
• Lua
• Lisp
• Go
• Javascript
• Basic

General purpose

• Provide general functionality for any purpose
• Can often be extended to offer additional specific functionality

Domain specific

• Language provides higher level abstractions for a specific domain
• Domain specific operations become simpler
• General purpose programming generally becomes harder

Domain specific languages

• R, SSPS (Statistics and numerical computation)
• Mathematica, Octave, MATLAB
• Csound
• Dot
• Scratch
• Pure Data/MAX etc.

Behind the scenes code generation

• Gibber https://vimeo.com/80698032
• Processing (a Java extension/code generator)
• Arduino (a code generator for C/C++)

Extensions to general purpose languages

• So many!!!
• Numpy and Scipy for Python (numerical/scientific computation)
• OpenGL libraries
• Basically make code that is often reused, reusable.