The Polyglot
Why more than one?
- Each computer language has different strengths and weaknesses
- Computer languages are designed for specific or general tasks
- The hardest language to learn is the first one
How to choose
- Simplicity/Familiarity
- Suitability for the task at hand:
- Task
- Performance
- Platform
- Distribution
- Maintanability/Code longevity
Hammer or screwdriver
Considerations
- Interpreted or compiled
- General purpose or domain specific
- Procedural, functional or declarative
Interpreted or compiled
- Compiled languages tend to run faster
- But the development cycle is typically slower
- A good combination is to prototype in an interpreted language, and if needed then port to compiled for speed
- Many interpreted languages offer "tricks" to achieve performance close to compiled languages, but this often requires specialized knowledge.
- But writing performante compiled code is hard
- i.e. writing performant programs is hard!
General purpose or domain specific
- Domain specific languages can simplify the learning curve and make code shorter and more manageable as they provide common functionality and abstractions
- But they can be harder to deploy on other people's systems
- And they do some things very well, but others badly or at all.
Procedural/Imperative languages
- The programmer must clearly describe the procedure and order in which things are to be done
- This often obscures the actual goal of the code for others reading it.
#include <stdio.h>
int main()
{
int a = 3;
int b = 2;
int c = a + b;
printf("%i", c);
return 1;
}
C
Functional languages
- All code is function calls
- Often, there is no persistent state
(println (+ 3 4))
Clojure
Declarative languages
- Describe the what, not the how
- Useful when building static systems (the system is static but it can respond to interactive events)
Rectangle {
width: 30
height: width
color: #ff0000
Circle {
width: parent.width /2
height: width
color: #00ff00
}
}
QML
The Polyglot
By mantaraya36
The Polyglot
- 977