What is Programming?

As with any learning adventure, we must begin somewhere. When learning how to write computer programs, one of the best questions to tackle first is “what is programming?” As it turns out, the answer to that question is key to understanding exactly what it is we are trying to learn.

A Simple Computer

Video Materials

At its core, a computer is simply an electronic device that is capable of following instructions to perform calculations. In computer science theory, there is a special kind of theoretical computer called a Turing Machine that represents the simplest version of a modern computer. It might look something like this, as imagined by an artist:

¹

A Turing Machine consists of an infinitely long tape that can be used to store data, and a small control box that manipulates the tape. The control box knows how to perform a few simple instructions, such as “Move Left” or “Write 0.” So, to program a Turing Machine, we must simply tell the control box which instructions to follow, and it can do it. For example, if we want the Turing Machine to write “101” on the tape, we could write the following program:

1. Write 1
2. Move Right
3. Write 0
4. Move Right
5. Write 1
6. Stop

Seems simple enough. We won’t go into the details here, but computer scientists have been able to prove that any computer program that can run on a real computer could also be performed on a Turing Machine, as long as the Turing Machine has infinite time and an infinitely long tape.

This video shows an example of what a Turing Machine might look like in real life.

So, all we really need to learn is how to write programs for a Turing Machine, right?

A Modern Computer

²

Well, it’s unfortunately not that simple. There are two major differences between a Turing Machine and a modern computer that we must deal with. First, a modern computer knows many more instructions than a Turing Machine. To learn how to write programs that a modern computer can understand, we’d have to learn an entirely different vocabulary of commands. At the same time, modern computers are very complex systems, so any program we write might not be very efficient at doing what we want.

So, to learn how to write computer programs quickly and easily, we really want to be able to do two things:

1. Use a vocabulary of commands that are familiar to us
2. Have those commands turn into programs that run efficiently on a modern computer

Compilers & Interpreters to the Rescue!

³

Developing computer programs was very difficult work in the 1950s, and many of those early programmers were looking for a better way to solve that exact problem. One of these was Rear Admiral Grace Hopper, shown above. Her team was one of the first to develop the idea of writing computer programs using English words, and then using a second program, which they called a compiler, to convert those English words into instructions a computer could understand.

Their compiler made developing computer programs much simpler, since programmers didn’t have to learn an entirely new vocabulary to tell the computer what to do. Instead programmers simply had to learn the rules of what a computer could and couldn’t understand, and the syntax, or grammar rules, of how the compiler expected the program to be written. These new programming languages that use English words are referred to as high-level languages.

Programmers would now write the source code for the program in a high-level language, and then use a compiler to generate the machine code that the computer would actually run. In addition, since the compiler was a program itself, it could make sure the machine code it generated was as fast and efficient as possible, eliminating lots of hard work programmers would have to perform to tailor each program to fit the hardware it was going to run on.

Today, programming languages such as C, C++, and Java use compilers to convert source code into machine code.

⁴

At the same time, other developers such as Steve Russell, shown above, were working on another type of program, called an interpreter, to solve the same problem. An interpreter can read source code and immediately tell the computer what steps to perform, without needing to generate the whole machine code first. This makes it much easier to write and edit programs on the fly, as the interpreter reads the source code directly each time the program runs. Today, programming languages such as PHP, JavaScript, and Python use interpreters to run the source code on a computer.

Programming

So, programming is simply the act of writing computer code in a way that a computer can run it. In most cases today, that means developing the source code for a program in a high-level language, then using either a compiler to generate the machine code for that program, or an interpreter to run the program directly on the computer from the source code. Of course, we can always write machine code by hand, but that is quite a bit more difficult.

In this class, we’ll learn how to write source code in one of two common languages, Java and Python. They both have their own unique features, especially since Java is a compiled language and Python is an interpreted language. However, as we saw above with the Turing Machine example, each language can be used to write any computer program. So, the choice of language is really more about personal preference and the unique features of each language than anything else.

This makes sense, because in general we can use both English and Spanish, as well as most other languages today, to express the same thoughts and ideas, even if we may not always have a word with the same meaning in both languages.

Running a Program

Modern computers share many theoretical similarities with the Turing Machine, but in practice they are much more advanced. In this section, we’ll discuss how a modern computer is able to run programs we’ve written in a high-level programming language like Java or Python.

Learning to program requires some basic understanding of how a computer works. In the most basic sense, a computer has a large amount of memory connected to a central processing unit (CPU). The memory consists of many bytes of data, each containing eight binary values (e.g. 01001101).

The memory itself is split into various regions, with different regions storing the machine code instructions that make up a program and the data that the program is operating on.

The CPU contains enough circuitry to interpret the instructions it is given and perform the requested mathematical and logical operations on the data provided.

Consider a single line of code in a program:

z = 5 + 7

In most languages, a line similar to this will instruct the computer add the values of 5 and 7 and save the value into a variable named z, which will be stored in some location in memory. In a machine language, this single line of code may be actually require different instructions to complete. The table below shows one possible interpretation of that line of code into machine language.

Machine language is pretty ugly, but this is really how a CPU functions. Thankfully, we have many higher-level programming languages available, such as Java and Python, so we usually don’t have to write code directly in machine language.

Estimating Program State

From this, we can come up with a pretty powerful abstraction that represents how a computer works. In programming, an abstraction is a simplified model of a complex system, keeping only the most important details to make using the model easy.

In our abstract model of a computer, we see that a program consists of an ordered list of instructions to be executed, and memory is a set of labelled boxes that can store any data our program requires. We also have a pointer that keeps track of which instruction the computer is currently executing. The list of instructions, the instruction pointer, and the memory combined all make up a program’s state when it is running on a computer. Tools such as Python Tutor or Java Tutor use a model very similar to this to help us understand how our programs actually work.

The Software Development Lifecycle

Another big idea in computer programming is the software development lifecycle. There are many different ways to approach developing a large computer program, but most of them can be divided into a few clear steps, as shown in this diagram.

Requirements

The first step is the requirements analysis phase. In this phase, a software developer works to clearly describe and understand the problem to solve and the features that the solution should have. Throughout this course, the problem statement or programming project description will form the requirements for the program to be written.

Design

Next comes the design phase. In this phase, we describe the solution in terms of the actual structure of the code. In Object-oriented programming, this generally consist of designing the classes and the interactions between them which meet the requirements.

A class is a collection of data and the methods (procedures) by which data is accessed and manipulated. Classes are generally organized around things (the shopping cart from any online store is a classic example) or functions (communication with your computer’s network card is probably handled by a class). Similar to the requirements, the design of most of the projects in this course will be provided, but it is important to pay attention to those designs since later courses may leave the design up to the developer.

Code

Following the design phase is the development, or coding, phase. In this phase, actual code is written to match the requirements and design from the previous phases.

This course primarily focuses on this phase of the software development lifecycle. In this course, we’ll cover how to write code in classes and methods to perform various operations on data. We’ll also learn how to store and manipulate data in variables and data structures.

Testing During Development

Testing while coding is an important part of the coding process. Good programmers constantly test small parts of their code, both by running the code and by continually thinking about the code, any time they complete a task. When a program fails to run correctly, we say that it contains a “bug” and the process of finding and fixing those bugs is known as “debugging.”

There are many advantages to testing while coding:

1. This allows the developer to develop their own test cases and have a better understanding of what the code is supposed to do and where any problems may be.
2. Any bugs introduced or revealed by the most recent addition of code become much easer to find.

We recommend a “code a little, test a little” approach in this class. This helps us narrow down which parts of the code are working properly and which ones contain bugs to be removed. If we write too much code without testing any of it, it can be very difficult to isolate and fix the bugs.

Test

When the development phase is complete, we move on to the formal testing phase of development. In this phase, the software is formally tested to ensure that is matches the specifications and design, usually by specially trained test engineers who can report any bugs that are found back to the developers.

In this class, the autograder that we use serves as a formal test process. It isn’t perfect and won’t catch every bug, but it does a good job of helping you determine if your program meets the specifications for the project.

Deploy

Finally, there is the deployment phase. This is where the code is packaged and released to the users. In this course, this is analogous to turning in our project and marking it complete. At this point, it is very difficult to fix any bugs without making a new release of the software.

OOP Features

In this course, we will be teaching programming from the object-oriented perspective. Both Java and Python support this programming paradigm, and it is commonly used in industry today. However, it is easier to understand some of these concepts (such as classes and objects), after first covering some of the fundamental concepts.

Thus, the first few modules of this course will contains some pieces of starter code that we’ll use without really explaining much about how they work, but rest assured that we’ll cover those concepts at the correct time in the course. Likewise, many of the early programming assignments will include some frozen starter code that provides the structure, and the point of the assignment will be simply to fill in the middle portion of the program that contains the actual code.

Before we begin programming, let’s go over some basic terminology and rules for high-level programming languages.

Syntax

First, all high programming languages have a syntax, which is a pattern of words and symbols that all correct instructions in the language much follow. So, when we receive a syntax error from the compiler or interpreter, we should know that one or more lines of code in our program doesn’t follow the rules of the language. In a very real sense, this is the grammar that all correct programming lines will follow.

Natural languages also have a grammar, but most natural language listeners can understand improper grammar. The question “Which room is the dog in?” which contains the dreaded dangling preposition, will almost certainly be correctly interpreted as “in which room is the dog?” by most listeners.

Programming languages are not like this – the slightest violation of a grammar rule will result in code that most likely will not run. So, as developers we must learn to be very precise and exact in our code. This is a difficult skill to learn, but we’ll develop it over time with practice.

Semantic

Second, all programming instructions have a specific meaning, known as the semantics of the language. When one of our programs compiles and executes but produces an incorrect result, this is due to a semantic error (sometimes known as a logic error) in our code.

Unfortunately, the compiler or interpreter isn’t able to help us find any semantic errors, but we can find them through careful testing of our code.

Initial Vocabulary

Finally, all programming languages have various defined structures, which we can think of as the “parts of speech” in the programming language. This is similar to the concept of nouns, verbs, and adjectives in the English language.

This is a list of some of the most common parts in a programming language:

• keywords - a word that has a special meaning. These words tell the program exactly what to do, and we cannot use these words as identifiers. For example, in both Java and Python the word class is a keyword. We cannot use a keyword for any purpose other than the one(s) intended by the programming language.
• literal – text evaluated as itself, for example numbers (such as 6.626) are meant to represent themselves. Similarly, text literals are most often enclosed in double quotation marks (such as "wildcat"). Text literals are most commonly referred to as strings.
• identifiers - any class, method, or variable name is considered an identifier. Each language will have both rules (syntax) about what can be identifiers and conventions (style guides) about how to construct them.
• variable – a programmer-defined piece of data. Programmers name their variables with an identifier.
• method - a piece of code that performs an action in our program. Methods are sometimes referred to as functions or subroutines as well, but we’ll use the term method. Programmers name their methods with an identifier.
• class – a collection of variables and the methods used to manipulate that data. Classes lay at the heart of object-oriented programming and will be the focus of the last third of this course.
• symbols - any non-alphanumeric character which has special meaning. Common ones include the math symbols (+ - / * =), brackets of all forms (() [] {}), and then some not so obvious ones ( . , ; " !).

Program Structure

When writing code in an object-oriented language, many times we must use some additional syntax and structures in our programs that are required by the language, even though they aren’t strictly required for the program we are trying to develop.

In this section, we’ll briefly discuss some of those limitations. Later on in this course, we’ll come back to these concepts and explain them more fully, but for now we’ll provide some of this structure as starter code for the first few modules.

Everything Is In a Class

Classes are the blueprints of objects. In an object-oriented program, everything is inside a class definition. In this class, every file will contain exactly class definition or class-body.

// Java
public class ClassName {
    class-body
}

# Python
class ClassName:
    class-body

Code Block Delimiters

The class-definition is a “code-block.” Code-blocks signal a boundary to the compiler. This helps the compiler manage identifier names, memory and other things.

Code-blocks are delimited, or set apart, by various symbols. In Java, this is done with brackets { code block }. Spacing is often also used to make the code more readable, so that everything in a block lines up.

In Python, each code block is indented to the right. In this class we will use 4 spaces. When a line ends with a colon, Python expects the next line to be the start of a new code block (and therefore indented).

public class Foo{
    statement one
    statement two
}

public class Bar{
    statement_three
}

class Foo:
    statement_one
    statement_two

class Bar:
    statement_three

In the preceding examples, statement_one and statement_two are the code block for the class named Foo, and statement_three is part of the code block for the class named Bar.

Every Program Starts Somewhere

By tradition (and by rule in many languages), object-oriented programs start in a method called main().

// Java
public class ClassName {
    public static void main(String[] args){
        // method-body
    }
}

# Python
class ClassName:

    @classmethod
    def main(args):
        # method-body

For the first several modules, all of the code we need to write will be in the method body for the main() method. This area will be clearly marked in the starter code using a comment similar to WRITE YOUR CODE HERE.

Hello World

Tradition dictates that the first program in any introduction course is Hello World. Hello World simply prints the text “Hello World” to the screen.

Below is an example of an object-oriented version of this program in both Java and Python:

// Java
public class HelloWorld{
    public static void main(String[] args){
        System.out.println("Hello World");
    }
}

# Python
class HelloWorld:
    @classmethod
    def main():
        print("Hello World")


HelloWorld.main()

In each program, the actual work is done by the line containing the word print. In the rest of this module, we’ll explain the details behind the structure of this program.

Summary

We’ve now written our first program in a real, high-level programming language. While the program may only consist of a few lines of code at most, it is still a very big step toward writing bigger and more advanced programs.

Computer technology has some quite a long way since the 1950s, but the same needs that drove the development of compilers and interpreters continue today to drive the development of more advanced programming languages and related tools. It’s a very exciting field to experience firsthand, and once we understand a bit of code, we’ll be able to see it for ourselves.

In the next chapter, we’ll dive in head first to learn all about how to store and manipulate various types of data in our programs.