What Is Computer Programming?
Now available here in Finnish thanks to Oskari Laine, Helsinki, Finland.
Mikä Computer Programming?
Today, most people don’t need to know how a computer works. Most people can simply turn on a computer or a mobile phone and point at some little graphical object on the display, click a button or swipe a finger or two, and the computer does something. An example would be to get weather information from the net and display it. How to interact with a computer program is all the average person needs to know.
But, since you are going to learn how to write computer programs, you need to know a little bit about how a computer works. Your job will be to instruct the computer to do things.
Basically, writing software (computer programs) involves describing processes, procedures; it involves the authoring of algorithms. Computer programming involves developing lists of instructions – the source code representation of software The stuff that these instructions manipulate are different types of objects, e.g., numbers, words, images, sounds, etc. Creating a computer program can be like composing music, like designing a house, like creating lots of stuff. It has been argued that in its current state it is an art, not engineering.
A fair question you may have is “Why should I learn how to program a computer?”
An important reason to consider learning is that the concepts underlying this will be valuable to you, regardless of whether or not you go on to make a career out of it. One thing that you will learn quickly is that a computer is very dumb, but obedient. It does exactly what you tell it to do, which is not necessarily what you wanted. Programming will help you learn the importance of clarity of expression.
Mark Guzdial has a blog post, Why should we teach programming, that elaborates on many reasons learning computer programming is important.
One final reason, it can be lots of fun! A co-worker once said: “I can’t believe I’m paid so well for something I love to do.”
So. just what do instructions a computer understands look like? And, what kinds of objects do the instructions manipulate? By the end of this lesson you will be able to answer these questions. But first let’s try to write a program in the English language.
Programming Using the English Language
Remember what I said in the Introduction to this lesson?
Before we see what a computer programming language looks like, let’s use the English language to describe how to do something as a series of steps. A common exercise that really gets you thinking about what computer programming can be like is to describe a process you are familiar with.
Rather than write my own version of this exercise, I searched the Internet for the words “computer programming sandwich” using Google. One of the hits returned was http://teachers.net/lessons/posts/2166.html. At the link, Deb Sweeney (Tamaqua Area Middle School, Tamaqua, PA) described the problem as:
When this exercise is directed by an experienced teacher or mentor it is excellent for demonstrating how careful you need to be, how detailed you need to be, when writing a computer program. A demonstration of this exercise is available on YouTube.
Programming in a natural language, say the full scope of the English language, seems like a very difficult task. But, before moving on to languages we can write programs in today, I want to leave on a high note. Click here to read about how Stephen Wolfram sees programming in a natural language happening.
Programming Using A Blocks Environment
A little less than fifteen years ago, Mitchel Resnick and friends at MIT introduced a programming environment called Scratch. It provides a new approach to teaching computer programming through a graphical user interface that eliminates the possibilty of making certain mistakes common in text-based programming.
Programs are constructed by connecting blocks, each representing some functionality available in the system. Figure 1.1 shows a simple program that asks the person running it for their name and then says hello. Color is used for categories that the blocks belong to. Notice that the ask block and the corresponding answer block are the same shade of blue. The shapes of the blocks determine where they can be placed to form an acceptable program. The rounded green join block fits into a rounded hole in the violet say block.
One of Scratch’s strengths is the ease with which students can construct games and animated simulations and stories. Another strength is the Scratch website itself which provides access to many tutorials and a community of users with programs they’ve written. To learn more about Scratch, visit the Scratch Wiki.
Blocks-based programming is a great way to get started. But, as the size of the programs you want to write grows or you need a feature not available in Scratch, it’s time to switch to text-based programming.
Programming Languages – High-Level Languages
Almost all of the computer programming these days is done with high-level programming languages. There are lots of them and some are quite old. COBOL, FORTRAN, and Lisp were devised in the 1950s. As you will see, high-level languages make it easier to describe the pieces of the program you are creating. They help by letting you concentrate on what you are trying to do rather than on how you represent it in a specific computer architecture. They abstract away the specifics of the microprocessor in your computer. And, all high-level languages come with large sets of common stuff you need to do, called libraries.
In this introduction, you will work with two computer programming languages: Logo and Java. Logo comes from Bolt, Beranek & Newman (BBN) and Massachusetts Institute of Technology (MIT). Seymour Papert, a scientist at MIT’s Artificial Intelligence Laboratory, helped Wally Feurzeig at BBN design Logo in the 1970s. More research of its use in educational settings exists than for any other programming language.
Java is a fairly recent programming language. It appeared in 1995 just as the Internet was starting to get lots of attention. Java was invented by James Gosling, working at Sun Microsystems. It’s sort-of a medium-level language. One of the big advantages of learning Java is that there is a lot of software already written ( see: Java Class Library) which will help you write programs with elaborate graphical user interfaces that communicate over the the Internet. You get to take advantage of software that thousands of programmers have already written. Java is used in a variety of applications, from mobile phones to massive Internet data manipulation. You get to work with window objects, Internet connection objects, database access objects and thousands of others. Java is the language used to write Android apps.
So, why do these lessons start with the Logo programming language?
No other computer programming language has the depth of research as Logo, based on its use in educational settings. It’s roots are in the development of interactive learning environments. Wally Feurzeig was researching the use of a timeshared computer to improve teaching mathematical concepts while at BBN (Bolt, Beranek, and Newman). The question he wanted to answer was whether kids would embrace the new technology and learn using it. With some success demonstrated using an existing programming language, Wally contracted Seymour Papert to help with Logo’s design. Seymour wrote the functional specification for Logo. Daniel Bobrow then wrote the first Logo interpreter. Since these early days, hundreds of books and research papers have been written about its use in the classroom. Cynthia Solomon, who started MIT’s Logo Group with Dr. Papert, has put together a comprehensive website on Logo: logothings.wikispaces.com.
I like using the Logo language to teach introductory programming because it is very easy to learn. The faster you get to write interesting computer programs the more fun you will have. And. having fun is important! But do not let Logo’s simplicity fool you into thinking it is just a toy programming language. Logo is a derivative of the Lisp programming language, a very powerful language still used today to tackle some of the most advanced research being performed. Brian Harvey shows the power of Logo in his Computer Science Logo Style series of books. Volume 3: Beyond Programming covers six college-level computer science topics with Logo.
Both Logo and Java have the same sort of stuff needed to write computer programs. Each has the ability to manipulate objects (for example, arithmetic functions for working with numbers). Each lets you compare objects and do a variety of things depending on the outcome of the comparison. Most importantly, they let you define named procedures. Named procedures are lists of built-in instructions and other named procedures. The abstraction of naming stuff lets you write programs in a language you yourself define. This is the stuff that programming is really all about, as you will see.
Just to give you a feel for what programming is like in a high-level language, here’s a program that greets us, pretending to know English.
This is one of the simplest programs that can be written in most high-level languages. PRINT is a command in Logo When it is performed, it takes whatever follows it and displays it. The “Hello world” program is famous; checkout its description on Wikipedia by clicking here.
In addition to commands, Logo has operators that output some sort of result. Although it’s a bit contrived, here is a program that displays the product of a constant number (ten) and a random number in the range of zero through fourteen.
In this source code, the PRINT command’s input is the output of the PRODUCT operator. PRODUCT multiplies whatever follows it by whatever follows that and outputs the result. So, PRODUCT needs two inputs. RANDOM is an operator that outputs a number that is greater than or equal to zero (0) and less than the number following it. So, PRODUCT gets its second input from the output of RANDOM.
Figure 1.2 shows a plumbing diagram, a graphical representation of how all these procedures fit together.
Still confusing? Don’t worry, we will get into the details of Logo operators in lesson 8.
Finally, here’s a snipet of advanced Logo source code, just to give you a feeling for what it looks like. This is a procedure definition for selecting the maximum number from a list of numbers.
Again, do not worry if you do not understand exactly how this procedure works. It will be a while before you will be writing anything like this. But, I want you to see that the words that make up the program’s instructions and the instructions themselves are similar to English sentences, e.g., the first line and a half in the procedure are similar to the sentences:
So, a high-level programming language is *sort-of* like English, just one step closer to what the language a computer really understands looks like. Now let’s move on to what a computer’s native language looks like when it is given a symbolic representation.
Programming Languages – Assembler Language
One abstract layer above a computer’s native language is assembler language. In assembler language, everything is given human-friendly symbolic names. The programmer works with operations that the microprocessor knows how to do, with each given a symbolic name. Objects in the microprocessor and addresses of stuff in the computer’s memory can also be given meaningful names. This is actually a very big step over what a computer understands, but still tedious for writing a large program. Assembler language instructions still have a place for little snipits of software that need to interact directly with the microprocessor and/or those that are executed many, many, many times.
Table 1.1 is an example of DEC PDP-10 assembler language, a function that returns the largest integer in a group of them, named NUMARY. The group contains NUMNUM members.
Hopefully this gives you feel for how primitive computer instruction sets are. I’m not going to go into the details of every instruction. If you want to go through it in detail on your own, the PDP-10 Machine Language is detailed here.
A few points I want to expose you to are the general kinds of things being done.
- moving objects (numbers) into the computer’s registers – very fast temporary storage,
- decrementing the value in a register,
- comparing the contents of a register to some value in the computer’s memory, and
- transfering control to an instruction that’s not in the standard sequential order – down the page.
So, as you’ve seen, higher-level programming languages provide similar functionality and in a form that is closer to the English language.
But there is a problem with assembler language – it is unique for every computer architecture. Although most deskside and notebook computers these days use the Intel architecture, this is only recently the case. And. a variety of computer architectures are commonly used in game systems, smart phones, tablets, automobiles, appliances, etc.
Ok, we are almost at a point where I can show you machine language, the *native* language of a computer. But for you to understand it, I’m going to have to explain how everything is represented in a computer.
Inside Computers – Bits and Pieces
Basically, computer instructions perform operations on groups of bits. A bit is either on or off, like a lightbulb. Figure 1.3_a shows an open switch and a lightbulb that is off – just like a transistor in a computer represents a bit with the value: zero. Figure 1.3_b shows the switch in the closed position and the lightbulb is on, again just like a transistor in a computer representing a bit with the value: one.
A microprocessor, which is the heart of a computer, is very primitive but very fast. It takes groups of bits and moves around their contents, adds pairs of groups of bits together, subtracts one group of bits from another, compares a pair of groups, etc. – that sort of stuff.
Inside a microprocessor, at a very low level, everything is simply a bunch of switches, also known as bits – things that are either on or off! Time to expand on how this is done; first let’s explore how groups of bits can be used to form numbers.
Numeric Representation With Bits
Computers are full of zillions of bits that are either on or off. The way we talk about the value of a bit in the electical engineering and computer science communities is first as a logical value (true if on, false if off) and secondly as a binary number (1 if the bit is on and 0 if it’s off). Most bits in a computer are manipulated in groups, so we humans need a way to describe groups of bits, things/objects a computer manipulates. Today, bits are most often grouped in quantities of 8, 16, 32, and 64.
Think about how you write down sequential numbers starting with zero: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, etc. Our decimal number system has ten symbols. In this sequential series, when we ran out of symbols, we combined them. You learned how to do this so long ago, in grade school, that today you just naturally think in terms of single digit numbers, then tens, hundreds, thousands, etc. The decimal number 1234 is one thousand, two hundreds, three tens, and four units.
So, how does the binary number system used inside computers work?
Well, with only two symbols, we would write the same sequential numbers as above: 0, 1, 10, 11, 100, 101, 110, 111, 1000, 1001, 1010, 1011. The decimal number 1234 in binary is 10011010010.
Since even reasonable numbers that we use all the time make for very long binary numbers, the bits are grouped in 3s and 4s which are simple to convert into numbers in the octal and hexadecimal number systems. For octal, we group three bits together. Take the binary equivilent of decimal 1234, 10011010010, and put spaces in between each group of three bits – starting at the right and going left.
10011010010 = 10 011 010 010
Now use the symbols 0, 1, 2, 3, 4, 5, 6, and 7 (eight symbols, so OCTAL) to replace each group.
10 011 010 010 = 2 3 2 2 = 2322
The octal representations of the binary patterns are certainly easier to read, write, and remember than the binary counterparts. An even more compact representation can be achieved by grouping the bits in chunks of four and converting these to hexNumerals.
When you group four bits together and use sixteen symbols (0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, A, B, C, D, E, and F) as their abbreviations, you have a hexadecimal representation.
10011010010 = 100 1101 0010 = 4D2
As you continue to explore how computers work, you’ll hear more about numbers expressed in octal and hex; these are just more manageable representations of binary information – the digital world.
Table 1.2 compares the decimal, binary, octal, and hexadecimal number systems.
So, if the most common groupings of bits in a computer are 8, 16, 32, and 64, what kinds of numbers can these groups represent?
A group of eight bits has binary values 00000000 through 11111111, or expressed in decimal 0 through 255. A group of sixteen bits has binary values 0000000000000000 through 1111111111111111, or decimal 0 through 65535. I’m not going to type in binary representations for groups of 32 and 64 bits. The range of decimal values for a group of 32 bits is 0 through 4,294,967,295. The range of decimal values for a group of 64 bits is 0 through 18,446,744,073,709,551,615 – or almost eighteen and a half quintillion.
But wait. these numbers are all positive (Whole Numbers). If we are going to allow for subtraction operations on numbers, which can result in negative numbers, we need Integers. Modern computers use one bit in each of the groups to represent the sign (positive or negative) when the groups are used to represent integers. Table 1.3 shows the range of numbers that can be represented with groups of 8, 16, 32, and 64 bits.
That’s about as deep as I want to get into the representation of numbers in computers and the binary, octal and hexadecimal number systems. Yes, computers have division operators but I am not going to cover numbers that include fractional parts, i.e., the “rational” and “irrational” numbers due to the complexity of their implementations. If you want to read more, I googled and found what looks like a good place for you to read more. Start at All About Circuits – Systems of numeration and read through it and continue on for a few more web pages in the series. The link to binary number at the start of this section points to a Wikipedia page that also will give you additional depth including the history of the binary number system.
Symbols as Bits – ASCII Characters
Ok, so numbers are simply groups of bits. What other objects will the computer’s instructions manipulate? How about the symbols that make up an alphabet?
It should come as no surprise that symbols that make up alphabets are just numbers, groups of bits, too. But how do we know which numbers are used to represent which symbols, or characters as I’m going to call them from this point on?
It’s all about standards. In these lessons, we will use the American Standard Code for Information Interchange (ASCII) standard. It is so ubiquious that it even has its own web page, www.asciitable.com.
Let’s walk through a couple of examples, entries in the table. Here are some characters, their decimal value and their binary value which is then transformed into an octal number.
Table 1.4 is a small slice from the full ASCII character set, just enough to give you a flavor of its organization.
Here’s a little trivia, from days long past when computers were so slow (compared with today).
- Look closely at the binary representations of uppercase and lowercase letters. You can convert uppercase to lowercase by turning on a single bit. Or clearing the bit converts lowercase to uppercase.
- Clearing two bits will convert an ASCII digit to its numeric value. Setting the same bits converts a number in the range 0. 9 into its ASCII character representation.
Check Your Understanding So Far
Take a moment to see if you understand the explanation of binary numbers.
HERE is a jLogo program that converts an 8-bit byte into numbers and ASCII characters. Try it out! .
The image on your computer’s display (actually all digital stuff) consists of a bunch of colored points called pixels. A pixel is an object. It has a color and a position (its coordinates) which consists of the row and column it is at. Figure 1.4 shows an artist’s rendition, a magnification of a display with a circle drawn in yellow. The tiny black dots are the pixels and the big yellow dots are the pixels that have been colored.
As an example, to display a thin vertical line, the color values of a column of pixels are set to the desired color of the line. If you want a thicker vertical line, you set the color values of the pixels of a group of consecutive columns to the desired color. Figure 1.5 shows a red line that’s a single pixel wide and an orange line that’s three pixels wide. The orange line is actually a very thin rectangle.
Follow this link to an applet which lets you experiment with painting magnified pixels.
So, the location of each pixel is obviously specified by a pair of numbers; what about the pixel’s color?
So, just as groups of bits represent numbers and symbols, they are used to form pixels.
Optional exploration: Why red, green, and blue? Why these colors? There are many good explanations on the Internet. If you are interested, search the net or checkout the Wikipedia entry for color vision.
Ok. I’ve exposed you to a variety of objects that you commonly see when you are using a computer, things that can be manipulated with instructions in a computer programming language. Now let’s move on to the computer’s instructions, one more thing that is just a bunch of bits!
The Microprocessor’s Language
So, all a computer has in it is bits. You’ve seen how they are used to represent stuff, pixels, numbers and characters. I’ve mentioned that computers perform operations on the bits, like move them around, add pairs of them together, etc. One final obvious question is: how are instructions that a computer performs represented?
Well, if you instructed a computer in its native language (machine language), you would have to write instructions in the form of (yes, once again) binary numbers. This is very, VERY hard to do. Although the pioneers of computer science did this, no one does this these days.
Just to give you something to look at, just to compare, Table 1.5 shows what the assembler language program in Table 1.1 could look like assuming that the machine instructions are loaded into memory at addresses 100 through 107. Also, the group of numbers starts at memory address 111 and the size of the group is in memory address 110.
A detailed explanation of any computer’s instruction set is beyond what can be presented here. I just wanted you to see how the symbolic information in assembler language programs needs to be converted to numbers (bits) before a computer can perform it.
The lesson walks you through programming a very simple robot computer. And, I wrote a simulator for the Robot Computer in jLogo that you can play with HERE.
No introduction to computer programming would be complete without at least mentioning debugging. The term refers to the discovery and correction of mistakes in computer programs. The computer is doing what you instructed it to do, not what you meant it to do. If you enjoy puzzles, there’s a good chance you will find the process of debugging an interesting challenge.
The origin of the term came from a bug (a moth) found in a relay of a computer in 1947, by Admiral Grace Murray Hopper. She found why her program was not working. Figure 1.6
Debugging is a process. The good news for us is that mistakes in introductory level programs are not that hard to diagnose and fix. It’s basically narrowing in on the instruction, or two, that are not doing what you intended. Steps you take are like solving Sudoku or Mastermind puzzles.
Debugging a program can be done in steps that match the Scientific Method.
In the observation step, you think hard about what is happening versus what you expected to be happening. An example is a program that is drawing something, but the jumble of lines you see on the display does not look right. Or, your program has a button that doesn’t do anything when you click on it. As a programmer, you approach these bugs like the legendary Sherlock Holmes approached his cases. Review the program you’ve written and ask questions like: “How could these instructions produce what is happening?”
Observation should lead you to the point where you can make a hypothesis about the bad behavior. For the drawing not coming out right, a hypothesis might be something like: “If computing the metrics (orientation, length, . ) of the first line is off, that might produce what I’m seeing.” For a button not working, a hypothesis could be: “If the mouse’s location when it is clicked is computed wrong, the program will ignore the click.” The hypotheses made are often based on intuition. This is because usually the person debugging either wrote the program or at least modified it.
Given a hypothesis, the next step is to figure out how to test it by predicting what should be happen if the hypothesis it correct. Continuing with the example of the misbehaving drawing program, let’s say that you notice some source code that might not produce proper results if what it is given is not in the bounds expected. This code could produce strange results. This is a prediction.
Finally, either the program is modified or a debugging feature in the programming environment is used to test the prediction. Modification of the program can be the addition of instructions that print stuff on the display. Most programming environments include debugging features, like tracing or setting breakpoints to suspend a program to examine its internal state. Either way, the programmer gathers more information about what the program is actually doing, which is, in the case of bugs, not what you expect.
Testing, even if it does not provide an answer, gets you additional information that can be used to repeat the process. You narrow your exploration until you find your mistake.
Finally – Who was the first computer programmer?
Ok. since you are reading this, accessing this web page on the net, you have Google and it takes a fraction of a second to answer this question.
Computer programming is composing/authoring of a process/procedure for doing something, the source code representation of algorithms – in great detail.
Since computers do not understand English and it would be impossible for a human to write a large program as a series of binary numbers that the computer can understand, we need something in between. High-level programming languages currently fit in this category. Given a programming language that you have chosen, you then follow its rules for composing statements (or expressions) that instruct the computer to do what you want.
Because a computer is simply a very fast manipulator of bits (ones and zeros), through the the power of abstraction, computer scientists have layered levels of object representation and functionality, one on top of another. We have now been working on refining and extending these layers for over half of a century. Table 1.6 should give you a feel for where we stand with computer programming languages today.
In the next lesson you’ll start to write programs in Logo!
Examples: Mathematica, SQL
Examples: Logo, Python
Example: Intel X86
Example: Intel X86
Gabriele Cirulli has created a game, very popular at the moment, which will teach you the powers of two (the binary number system) as a side effect. But beware, the game has been tagged very addictive.
Click here to go to the game’s website and play – and learn.
Cisco’s Learning Network provides a game where you can test your ability to do binary to decimal and decimal to binary conversions. If you do not mind registering with Cisco, you can find it here:
Click here to go to the website to test your ability.
The Book “Blown to Bits” is available on-line for free!
Find parts of it that interest you and read more about BITS.
This work ( BFOIT: Introduction to Computer Programming , by Guy M. Haas ),