Computer Program

V

History

People have been storing sequences of instructions in the form of a program for several centuries. Music boxes of the 18th century and player pianos of the late 19th and early 20th centuries played musical programs stored as series of metal pins, or holes in paper, with each line (of pins or holes) representing when a note was to be played, and the pin or hole indicating what note was to be played at that time. More elaborate control of physical devices became common in the early 1800s with French inventor Joseph Marie Jacquard’s invention of the punch-card controlled weaving loom. In the process of weaving a particular pattern, various parts of the loom had to be mechanically positioned. To automate this process, Jacquard used a single paper card to represent each positioning of the loom, with holes in the card to indicate which loom actions should be done. An entire tapestry could be encoded onto a deck of such cards, with the same deck yielding the same tapestry design each time it was used. Programs of over 24,000 cards were developed and used.

The world’s first programmable machine was designed—although never fully built—by the English mathematician and inventor, Charles Babbage. This machine, called the Analytical Engine, used punch cards similar to those used in the Jacquard loom to select the specific arithmetic operation to apply at each step. Inserting a different set of cards changed the computations the machine performed. This machine had counterparts for almost everything found in modern computers, although it was mechanical rather than electrical. Construction of the Analytical Engine was never completed because the technology required to build it did not exist at the time.

The first card deck programs for the Analytical Engine were developed by British mathematician Countess Augusta Ada Lovelace, daughter of the poet Lord Byron. For this reason she is recognized as the world’s first programmer.

The modern concept of an internally stored computer program was first proposed by Hungarian-American mathematician John von Neumann in 1945. Von Neumann’s idea was to use the computer’s memory to store the program as well as the data. In this way, programs can be viewed as data and can be processed like data by other programs. This idea greatly simplifies the role of program storage and execution in computers.

---

---

VI

The Future

The field of computer science has grown rapidly since the 1950s due to the increase in their use. Computer programs have undergone many changes during this time in response to user need and advances in technology. Newer ideas in computing such as parallel computing, distributed computing, and artificial intelligence, have radically altered the traditional concepts that once determined program form and function.

Computer scientists working in the field of parallel computing, in which multiple CPUs cooperate on the same problem at the same time, have introduced a number of new program models (see Parallel Processing). In parallel computing parts of a problem are worked on simultaneously by different processors, and this speeds up the solution of the problem. Many challenges face scientists and engineers who design programs for parallel processing computers, because of the extreme complexity of the systems and the difficulty involved in making them operate as effectively as possible.

Another type of parallel computing called distributed computing uses CPUs from many interconnected computers to solve problems. Often the computers used to process information in a distributed computing application are connected over the Internet. Internet applications are becoming a particularly useful form of distributed computing, especially with programming languages such as Java . In such applications, a user logs onto a Web site and downloads a Java program onto their computer. When the Java program is run, it communicates with other programs at its home web site, and may also communicate with other programs running on different computers or web sites.

Research into artificial intelligence (AI) has led to several other new styles of programming. Logic programs, for example, do not consist of individual instructions for the computer to follow blindly, but instead consist of sets of rules: if x happens then do y. A special program called an inference engine uses these rules to “reason” its way to a conclusion when presented with a new problem. Applications of logic programs include automatic monitoring of complex systems, and proving mathematical theorems.

A radically different approach to computing in which there is no program in the conventional sense is called a neural network. A neural network is a group of highly interconnected simple processing elements, designed to mimic the brain. Instead of having a program direct the information processing in the way that a traditional computer does, a neural network processes information depending upon the way that its processing elements are connected. Programming a neural network is accomplished by presenting it with known patterns of input and output data and adjusting the relative importance of the interconnections between the processing elements until the desired pattern matching is accomplished. Neural networks are usually simulated on traditional computers, but unlike traditional computer programs, neural networks are able to learn from their experience.

	More from Encarta
	•  Presidential Myths Quiz •  Coffee break: Recharge your brain