The Atari Home Computer: Human engineering as art

9.1 The Atari Home Computer: Human engineering as art

The Atari 800 and its less expensive sibling, the Atari 400, garnered a respectable share of the home-computer market in the early 1980s. Primarily targeted as a machine for game enthusiasts, its claim to fame was its advanced graphics and sound capabilities. Star Raiders, a Star Trek simulation incorporating stars and photon torpedoes whizzing by in 3-D hyperspace, propelled it to a strong position in the home market until it was overtaken by the popular IBM PC and the bargain-priced Commodore 64. Today its "player-missile" graphics (sprites) and display list processor are considered primitive.

The designers of Unix and the Atari Home Computer's operating system had a common goal in that they both wanted to build a system for playing games-namely, Space Travel on Unix and games in general on the Atari. (Pac-Man?) That may be one reason why they were successful. People will often work long and hard on something that brings them pleasure. Today, many developers write software for Linux for the sheer fun of it. To them, this is entertainment, geek-style. Beyond a need for survival and social order, people have a built-in need to be entertained. It may seem strange to those outside of the computer world, but many Linux geeks find software development to be a great pastime.

Chris Crawford, a member of Atari's staff at the time, heavily influenced the design of the Atari Home Computer's operating system. His application software (games, really) set the standard by which all later software for the machine was to be judged. Crawford documented much of his design philosophy in Appendix B of De Re Atari: A Guide to Effective Programming. While the section's main focus is human engineering, it also sheds much light on Crawford's general approach to computing.

Crawford views the computer as an intelligent being lacking outward physical traits. Its thought processes are "direct, analytical, and specific." He contrasts these with the thought patterns of human beings, which are "associative, integrated, and diffuse." The differences in these thought processes create a communication barrier between these human beings and this smaller intelligence, or homunculus, as he calls it. The goal of the programmer, he believes, is not so much to make the homunculus more powerful as it is to break down the communication barrier between the homunculus and the person interacting with it.

This is where Crawford's approach and the Unix philosophy part company. For while Crawford emphasizes that the most important thing a piece of software can do is to communicate with a human being, a Unix program's highest priority is to communicate with another program. That the Unix program must eventually interact with a human being is secondary.

In their approach to user interface design, the Atari Home Computer and the Unix operating system are galaxies apart. Most software for the Atari machine strives for "closure," or the narrowing of options. According to Crawford,

The ideal program is like a tunnel bored through solid rock. There is but one path, the path leading to success. The user has no options but to succeed.^[1]

To create the optimum Atari Home Computer program you must limit choices. The user should be given only one way to perform a task, and that way should be obvious. There must be little opportunity for error, even if it means limiting choices. Using Crawford's line of thinking, the flexibility and adaptability found in the Unix and Linux worlds can only lead to chaos.

Modern game machines implement the same philosophy as the software found on the Atari Home Computer. They limit choices. They make it nearly impossible for kids to make a wrong selection. They ensure that the gamer will usually have a pleasurable, if somewhat limited, experience with the computer.

Limiting choices within a computer system can be challenging. For while users are constantly seeking out the latest whiz-bang thrill, they want to do so within the confines of a constrained world. They want to be adventurous, but not too adventurous. They must be able to walk away from the ride afterwards so they can ride again.

The Atari approach suggests that if the average person is given a gun, he is likely to shoot himself in the foot. By contrast, the Unix system effectively hands the uninitiated user an assault rifle, plugs in 20 rounds, and points it at his foot. A person with his foot shot off doesn't walk away from the ride very easily.

As you might expect, the Unix way can wreak havoc. Choices abound in the Unix environment, and limitations are few. There usually exist at least a dozen ways to perform any task. This comparative abundance of freedom often pushes new users' patience to the edge, as ill-chosen commands plunder their data and leave them wondering how to salvage it. But eventually most Unix power users find that understanding replaces chaos, and flexibility and adaptability find favor over artificial limitations. For those who have no interest in becoming Unix power users, the various Linux distributions come with GUIs that hide most of the complexity underneath the covers.

Mainstream computer users have grown quite comfortable over time with a progressively narrower set of choices. At one point, there were several office application suites in common usage on PCs, each offering its particular set of features and strengths. Today Microsoft Office clearly dominates this space. Many business offices refuse to use anything else. Some say this simplifies things in that everyone can share documents more easily if the same application suite produces them all. The downside is that this limits creativity. If the document format were made a public standard, then new and interesting applications would come along that would adhere to the standard and offer added value beyond that provided by a single vendor. As good as Microsoft Office is-it is a feature-rich application suite-one must believe that Microsoft doesn't hold the world's entire cache of creativity captive in Redmond.

One has only to look at the cable-television market to realize that the number of choices that consumers are faced with is increasing, not decreasing. About 40 years ago, back at the tail end of the Industrial Age, you had a choice of three television networks in America. Today the typical digital cable system offers hundreds of channels. And that number will only increase as the population grows and its viewing interests become more diverse.

In the Information Age, the number of choices increases dramatically. The "tunnel through solid rock" becomes a series of passageways each arriving at an end point tailored for a particular user. Like the users of the Atari Home Computer, new or occasional computer users require user interfaces that limit choices, making it difficult to fail. Intermediate users want more features and options but not at the expense of making the simple tasks too difficult. No one wants to drink from a fire hose directly. The "power users" want access to all that the system offers without the impediments of user interfaces that hide their options.

This is where Linux steps in. With GUIs like KDE and other environments, it offers multiple entry points for novice users. As these environments improve, Microsoft Windows and other GUI-based OSs will lose ground as the only game in town for new users, slowly succumbing to an onslaught of desktops, some of which may be even easier to use. Power users under Linux find that its distinctly open architecture and interfaces leave room for plenty of flexibility in ways that closed architectures cannot.

Now, if we could just get Chris Crawford to write a really cool version of Star Raiders for Linux.