~3 minute readPERSONAL COMPUTERS
Key Influence of Games Is Seen in New Trends
COMPUTER games are pretty much passe today, except perhaps among a small core of die-hard fanatics. All the same, their influence on personal computing has been tremendous, not merely because of the boost they gave to "home" computer sales, but also because of their profound effect on the direction of personal computing as a whole.
At first glance, there seems to be a considerable gap, indeed a chasm, between consumer games and such practical applications as word processing, spreadsheet manipulation and data-base management. Where the influence of those games is unmistakable, however, is in the design of the hardware and software now on the market.
Consider, for example, two of the latest hardware contenders that are vying for popularity at the low end of the price scale: the Atari 520ST and Commodore's Amiga. These much ballyhooed machines stress, above all, the video aspect of computing. The advertising hallmark of both is high-resolution color graphics. A midi connector -- for musical instrument digital interface -- permitting the coupling of the machines to a music synthesizer is tossed in as an extra selling point.
Here the influence of computer games appears obvious. But is vivid color display or hi-fi sound what a business really needs most in a computer? Of course not. Although such features may well serve other useful purposes, a major attraction appears to be entertainment.
In the long run, the trend toward higher-resolution screens, at least in monochrome monitors, will benefit the computer user in a roundabout way -- if it contributes nothing else, it will increase eye comfort.
Another outcome of personal computing's game-oriented birth is the rapidly expanding application of parsing. Parsing, which may well become the hot programming concept of 1986, is the activity of separating sentences into their parts, describing the grammatical form and function of each of those parts and explaining how they relate. If the process sounds like something from your old high-school English class, it is, but in computerized garb.
Take, for example, the simple sentence "The dog whispers the truck." "The dog" is the subject, "whispers the truck" the predicate. We can break the sentence down further by saying that the subject consists of an article, "the," followed by a noun, "dog." The predicate consists of the verb "whispers" followed by its object and so on. We can diagram the sentence as well, in which case we would draw a parse tree. Of course, the sentence doesn't make sense, which is a problem.
Though it is a problem that we recognize, the computer does not. THIS is one of the major stumbling blocks facing software developers trying to imbue personal computer programs with artificial intelligence. However, crude attempts to come up with a "natural language" approach, which strives to permit computers to "understand" English rather than computerese, are already present in -- you guessed it -- games.
Verbal adventure software, first popularized by Infocom with the likes of Zork, Deadline and The Hitchhiker's Guide to the Galaxy, employ simple computer parsing to interpret a player's statements.
A computerized parser works essentially as a list manager. Long tables of acceptable, or "recognized," words are built into a program, broken down by category. Commands such as "Go," "Look," "Take" and the like might make up one table in a game. These directives could then be coupled to items from an object list such as "house," "jewels," "candle" and "pistol." To avoid the acceptance of a nonsense entry by the computer, tables or lists of grammatical rules are included in a parsing program. So, to determine what a given entry means, the software first checks its vocabulary list for the words typed in and then locates the itemized rules or rules governing how those words may be used. It would accept the coupling "Take candle," for instance, but not "Take house."
Research in such so-called natural language processing can be expected to receive a lot of publicity in the next year or so as promoters proclaim, "Look, folks, a computer that understands English!" We can also expect numerous programs promising fast information retrieval using simple English commands in an attempt to make personal computing as easy as writing home to Mother.
How close all this will come to matching the expectations it raises remains to be seen. One safe bet, however, is that at least two sentences will confound any program, no matter how the parser slices them: "This software doesn't work" and "I want my money back."
These historical, out-of-print articles and literary works have been GNUSTOed onto InvisiClues.org for academic and research purposes.
