Phase IV: Shakedown and Consolidation

With the Atari 2600 and Apple ][ on the market, there was a clear sense of direction in the games business. Thus, the period comprising the late 1970s and early 1980s is sometimes dubbed "The Golden Age." All that was needed was to keep improving these initial products, to create better platforms and games at each iteration. On the console side, Atari reigned supreme for many years, even when Nolan Bushnell left the company in 1978.

Following the departure of its founder, the company enjoyed prosperity due to the huge number of games being published. This period is also characterized by the entrance of Japanese developers into the equation. Taito (a pachinko manufacturer) introduced Space Invaders; Nintendo (remember the Hanafuda card company?) began by selling an Othello arcade game only to later introduce Donkey Kong (the first game by now-celebrity designer Shigeru Miyamoto) in 1981; and Namco (known as the Nakamura Manufacturing Company, maker of merry-go-rounds) introduced such classics as Galaxian, Pac-man, Galaga, and Pole Position.

Pac-man, whose initial character concept comes from a pizza with a slice missing, is based on a Japanese folk hero called Paku, known for his appetite. The game was to be called Puckman, but the potential risk of American graffiti artists changing the P to an F made Namco go for Pac-man as the final name.

Atari's golden days were about to end. The company did not keep a careful watch on the quality of games for the 2600, and some highly anticipated games for the platform were a flop. Pac-man and E.T. games were released with great hype, only to see consumers reject them due to their low quality. The legend says that thousands of copies of these games ended up in a massive landfill in New Mexico. By December 7, 1982, Atari announced that 2600 sales had, for the first time, failed to meet predictions, making its parent company, Warner Communications, suffer a 32 percent stock price drop in a day. In early 1983, with Atari still struggling to recover from E.T., a crash occurred in the video game market. Retailers had tons of unsold games on store shelves, and many of them offered very low quality. Game prices dropped sharply. Titles that previously sold for $40 were selling for 99 cents, which resulted in many companies going out of business. Others would survive, but would never quite recover from the hit.

It took a while for the industry to recover from such a bad experience. Luckily, by 1985, some companies introduced a new wave of products and a revised quality policy. Nintendo introduced the Nintendo Entertainment System (NES) (Famicom in Japan), backed by a roster of internally developed, high-quality games. Shortly thereafter, some of the Atari developers such as Namco became Nintendo licensees. By 1988, classics such as Super Mario Bros and Legend of Zelda had been introduced, making Nintendo the top seller. Sega did not stand still either, introducing the Master System in 1986. Internally, both the NES and Master System were equipped with 8-bit processors (a 6502 and a Zilog Z80, respectively). Spec by spec, the Master System was a slightly better platform: More sprites were supported, and RAM size was increased. But Nintendo had the advantage of being the first mover with an impressive game catalog and reputation.

During these years, the consolidated console business model finally became clear. The console manufacturer created the hardware and produced (internally or through contractors) a first batch of titles, which showcased the hardware's capabilities and helped sell the first units. Then, developers wanting to jump on the bandwagon needed to apply to become licensees, and each game they built had to be tested for quality by the manufacturer. Only by enforcing this policy could consoles offer a consistent-quality game catalog, which made consumers happy with their purchases. Developers still received documentation and sometimes technical support from the console maker. They also had to pay a royalty fee to cover the console's development cost. In the case of Nintendo, there has always been a very strong presence of Nintendo-built games for its game systems, which ensures that a core of high-quality games is always available. By 1988, Nintendo was an established market leader with a well laid out licensing policy.

Let's press the pause button and explore the internals of the flagship product for this era, the NES. To begin with, Nintendo's console was built on top of a customized 6502 MOS CPU, which was enhanced to perform audio waveform computation directly on the CPU. As with all 6502s, the NES could only address 64KB of memory, but some techniques were devised to overcome that limit. Cartridges could implement a paged approach using a memory mapper to increase the addressable memory.

Graphics chores were carried out by the picture processing unit (PPU), whose main feature was support for tile-based backgrounds, sprites, and scrolling on hardware. Contents of the screen were described using four main data structures: the pattern table, the name table, the attribute table, and the sprite table. The pattern table was a list holding the definition of each 8x8 tile or pattern. Each tile was defined by a sequence of eight 16-bit values, each one representing one row of the tile. The idea is pretty straightforward. Each pixel in the tile can have one of four possible color values, and thus two bits per pixel are required (hence the 16 bits per row). So one tile in the pattern table could represent four different color values.

The name table assigned tiles to screen coordinates. The name table was a two dimensional, 30 row by 32 column matrix, with each position exactly one byte in length (thus selecting from 256 possible tiles). Multiplying the name table dimensions by the tile size would produce the total size of the background: 256 by 240 pixels. This was more than the available screen resolution, so only a portion of the background was visible at any given time. This background could be scrolled by simply changing the value of two offset registers.

An attribute table was used to modify how tiles were to be mapped to the screen. Attributes were specified for each 32x32-pixel screen block (which means one block every 4x4 tiles), so all tiles in that area shared the same attributes. Attributes were used to further refine the color scheme to be used. The attribute table provided two high-order bits for the color, whose two low-order bits were taken from the pattern table. This way a maximum of 16 colors could be used.

The sprite table was used to overlay sprites characters, potions, and so on on top of the background. The PPU could store up to 64 sprites (8x8 or 8x16 pixels each), which could be quickly mapped to the screen. Sprites were prioritized; whichever sequence you decided to paint, lower-number sprites were painted after higher-order ones were processed, thus providing an easy way of layering information onscreen. The main difference between sprites and tiles was that sprites could be blended with the background in a realistic manner; a key value was provided to specify which portions of the sprite were actually transparent and should expose the background.

As these special structures illustrate, the NES was a dedicated game platform. Its structure was clearly limited to what it was supposed to do. This is a common characteristic of all game consoles: They excel at what they were designed to do, but aren't useful for much else due to the system's design decisions.

The evolution of personal computers did not stand still for a decade. The Apple ][ had marked the course for the industry. Computers that sold at a reasonable cost were useful as working tools, but also kept the door open for home computer gaming. Each new iteration offered more powerful hardware, easier to use programming environments, and a more compelling game experience. The introduction of the IBM PC in 1981, with its open hardware design, fostered innovations by specific peripheral manufacturers. Graphics cards and hard drives began to be manufactured by different companies; each one competing for a piece of the market share. The evolution was enormous: From the crude, 4-color CGA adapter to the 256-color, 320x200 VGA adapter, only five years passed. VGA was the graphics mode that made computer games so popular in the late 1980s. It offered full-screen titles with a resolution similar to game consoles and color depth to display rich game worlds. Masterworks such as Day of the Tentacle by LucasArts or the Alone in the Dark saga are classic representatives of the VGA days.

Core Techniques and Algorithms in Game Programming2003
Core Techniques and Algorithms in Game Programming2003
Year: 2004
Pages: 261 © 2008-2017.
If you may any questions please contact us: