From printouts to pictures you can change
For most of the 1940s, a computer’s output was numbers on paper or lights on a panel. The change that made modern graphics possible was the arrival of a screen that showed a computer’s results as a picture, followed by devices that let a person respond to that picture. Together these created interactive graphics: a loop in which the user looks at an image, points at something, and the machine changes the image in response.
This article covers that early loop in the 1950s and 1960s. The longer technical story of how pictures are drawn is told in the history of computer graphics, and the film side is covered elsewhere. Here the question is how people first got to draw and manipulate images, and who contributed. Credit is shared among several projects, so we avoid naming a single first.
Putting a screen on a computer
Early computers sometimes used a cathode-ray tube, the same type of tube as in an oscilloscope or early television, to show the contents of memory. In the late 1940s and 1950s, some machines used it as a display and not just as a memory device. The Whirlwind computer at the Massachusetts Institute of Technology, built for the US Navy and developed into the early 1950s, had a screen that could show text and simple graphics. Its design influenced a number of later systems. The role of that kind of tube is described under CRT displays.
These displays were generally vector displays. Instead of filling in a grid of dots, the electron beam was steered to trace lines directly on the screen, like drawing with a pen. Drawing lines this way was fast, but the amount of detail depended on how often the computer could refresh the picture before it faded.
SAGE and the light pen
A major step came from air defense. In the 1950s the United States built SAGE, a network of large computers that tracked aircraft. Operators sat in front of display screens that showed radar data as symbols and lines, and used a handheld light gun, a pointing device related to the light pen, to select a target on the screen. It was among the first systems in which many people routinely used a computer by looking at a display and pointing at it, rather than by submitting punched cards.
The light pen detects the moment a point of light passes beneath it on a screen. Because the computer knows what it was drawing at that moment, it can tell where the pen is pointing. This approach was used in various research and commercial systems in the 1960s. It is one of the ancestors of today’s pen input, though the technology is quite different.
Sketchpad and its place in the story
In 1963, Ivan Sutherland submitted a doctoral thesis at MIT describing Sketchpad, a program running on the TX-2 computer at Lincoln Laboratory. The user drew with a light pen, with lines and circles that the program stored as objects. Constraints could be applied, so that, for example, lines remained parallel or joined at a point when other parts of the drawing were changed. Shapes could be copied and reused.
Sketchpad is widely and fairly described as a landmark. It is often credited as an influence on later work in graphical interfaces and computer-aided design. It was not the only project of its time, however. Researchers elsewhere were working on related ideas, and Sketchpad built on the hardware and the culture of Lincoln Laboratory and MIT, where earlier display work had already been done.
Industry projects: design and manufacturing
At about the same time, industry looked at graphics for engineering. General Motors worked with IBM on a system known as DAC-1, a design system for automobile parts that was shown in the mid-1960s. Other companies, including aircraft manufacturers, experimented with graphic terminals for drawing parts and checking designs. The term “computer graphics” is often credited to William Fetter, a designer at Boeing who studied how a pilot’s cockpit view could be modeled with a computer, although other uses of the term may have appeared nearby.
These industrial uses are why the topic overlaps with computer-aided design, where interactive drafting became a goal of its own. Drawing with a computer, in this period, was typically an engineering tool rather than an art form.
Other devices for pointing and drawing
Alongside light pens, there were graphics tablets, on which a stylus could be moved to trace out lines. The RAND Corporation developed a tablet in the mid-1960s that gained attention. Meanwhile, at the Stanford Research Institute, Douglas Engelbart’s group developed the computer mouse and demonstrated it, along with other interface ideas, in the late 1960s. Another thread is the early work in the 1960s and 1970s on the graphical user interface, which later came to rely on pointing devices. The question of who contributed which part of that development is complicated, and it is covered in its own articles.
Limits of the early systems
The machines of the time had serious constraints:
- Cost. A computer able to run a display cost very large sums, and graphics terminals were expensive on top of that. Only government agencies, universities and large firms could afford them.
- Memory. Storing even a simple image required memory that was scarce and expensive, which pushed designers toward vector displays that stored lists of lines rather than grids of dots.
- Speed. Complex drawings had to be redrawn continuously, and too many lines caused flicker.
- Scale. Most systems were individual projects, and software written for one display often could not run on another.
Despite these limits, the ideas of the era were durable: an image can be an object that a program understands, and a person can edit it by pointing.