Ins and outs of X Window
Jim Flatten
Electronic Engineering Times
October 30, 1989
Every few years, there's a hot new technology touted as revolutionary in the computer graphics industry. Some innovations are quickly eclipsed by newer technologies, while others make a deep and lasting impression. For a variety of reasons, the X Window System will have a significant impact on computer graphics users.
The X Window System is a base window system providing high-performance graphics and window-management mechanisms for a hierarchy of resizable, easily moved windows. Development of the X standard grew out of the Massachusetts Institute of Technology's Project Athena, an experiment in the educational uses of high-power networked personal workstations.
The X Window System provides a common windowing system across different vendors' workstations, as well as providing network-transparent graphics in a standard fashion. While some vendors have offered proprietary methods for doing network-transparent graphics, X is the first to be widely accepted in the industry.
The advent of X means that workstations are gradually becoming interchange
able pieces in a larger networked computer system. If all the workstations have an X 11 interface, then an engineer could use them interchangeably.
The importance of a standard windowing interface available across competing workstations is particularly evident in scientific visualization. A user with an X-compatible workstation could be running a complex application, such as a finite element analysis, remotely on a Cray. The user could then use any X workstation to visualize the result.
As a result, supercomputer time can be used more efficiently and by a larger number of individuals. While those who operate their workstations primarily as terminals do not need a large number of Mips on the desktop, the memory-intensive nature of graphics applications means that those who use the visualization capabilities of their own workstations will always find a use for extra power.
The impact of X on graphics cannot be viewed in isolation. During the last three to four years, a number of graphics standards have been established, including several that had been under development since the late 1970s.
The new standards include: X 11, the X Window System standard; Graphical Kernel System (GKS), the two-dimensional graphics standard; Programmer 5 Hierarchical Interactive Graphics System (PHIGS), a three-dimensional standard; PHIGS+, an expansion to PHIGS providing standards for three-dimensional shading and solid modeling; and PHIGS extension to X (PEX), which provides three-dimensional extensions to X's two-dimensional windows.
Now that graphics standards and windowing standards are both in place, it is possible to work on integrating graphics applications across multiple hardware platforms. In the X environment, it is possible to view a CAD image in one window and text in another.
The Compound Document Architecture is Digital's standard for data integration. When used with the X Window System, it provides a common backbone between applications running in different windows, allowing users to take the CAD image and insert it into a document in another window.
The next integration challenge to tackle is imaging. Many organizations have archives of images that they would like to make available on-line. For example, an aircraft manufacturer may have a library of blueprints that are modified as the planes are updated. Rather than redrawing each blueprint with a CAD system to enter it into the computer, users want to read the images into the system with a scanner. Using PHIGS, an individual would be able to extract the geometry of the drawing, modify it or generate a new hard copy.
Imaging encompasses scanning a drawing into a system, as well as rotating it, enhancing the contrast and modifying the size. To do that, there must be very good integration of imaging and graphics, which does not exist right now. The X Window System is already a convenient way for users to integrate multiple nongraphics applications, so providing imaging capabilities will more fully support graphics integration. X users are going to want to use X as a basis for all kinds of integration.
As a result, PEX and imaging extensions to X are both key areas for continuing development.
To a graphics system like GKS or PHIGS, the windows in X appear two-dimensional. A user drawing in an X window on a system such as a VAKstation 3520 is unable to take advantage of the three-dimensional hardware features of his workstation. PEX extends the two-dimensional X window and makes it appear three-dimensional to PHIGS. Using PEX, the user will get better performance because of access to all the features of the three-dimensional hardware.
Right now, standards for integrating text, geometric graphics and images do not exist. How we integrate imaging and geometric graphics so that they work smoothly and seamlessly together is a challenge for future development.
From the point of view of graphics systems such as GKS and PHIGS, X is a device-dependent protocol that applies only to workstations. Although workstations are extremely popular at the moment, there are still many terminals around. X will be important for those users interested in distributed computing, but developers of graphics applications are encouraged to write their programs to be compatible with graphics standards rather than a windowing standard. In that way, their programs will be easier to update to work with future innovations.
Specialists and generalists
The need for integration between specialized applications and general office systems has become clear. We realize that even the users of the most specialized applications need to communicate some of their information to others, and they must also interact with their fellow workers in an office environment.
Windowing systems effectively integrate graphics with electronic mail, word processing and other office applications. As the first independent windowing system designed to make graphics applications transparent across a network, X is more than the technology of the month. X provides a setting where users can realize the dramatic benefits of the recently established graphics standards.
Jim Flatten, product manager of Digital Equipment Corp. PHIGS and GKS, was a member of the ANSI X3H3 graphics standards committee for nearly four years. He has worked with Digital Equipment hardware and software both as a user and as a Digital Equipment employee.
Copyright 1989 CMP Publications, Inc. All rights reserved.