The Desktop-Publishing Phenomenon

Personal printing has come a long way in a short time

John W. Seybold
BYTE

May 1, 1987

EVERYBODY UNDERSTANDS WHAT you mean when you say ''desktop publishing,'' even those who have never used a computer. As a concept, it's compelling. As a product offering, it's the best thing that ever happened to personal computers. As a fillip for a market that--at least temporarily--had lost its momentum, it has been a godsend. As a harbinger of personal-productivity tools, it has been and will continue to be a prime example. Yet, in a sense, desktop publishing is a slippery product without a clear-cut definition.

The term is not old, and its origins are readily traceable. Paul Brainerd of Aldus, father of PageMaker, gets credit for coining the phrase. Apple Computer, looking for a vehicle to dramatize the Macintosh's capabilities, had the perspicacity to go all out on a promotional campaign. Without the Apple LaserWriter, desktop publishing would never have existed--desktop publishing didn't really make its debut until Apple announced the LaserWriter in January 1985. That was a short time ago, and a lot has happened since.

But the concept didn't just come out of the blue; an interesting history lies behind it. It originated in 1973, when an experimental multifunction workstation was developed at the Xerox Palo Alto Research Center (PARC). That workstation, the Alto, was the most important unannounced computer product of the 1970s. It was conceived as a completely self-contained, single-user computer, although its principal mission was to communicate with other similar devices over the original 3-megabit Ethernet local network. Because its designers had graphics applications in mind, the Alto had a high-resolution bit-mapped display screen and a mouse-pointing device.

No one intended the Alto to be a widely used system. Each run of Altos was supposed to be the last, but by 1981 over 1200 had been installed on 46 Ethernet networks in seven geographic areas. Outside of Xerox, Altos were assigned to a few selected sites, including the White House, the House of Representatives, the Senate, a few universities (especially Stanford), and several undisclosed locations in the U.S. and Europe. The large base of users provided input as to what was right and what was wrong with the Alto and the various software packages that had been developed for it. These included Bravo, Gypsy, Markup, Draw, SIL, and Laurel--for text editing and formatting, creating pictures and diagrams, and providing electronic mail by means of a central file server.

Even more important, PARC alumni disseminated their fascination with this product, and the ideas it generated, throughout the computer industry--especially on the West Coast--wherever they took on new roles as key members of advanced development teams.

From the Alto came the Star, developed under the aegis of Dave Liddle, vice president of office systems within Xerox's Office Products Division, and announced on April 27, 1981. The Star perpetuated four key concepts. The first was ''seeing and pointing'' rather than ''remembering and typing.'' ''Progressive disclosure'' reduced the apparent complexity of the system by presenting you with only the information relevant at that moment. ''Uniformity of commands across domains'' made system procedures consistent. Once you understood the basic logic of the machine, you could deduce what to do next without recourse to manuals. Last was WYSIWYG (what you see is what you get). The display on the screen should always be a close facsimile of the final printed docu- ment.

The Descending Star

The Star never reached any kind of ascendancy. It was poorly marketed, was too expensive, and lacked a reasonable output device. The 8044 Print Server, offered as an Ethernet peripheral, cost $30,000.

But Steve Jobs had visited PARC and saw it all happen. He enlisted the support of some of the Star's major developers and brought out the Lisa, a $10,000 desktop unit with a high-resolution bit-mapped display screen, a Motorola 68000 processor with a million bytes of memory, two high-capacity floppy disk drives, a 5-megabyte hard disk drive, a mouse, a user interface inspired by the Xerox Smalltalk system, and an impressively complete set of office application programs (LisaCalc, LisaGraph, LisaDraw, LisaProject, LisaList, and LisaWrite).

But the Lisa didn't have anything to write with that wasn't anticlimactic, even though its dot-matrix printer (with a standard resolution of 96 by 72 dots per inch) could move into a double-write mode of 160 by 144 dpi. The Lisa, too, was overpriced, ahead of its time, and not quite all there.

Still, the Lisa perpetuated the desktop metaphor of the Star and Alto. This metaphor is the origin of the ''desktop'' in desktop publishing.

The concept used almost all the screen to portray a desktop environment on which icons represented various familiar objects, images that you could point to and move about to make things happen. The Seybold Report on Publishing Systems, in describing this product in its January 31, 1983, issue, included a footnote that argued that, while some believe that the physical environment of an office needs significant change and shouldn't be emulated and ''while we recognize many of the failings of the existing office, we also object to the arcane user interfaces typical of most computer systems . . . There can be no doubt that Lisa is different from many systems, and we are convinced that it is better than many systems because it uses icons.''

Still, there was no desktop-publishing phenomenon. Nor did the phenomenon begin when Apple moved away from the Lisa and introduced the Macintosh in January 1984. It began in January 1985, when the LaserWriter made its debut. This $7000 plain-paper ''typesetter'' incorporated Adobe's RIP (raster-image processor) and Adobe's professionally designed fonts licensed from the International Type Face Corporation. Moreover, the event received the blessing of the most famous professional typesetting manufacturer: Allied Linotype (formerly Mergenthaler) threw tradition to the winds and embraced the technology, offering an interface, via Adobe's PostScript, to drive its Linotronic 101 and Linotronic 300 typesetters to produce truly professional output quality.

Aldus president Paul Brainerd had been in on the ground floor and played a role in persuading Apple that a typesetting application would be popular. The word ''typesetting'' is used in this context only because desktop publishing hadn't yet been invented, although with Aldus's PageMaker, and the hype that surrounded it, the term quickly emerged and conquered.

What Is Publishing?

As you can see from its history, the ''desktop'' part of desktop publishing was a product of the Xerox desktop metaphor, implying icons, bit-mapped screens, and the other features described above. The ''publishing'' part suggests that the LaserWriter would be more than a typesetter, since it could also serve as a final output device.

A typesetter produces only camera-ready copy from which printing plates are made, but you could use the LaserWriter to print multiple copies for distribution. This isn't very practical, however, because you'd want to print on both sides of the sheet (which is awkward), and the duty cycle of the LaserWriter won't support extensive emulation of a printing press.

''Publishing'' also implies going beyond typesetting to produce graphics. While typesetting devices (now of the raster-image variety) have become increasingly capable of producing illustrative matter, this (with a few exceptions) was and is not commonly the way that graphics are reproduced. Generally, the platemaker or the camera operator combines graphics with text by stripping film onto the indicated page positions or double-burning illustrations onto printing plates that already contain the text.

Finally, ''publishing'' implies distribution to a readership. In its broadest sense, the word means to make known, presumably in a general way, to differentiate the activity from mere communications. Traditionally, publishing has implied the use of printing presses and paper, although today many publishers have ventured into nontraditional activities that make use of alternative media.

Thus, just as ''desktop'' is either too specific or too vague a term to provide a definitive description of the activity in question, so ''publishing'' is probably too ambitious or presumptuous. But the fact remains that the new technology has spawned a whole host of desktop publishers who are publishing--writing, producing, and distributing--to a readership, and the publishing process has suddenly become democratized and affordable beyond the dreams of only a few years ago.

Not Just the Mac

So far, I have focused on Apple's Macintosh. This has been intentional because, without the Mac, I doubt if desktop publishing as we know it today would ever have happened. But nowadays desktop publishing isn't limited to the Mac, and it may not be principally pursued on the Mac in the future. However, desktop publishing does not necessarily include any and all efforts to apply personal computers to text processing (or even typesetting). No product deserves inclusion in the category unless it meets all or, at least, most of the following criteria (although many people might be inclined to argue with me).

First, the system should let you compose text in a manner that comes close to the requirements of typesetting--that is, well-designed and proportionally spaced characters--along with the ability to produce justified lines if desired, including automatic hyphenation. The latter is what the graphic arts industry calls h&j (hyphenation and justification). The process of composition also generally implies other typographical considerations, including the use of more than one type font or size, and a character repertoire more extensive than that available on a typewriter. It also embraces the setting of tabular matter, indents of all kinds, and perhaps aesthetic character kerning.

Second, the system should let you perform the composition tasks I've alluded to in a manner that is considerably less code-intensive than that which characterizes traditional computer typesetting. This strongly supports the WYSIWYG bit-mapped screen and an interface that lets you point to the effects desired rather than requiring that you describe them by some sort of command language. Command codes might inevitably be embedded in the text stream or in some sort of pointer file, but in either case they are transparent to the user.

Already I have excluded from my definition a good many typesetting programs written to perform professional typesetting, or which emulated them. These were code-intensive and required considerable user experience in markup and composition programming. There are a good many programs of this sort. In fact, they antedate desktop publishing. Some of them are powerful and go well beyond the capabilities of desktop publishing, but they are something else again.

Third, the system should let you include graphics and output illustrative matter along with the text. This implies, for example, the existence of some sort of pagination or area composition program--not just what the industry calls ''galley composition.'' This feature also implies that you can create graphics on your system, import graphics into the system, or both. Graphics features created within the system do not necessarily need to be produced by the same software that handles composition, but the relationship between the handling of text and graphics must be as seamless as possible.

Illustrations can be generated by business graphics programs that pull figures from a spreadsheet and convert them into bar or pie charts. They can be created by painting, drawing, or CAD programs, or be scanned in from a peripheral device such as a scanner. They can even be captured by means of a video camera. Or perhaps the pictures can be moved over from a digitized clip-art collection. The pictures need not be derived from photographs (continuous-tone images) and screened to become halftones for purposes of reproduction, but they might well be. The system should also provide some means, within itself, to perform a variety of manipulative tasks, such as sizing, cropping, image rotation, or enhancement.

Fourth, the system should offer some reasonably good editing capabilities. Ideally, you should be able to use the same kinds of editing tools offered by the best word processors and to make changes interactively, not merely to uncomposed text, but to text in the process of interactive composition. In other words, you want a system sophisticated enough to let you look at a composed page on the screen, identify a paragraph that needs rewriting, and make that change then and there, without reverting to some sort of input file. Whatever the system's capabilities are, desktop publishing clearly demands an effective editor as part of the package.

Finally, the system must be able to produce output on at least one device that offers a resolution of 200 or more (I would prefer to say 300 or more) dpi. Without this, you can't handle graphics, except a few bar charts and the like, and the quality of type doesn't deserve to be considered publishable. Beyond that, the system should interface to (or contain) one or more RIPs that work with PDLs (page-description languages) capable of producing a variety of outputs, including output to a high-resolution typesetter. This is the beauty of PostScript, Interpress, and Document Description Language as page-description languages. And this is what RIPs are for--the hardware and software necessary to interpret and translate that language into commands to drive the desired device.

Not all so-called desktop-publishing systems meet all the criteria outlined above. Some of them fall quite short. Even PageMaker was unable to perform automatic hyphenation until recently. Many program developers have never heard of kerning. Some programs use type-width values that are simply too gross to produce good-quality output. I believe that type-width increments as tiny as 1/216 inch are essential, and I would prefer much finer increments. This figure is based on the reasoning that an increment should not be less than 1/18 of a printer's em (a unit of measure based on a piece of type that is as wide as it is tall), and an em needs to be at least as small as six points (1/72 inch). Putting It All Together A happy coincidence of events has been engendered or accelerated by powerful, low-cost computer chips: the advent of scanning and graphics packages, for example, and the ability to store a hyphenation dictionary in memory along with tables of width values and typesetting macros.

Another important technology was pioneered by Bitstream. It was able to transfer the art of type design into a computer science, and develop the contacts and initiative to get the rights to a library of classical type fonts and transform them into a database. The authority and persuasiveness of true type makes a significant difference between publishing and mere communications.

For many users, desktop publishing will end with laser-printer output. For others, such 300 dpi output, while a significant improvement over letter-quality daisy wheels, is the first step. The laser printer often furnishes merely the proofing copy, and then the output can be redirected to a true typesetting system, either convenient to the desktop of the user or by arrangement with a neighborhood trade typesetter.

Again I come back to the desktop. The implication is clear that the devices must be inexpensive, small, easily installed without the need for professional assistance, and easily maintained. One person should be able to easily take control of the entire process. He or she can learn and practice all phases of it, more or less intuitively, keep track of all the components within the system, and function effectively in a creative and a production environment to produce publication-worthy products.

Publishing Platforms

Although the Macintosh and the LaserWriter started it all, others soon got into the desktop-publishing act. The IBM PC was, in some respects, an ideal platform for the development of a desktop-publishing system because it is ubiquitous, powerful, and open. In other respects, the PC has been awkward and frustrating. Graphics capability does not come easily to it, and no intuitive and seamless way exists for the user to flow copy and graphics into and through the system.

When applied to the Macintosh, the term ''platform'' is easy to understand. It consists of a certain amount of hardware and a clearly understood set of operating conventions sufficient (with the aid of an RIP in the output device) to provide a platform for the development of application software development.

But the PC has no comparable platform. You must specify what kind of graphics board is available, and, to get an appropriate environment, you need Windows under MS-DOS or GEM under its own set of conventions.

Nevertheless, in describing available systems, you will often hear of a PC platform as if it did exist. I would caution you to make sure that some degree of seamlessness is possible with the aid of the particular planks used to build the platform you choose.

In a sense, the IBM platform, despite its ambiguities, is more congenial for the development of future desktop-publishing applications than the Macintosh, simply because it's more ''naked'' and can be clothed or ornamented more easily by clever designers. Some of the desktop systems for the PC platform are so powerful that it becomes impossible to draw a clear-cut distinction between desktop and professional publishing.

To draw such a distinction, once you are satisfied that all the typographical niceties are provided for, you have to look closely to determine whether you can ''grow'' a given desktop system for simultaneous use, in some sort of sharing and networking mode, with a fairly large number of users.

In my opinion, the primary difference between a professional front-end system and a desktop-publishing system lies not so much in the system's composition and graphics capabilities as in supporting a fairly large number of users. In other words, the file-handling capabilities of, say, a newspaper or magazine editorial system must be powerful and sophisticated. I haven't seen anything in the low-end, desktop market that offers the queue and directory structures essential to the copy-flow requirements of a major publishing venture. These will no doubt come in time, and when they do, the distinction between desktop and other publishing systems will probably fade away.

This is not to suggest that all systems will offer the same capabilities. Significant differences exist between desktop offerings, not only in terms of typographical niceties, the ways in which illustrative matter is handled, and throughput speeds, but more specifically in terms of the type of work the system can best handle.

For example, some systems are built around yet a third platform: workstations that support UNIX, which are patterned after Donald Knuth's TEX and which handle multilevel mathematics, scientific notation, and book pagination in an environment that is primarily batch-oriented, rather than interactive in nature.

Some systems, built on either a Mac or PC platform, are superb at area composition (making up blocks of copy--like newspaper display ads) but can't cope with running the pages of text necessary for book composition. Still other systems intended primarily for short pieces--brochures and newsletters--try to offer something a bit more appealing than word-processing output. However, they don't come to grips with what makes for good-quality composition, such as control over interword spacing; letter spacing; frequency of hyphenation; handling of hangers, widows, and orphans (lines or portions of lines that stand by themselves when they should join to larger text blocks); and all of the other features that contribute to the artistry of composition.

Opportunities at Hand

Desktop publishing has come a long way in a short period of time. Now on the desktop are tools that take advantage of the knowledge and experience of generations of artisans and designers. Without serving a long term of apprenticeship in a specialty such as photoengraving or hand or machine composition, you can quickly and pleasurably become a desktop publisher.

-- John W. Seybold (P.O. Box 644, Media, PA 19063) has been a pioneer in the field of computer-aided typesetting and publishing since 1963. He is the founder of The Seybold Report on Publishing Systems and author of The World of Digital Typesetting and Publishing from the Desktop.