(FORTUNE Magazine) – A commercial real estate broker in Boston pops a compact disk into a player. She taps a few keys on her computer terminal. The floor plan and pictures of an office tower in Dallas pop up on her client's screen a continent away in Los Angeles. The broker's man in Dallas comes on line to take the client on a video ''tour'' of the building. With a few more commands, the computer produces a spreadsheet outlining the building's costs and revenues. The buyer taps in an offer for the building and the numbers appear on the broker's screen. Broker and buyer agree. Deal closed.

FROM THE FERTILE, science fiction-soaked imaginations of computer software engineers is springing a dazzling new vision of a futuristic office environment where the 21st-century TV cartoon character George Jetson might feel at home. Software that will enable people to collaborate across barriers of space and time is one of today's hottest frontiers of computer research. Like an electronic sinew that binds teams together, the new ''groupware'' aims to place the computer squarely in the middle of communications among managers, technicians, and anyone else who interacts in groups, revolutionizing the way they work. Groupware boosters make breathtaking, if extravagant-sounding, claims for what this new technology will do. Linked desktop terminals running the new software will coordinate schedules and route messages. Novel products will emerge as networks of computer work stations guide teams of workers through large shared databases; a pharmaceutical company, for example, might search a database of organic chemicals for possible new drugs. Managers will confer with colleagues, suppliers, and customers via wall-size video screens as cameras connected to computers record and store their conversations. And -- hold on to your space helmets -- even meetings will become more effective as today's low-tech conference rooms turn into multimedia ''war rooms'' controlled by software that helps keep everything on course. Software that supports group work may not be as far out as it sounds. Advanced prototypes are already in use at a handful of research labs around the country; the first commercial products are beginning to reach the market. Says Jerry Wagner, a professor of management information systems at the University of Texas Business School in Austin: ''This technology could be one of the most important contributions to management effectiveness in business history.'' In the new effort to develop software that helps teams of people work smarter, necessity is running in tandem with opportunity. Software engineers need better ways to manage teams of programmers writing vastly more complicated computer codes; as programs grow more intricate, the risk increases that the work of engineers collaborating on one of dozens of subsystems will distort the work of other teams. At the same time, companies are looking for opportunities to transform their huge investment in office automation equipment into tools that actually, finally, pay off in productivity gains. ENGINEERS at the Microelectronics & Computer Technology Corp. (MCC) research consortium in Austin, Xerox's Palo Alto Research Center (PARC), and Tektronix Laboratories in Beaverton, Oregon, have made groupware a priority for their research. ''We've reached the limits of what individuals can do,'' says Laszlo Belady, director of MCC's software research program. ''The biggest problem we face isn't so much a technical problem as a human coordination problem.'' Belady is quick to add that solving the problem casts software designers in a new role: ''As hard scientists, we've found it uncomfortable dealing with the messy, ambiguous relationships that exist among people in groups.'' Says Benn Konsynski, a professor of management information systems at Harvard Business School: ''Coordinating people is the business problem managers want to solve.'' Thomas Malone, who teaches the same subject at MIT's Sloan School, thinks companies will benefit most from software that helps quicken their ''information metabolism,'' the speed with which they digest and respond to information. A team Malone heads has been working for over two years on a program called Information Lens, with funding from GM, Wang, and Xerox. Information Lens blends into the office like a supercompetent servant with a team of scattered masters. It draws on an artificial intelligence programming tool that allows each user to tailor the system to his personal needs. The computer can be instructed to discriminate among electronic messages -- recognizing the difference, say, between a routine announcement of a new pencil procurement policy and an urgent reminder that your boss wants your group's report before the end of the day. And it can cull information from other computers' databases that relates to the projects a group is working on, routing it speedily to each group member's terminal and automatically updating his files. Malone thinks commercial versions of the system designed to work on personal computers could be ready in two to three years. Transforming groupware from an engineers' plaything into user-friendly programs for business poses formidable problems. ''The pace of acceptance of these new tools will be determined by the abilities of the most technophobic member of a work group to use it,'' says George P. Huber, a professor at the University of Texas Business School. Commercial software developers, having perfected word processing and financial spreadsheet programs to a market- saturating fare-thee-well, have a powerful incentive to try to solve the problem. Among the biggest new markets for business applications software are the roughly 280,000 networks that companies have bought to link their computers together but so far have discovered few ways to use. Microsoft and Lotus Development are making software that supports group work a major focus of their research, but both estimate that they are about three years away from launching products. Pride of place for bringing the first easy-to-use programs to market belongs to two California companies, Action Technologies and Broderbund. At its start- up in January 1985, Action Technologies introduced a product called the Coordinator that is based on a stunningly simple theory of human communication. It is the brainchild of a cerebral Chilean emigre named Fernando Flores, who served as finance minister in the government of Salvador Allende (see box). Business, according to Flores, is essentially the completion of a set of action-oriented conversations such as requests, offers, counteroffers, and promises. The trick to effective team coordination, he says, is having work groups label their conversations accordingly. The Coordinator lets a user compose a message on a personal computer, assign it to one of Flores's conversation categories, and transmit it via standard telephone lines to other users. The system tracks each user's conversations, reminds him of his pending commitments, and keeps a record of the status of a group project. Gtel, a California-based seller of telephone equipment, uses the Coordinator to track inventory at its 20 retail outlets and four warehouses, assemble daily store management reports, and communicate between headquarters and the field. Denise Schubert, Gtel's personal computer manager, estimates that the new system yields better data and costs $4,500 a month less than the electronic-mail system it replaced. But Schubert is most impressed with the fact that Gtel's store clerks, most of them aged 17 to 24, were able to start using the system after just a day's training. The Coordinator is winning converts among managers at General Motors' EDS subsidiary, whose account managers are using the system at GM offices in Dayton. It is also being tested at Aetna Life, where Bonnie Johnson, a corporate technical planning manager, says, ''This is the first effective group communications tool I've seen.'' Broderbund introduced a program called For Comment last November. It enables a team of up to 16 members to collaborate on writing, reviewing, and editing documents that can contain up to 236 single-spaced pages. Users relay a common document they are working on -- a business plan, for example -- through their network of personal computers, suggesting revisions and adding up to five levels of comments and comments on comments. The software tracks and saves each draft revision as well as all the commentators' contributions, creating a record of the ideas that influenced the group decision process. To prepare a finished document, the author simply moves suggested changes from the comment block into the main text almost as easily as a crazed teenager in a video game arcade zaps a space invader. ''Our background developing computer games gives us a big advantage in understanding how to make software easy to use,'' says Joanne Bealy, For Comment's product manager. In the research labs, software engineers are taking two approaches to create programs that will transform networks of computers into more advanced collaboration tools. The first, called hypertext, allows groups working in the same building or at a distance to use their linked computer screens as a shared work space, something like a common desk. The second, using electronic meeting rooms, harnesses the computer to increase the amount of information that can be brought to bear in face-to-face meetings. The computer is at the heart of both approaches, capturing the communication among team members, tracking and updating the information exchanged, and rendering it back precisely to all of its users. The term hypertext was coined in the 1960s by Ted Nelson, an exuberant Stanford University computer scientist who had the extraordinary idea of creating a computer-linked global information source he called Xanadu, after the mythic palace in Coleridge's ''Kubla Khan.'' Today researchers at Xerox PARC, MCC, and Tektronix are plucking hypertext from the realm of fantasy and trying to shape it into a focused research tool. Working with networks of powerful Xerox and Sun work stations equipped with large display screens that can be divided into windows, hypertext programmers segment data into interconnected idea-size chunks called nodes. HYPERTEXT PULLS together data from disparate sources and prompts the user to check information that may bear on what he's doing. In a simple commercial application, a toy company's market researcher checking a node called ''Frisbee sales'' might confirm that adolescent males are the toy's biggest buyers. Searching another node called ''population trends -- Minneapolis,'' he might discover that the city's male population aged 12 to 18 is decreasing far faster than the national average. By linking the two nodes, the researcher - could conclude that his company should reduce its Frisbee distribution to the Twin Cities. Individual hypertext users can navigate freely among nodes, following established links. As they roam, they can add new nodes and link them to the existing network, enriching the foundation of associated ideas and creating a sort of nonlinear electronic encyclopedia. More limited systems that allow users to browse through data without adding new information or creating new links are beginning to find commercial applications. This summer Ford plans to automate its repair manuals, using a browsing system called Guide to help mechanics service its 1988 models. Sold by Owl International, a year-old company based in Bellevue, Washington, the software will run on Hewlett- Packard computers connected to touch screens at mechanics' work stations. Instead of having to leave the repair bay to leaf through volumes of grease- stained manuals, a mechanic using Guide will simply touch symbols on the computer screen to search through servicing instructions stored on compact disks in the computer's memory. Software developers are also aiming to crack business's last computer-free inner sanctum, the conference room. For all the time managers spend in meetings, the process remains more a tribal rite than a structured, scientific process. After much talking and little listening, decisions too soon forgotten emerge by a process that no participant can describe. Armed with powerful work stations, researchers are mounting an assault on the meeting room and trying to integrate meetings with the work that individuals do independently. ''We use computers in every other aspect of our work,'' says Mark Stefik, a lean, intense engineer who designed an electronic meeting room called Colab at Xerox PARC. ''Yet each time we came to meetings, we had to leave them behind in our offices and resort to flip charts.'' The Colab room in Palo Alto accommodates eight participants seated around a U-shaped table at computers linked by cables. Facing the table at the front of the room stands a glowing 4-foot-by-8-foot projection screen that serves as an electronic chalkboard. Unconventional as the room looks, what goes on inside it is positively bizarre. A meeting moves through three stages, beginning with ''idea generation,'' proceeding to ''idea structuring,'' and ending with evaluation. In the generation stage, participants hammer away at keyboards, creating a frenzied staccato as they simultaneously enter their thoughts into computer memory. As new ideas are entered, they pop up automatically on other participants' screens and on the electronic chalkboard. Moving to the structuring stage, participants abandon their keyboards and shift ideas into related groups by using a mouse to shove a block of text across the screen and click it into place. Once the ideas are grouped to everyone's satisfaction and the clicking subsides, the group settles down to evaluate its work, supplementing the sorted thoughts with additional information drawn from members' personal databases. Colab's developers think all this chatterless clatter brings far greater cohesiveness to meetings. The computer records each participant's comments and displays them for all to see as the meeting progresses, a process Stefik describes as ''what-you-see-is- what-I-see,'' or WYSIWIS (pronounced ''whizzy-whiz''). The problems are equally obvious. Speed, unshakable concentration, and acute manual dexterity help a lot in getting one's point across. LIKE LATTER-DAY Dr. Jekylls, the researchers at Xerox PARC and MCC, which is building its own electronic meeting room, are studying computer-coordinated meetings by experimenting on themselves. At the University of Arizona's College of Business and Public Administration, researchers headed by Jay Nunamaker are investigating how ordinary executives adjust to this new idea. Arizona built a Decision and Planning Laboratory in 1985 and tested it on over 200 planning executives from such organizations as IBM and the U.S. Army. The lab lets users enter their comments anonymously. Managers like the candor the process allows, and the test groups have been uniformly enthusiastic. The work groupers' ultimate strategy is to marry the computer's memory with video and audio equipment. A small team of researchers headed by Robert Stults is exploring that far-out edge of technology at Xerox PARC in an experiment called Media Spaces. Last May, Stults and his team equipped a half dozen offices at Xerox's Palo Alto labs and four Xerox offices in Portland, Oregon, with video cameras, monitors, microphones, and speakers, and wired them together with recording gear -- all controlled by computers. They even wired the coffee lounges to capture all the creative chitchat that goes on during what employees used to think of as breaks. Media Spaces allows group members using these multimedia offices to see each other as they confer; the camera projects each user's image on television monitors in his colleagues' offices. Tapping into the computer's database, Media Spaces conferees can swap text and graphic data back and forth and exhibit them in windows on each other's screens. In time a new participant in the system will be able to command his computer to search a vast library of meetings recorded on video disks, zero in on a specific conversation, and play it back to reconstruct the rationale for a decision that the group made before he joined it. Stults thinks advanced versions of Media Spaces may emerge by the end of the century. ''Right now we may look like just a bunch of over-age kids playing at a funny farm in Palo Alto,'' says Stults. ''But when this technology is ready, it could take off fast.''