Path: utzoo!utgpu!jarvis.csri.toronto.edu!clyde.concordia.ca!uunet!snorkelwacker!tut.cis.ohio-state.edu!ucbvax!ENCORE.COM!bzs From: bzs@ENCORE.COM (Barry Shein) Newsgroups: comp.society.futures Subject: More ball gazing Message-ID: <9001032345.AA17793@encore.encore.com> Date: 3 Jan 90 23:45:45 GMT Sender: daemon@ucbvax.BERKELEY.EDU Organization: The Internet Lines: 364 Return-Path: Reply-To: Date: Wed, 3 Jan 90 16:55:24 EDT From: Bob_German@MAYTAG.CEO.DG.COM To: info-futures-request@encore.com Subject: More ball gazing I couldn't resist putting on my sage's hat and taking a crack at the coming decade. (_My_ dictionary, by the way, defines a decade as any period of ten years, so a new decade actually begins every year. I choose to make predictions for the one beginning in 1990 :-) .) Predictions for Computing in the '90s The last thirty years have seen such dramatic change in computing technology that it's tempting to extrapolate the advances of the past in making predictions for the future. Indeed, the sky's the limit on where our technology could go. In the '90s and beyond, however, technology will do less to shape the world of computing than will market and economic considerations (what people want and what they can afford). Workstations: The CPU chips used in workstations will continue to double in power, and memory and peripheral speeds will lag slightly but will eventually keep up. This will lead to the ability to put 100 MIPs machines sporting gigabyte memory sizes on the desktop. What will fall behind is applications to take advantage of these faster machines. Thus, the market's appetite for the latest, fastest workstation will diminish, and manufacturers will turn their focus away from high powered versions and toward reduced cost machines. Extremely powerful high end workstations will find their way to the desks of engineers and scientists, but are unlikely to penetrate commercial realms. Even as new applications become available, many non-technical people will prefer to stick with old, familiar applications. The high cost to retrain personnel and convert data for new applications will keep today's applications in the main stream long after the technology has passed them by. Parallel, often special purpose processors will become popular in mid to high end workstations for use as user interface accelerators, encryption devices and machine emulators. Machines will show up which can simultaneously emulate multiple "standard" workstations, such as IBM PC compatibles and Macintoshes. User interface shells will add "point and click" capability to old DOS and UNIX applications with text based user interfaces. The greatest strides in user interface design will come not on workstations, but in specialized devices such as hand held computers and pocket CD-ROM readers, where physical size will constrain both the display and input device. Pressure sensitive displays on notepad sized computers will encourage innovative new user interfaces. Voice will find little use in user interfaces except for telephone based interfaces and in specialized applications where a worker's hands are busy doing something else. Other voice based interfaces will have about the same level of success as the cars of the last decade which reminded us that "the door is a jar" and were generally written off as an annoying gimmick. In the commercial environment, diskless workstations attached to high powered file servers will become the prevailing trend. On machines equipped with mass storage devices, high speed removable optical disks will almost entirely replace magnetic media by the middle of the decade. Most software will be distributed on read-only optical media. Entire libraries of software will be available on CD ROM, along with improved help and tutorial facilities that will eliminate user manuals. Servers: Today's minicomputers, mainframes and supercomputers will slowly be replaced by sophisticated servers. These servers will be linked to workstations on a variety of local area networks. Twisted pair and fiber optic networks are likely to be the most popular. File and data base servers, communications servers, office automation servers, management servers and computing servers will all be available. These servers will fall into two camps: the proprietary, special purpose servers, and the standard, commodity servers. In the first category, high performance and low cost servers will perform generic functions such as file storage, communications and mail access. The standard, commodity servers will be platforms for custom server-based applications which will be written by large end users and value added resellers. A small niche will exist for very high performance programmable servers, which will replace today's supercomputers. Printers will generally come LAN equipped, and will not require servers at all (or, viewed another way, will be self-contained print servers). Although a typical workstation user may make use of a large variety of servers, their use will be hidden by data and applications management software which will attempt to present the image of a single, shared system to all users. Communications: Local area networks will continue to thrive. Prevailing technologies will be those which are cheap to install, which perform well, and which are most manageable. Users will be so dependent on their LANs that no technology will survive unless problems can be quickly diagnosed, and unless the network can continue to function in spite of a single failure. Large corporations and telecommunications vendors will continue to develop faster and faster media, probably based on fiber optics. Light switches will become available, but will be too expensive for all but telecommunications service providers. High bandwidth, fiber based networks will not reach the home in the 90s, however. In the middle of the decade, HDTV will finally emerge, but will be plagued for at least a few years by extremely high cost and a lack of programming. HDTV may eventually bring higher bandwidth cable into the home, and with it we may see interactive TV and two way, high speed home communications. This is unlikely to happen during the 90s, however, except for the very rich. ISDN will be offered for home use by the mid '90s, but it will cost more than the RBOCs are currently predicting. Information workers who telecommute may use it, but most of the population won't understand why they need anything more than their old, faithful telephone. The average home will have to make do with today's one-way cable and conventional telephone access. More sophisticated cable TV receivers are likely, however. It is likely that videotext systems running on one-way CATV based networks will reach the US in the '90s. Everyman's global communications network will remain an illusive dream during the 90s. Worldwide economic and environmental crises will lead to less global teleconferencing and more local, grass roots communication. Information technology will be of great use, however, in helping local groups to share ideas and to explore possible solutions to the problems of pollution, resource depletion and global warming. In general: In the 90s, the world will begin to reevaluate the meaning of "progress" as constant expansion of the economy and of endless cities, suburbs and shopping malls. Instead, progress will be seen as our ability to work together to achieve goals such as world peace, an end to hunger, and a clean environment. People will turn their minds to survival of the planet, and will apply our information technologies to that goal. The process of shifting our priorities will begin in the 90s, but will not be complete until early in the next century. It would be nice to believe that these changes will happen due to a general sense of altruism and good will. It is more likely to occur because we will be forced into it by ecological and economical crises. When pollution begins to make the disease rate skyrocket and global warming becomes more apparent, we will be forced to take environmental issues more seriously. The increasing cost of energy and other resources will force us to turn to local food production and manufacturing, and to turn "back to basics." The end: As the end of the decade nears, hoards of contractors specializing in Cobol and RPG will emerge to modify the thousands of old applications which will blow up when the year flips over from 1999 to 2000. And much of the rest of the industry will be occupied with endless arguments on whether the year 2000 or 2001 will mark the beginning of a new decade, a new century and a new millennium. :-) Meanwhile, have a happy new year! ------------------------------------ The opinions expressed herein are those of the author and do not necessarily reflect those of Data General Corporation or any of its related entities.