Windows/NT will overtake OS/2 by the end of the year, leading into an NT versus Unix showdown on the desktop; the world will wave goodbye to Systems Application Architecture as IBM’s Personal Systems Group launches an attack on IBM’s mainframe business; the ACE consortium will disintegrate; and Storage Technology Corp’s Iceberg disk array will be a big hit. Not only that, but Apple Computer Inc will lose the Microsoft Corp look and feel suit, and industrial-strength client-server software engineering tools will flood the market. These are just some of the controversial forecasts for 1992 being made by Technology Investment Strategies, a division of Framingham, Massachusetts-based market research group International Data Group, in a new report entitled Predictions ’92: Information Systems Hurtle Towards a New Order (CI No 1,830). By the year end, the report goes, Windows/NT, with an inherently stronger installed-base appeal, will storm past early starter OS/2 2.0, OS/2 being vulnerable to the snowballing challenge presented by desktop Unix.
Microsoft buys Banyan
As OS/2 falls by the wayside, can Microsoft win the last great battle of open versus proprietary operating systems? Meanwhile, the Technology Investment predictions continue, SunSoft gets off to a good start, being the first to deliver the rare combination of vision and product in the advanced OS wars; Digital Equipment Corp, frustrated by Santa Cruz Operation’s delays, will market its own (OSF/1) desktop Unix; and Microsoft will reduce the overwhelming advantage held by Novell Inc in local networking software by acquiring Banyan Systems Inc. As for IBM, Technology Investment sees its Personal Systems unit exercising its new-found quasi-independence, selling PS/2 and PS/1 models via mail order, acknowledging that value-added can’t beat price-reduced at the low end of the personal computer market. IBM Personal Systems will continue an unsuccessful middle market strategy of trying to force-feed high-end technology to the average user to justify higher prices, while IBM’s sales and marketing organisations demonstrating their own measure of independence from IBM’s personal computer group – will consider marketing clones from an Asian supplier. As the personal computer hardware market rises and blips, the winners are already named at the post as being Compaq Computer Corp, AST Research Inc, Advanced Logic Research Inc and CompuAdd Corp; losing ground, meantime, Technology Investment sees Dell Computer Corp relatively declining, and Northgate Computers Inc, Zeos International Ltd and Gateway 2000 sliding. While all this is going on, Apple Computer Inc Macintosh laptops will continue their strong sales of 1991. Apple’s suit against Microsoft, over the look and feel of Windows, will be decided in favour of Microsoft, states Technology Investment firmly. The bottom line being that the look and feel issue is a big one for the opens systems movement, since it addresses the area that makes users most captive to a software supplier – the user interface. When Apple loses the suit, the commoditisation that has been rampant in hardware, and growing in the operating systems area, will extend more firmly into the application software area, dashing the hopes of those that thought software would take over the proprietary supplier vacuum left by the systems supplier. Meanwhile, heavy investment in Unix as an off-the-shelf operating system will make proprietary operating system development increasingly infeasible.
IBM’s SAA goes open
IBM, says Technology Investment, will actively shift development and marketing energy away from proprietary products, towards its open systems offerings: the vendor’s proprietary vision of SAA will give up the ghost, with SAA becoming the framework for incorporating open systems standards; the Personal Systems group, if indeed it is as independent as IBM would have us believe, will challenge the likes of Hewlett-Packard and NCR Corp with its own programme – namely the RS/6000 – to move disgruntled mainframe customers to the open systems wor
ld. DEC will do the same. Open proprietary systems – DEC’s Open VMS, IBM’s Open MVS – will be the key meachanism for transitioning to open systems for companies with large investments in information technology. By the year end, Microsoft – last of the proprietary operating system suppliers to do so – will reluctantly commit to supporting open interfaces, such as Posix, in Windows/NT. Predictably, the defection of mini and mainframe users to Unix-based hardware will gather speed. Hewlett-Packard and NCR will notch up an increasing number of Unix wins, against the DEC VAX and IBM AS/400. There will be a proliferation of user-driven standards groups oriented towards specific industries. Retail, health care and the petrochemical sectors will lead the storm into the open systems arena. Standards profiles will become the new buzz phrase for 1992, confusing users and paving a crooked path to insert proprietary products into the unsuspecting user’s strategic plans. On the future of the Advanced Computing Environment, Technology Investment reckons DEC (OSF/1 and Alpha) and Microsoft (Windows/NT) will ship, while MIPS Computer and Santa Cruz struggle and Compaq focuses on re-establishing its own direction. The result: unhappy or uninterested partners abound; the consortium falls apart. Economic and political pressures will force together the efforts of the Open Software Foundation and Unix International; similar factors will push together, or oust, the Object Management Group, X/Open and the User Alliance for Open Systems. This will be the Year of Client-Server in the world of application development, says Technology Investment, as traditional names in software engineering – Intersolv Inc, Texas Instruments Inc, Andersen Consulting – join relative newcomers PowerSoft Corp and Co-Operative Solutions Inc. At least one of the top suppliers – Texas or KnowledgWare Inc – is destined to fail in the end to make the transition to the open client-server application development model. Distributed systems management products will proliferate.. distributed data access (IBM’s Information Warehouse) will percolate. Collector of old technology Computer Associates International Inc will extend its data centre management tools to the desktop and the local area network. Object-oriented technology will have its greatest impact – outside the world of graphical user interfaces – in distributed systems management products: Novell and Hewlett-Packard will ship simplified Novell local net management tools. The Open Software Foundation’s Distributed Computing Environment will ship in numerous vendors’ products; most system management tools launched this year will migrate to support the Distributed Computing Environment by next year at the latest (and the Distributed Management Environment, DME, by 1994). This year will also be the beta-test year for distributed data access, with implementations to follow next year. IBM’s Information Warehouse initiative will suffer the same organisational delays. At the high-end of the computer hardware market, StorageTek will, according to Technology Investment’s crystal ball, hit a triple – not a home run: there will be small delays in advanced functions – with shipment of its Iceberg large-scale disk array of commodity 5.25 drives.
CASE for maintenance
Small is beautiful: harnessing low-cost, powerful technology to build high-performance systems will be a growing trend. StorageTek and EMC Corp will do so in large-scale storage this year. NCR is building a successful business in systems and storage around this theme – all of which is more bad news for traditional large-scale mainframe and disk suppliers (not to mention the obvious name here). EMC will ship a large-scale disk array which, with StorageTek’s Iceberg, will put pressure on technology market leaders IBM and Hitachi Ltd. Multi-microprocessor based systems from the likes of NCR and Teradata will continue to nibble at core mainframe applications – technical computing, transaction processing and the like. IBM’s Repository Manager will continue to
stumble in performance until it is implemented on a dedicated repository machine (a good candidate for multimicro architecture). Specialised file servers with distributed file management intelligence is a next step; Auspex Systems points the way. And finally, high-end CASE tool suppliers, in the main, will bring out products that focus on improving the productivity of maintenance of legacy systems, in contrast to the community’s largely unsuccessful history of focusing on new system development. The 1992 plethora of maintenance-targeted tools will attack this giant problem and finally take software engineering out of the ivory tower, into the Cobol fix-it shop. We’ll see…
