Without that "vision thing" going, can IBM deal with a Java-enabled future or is it merely buying time for its big iron past?
Several decades after middle school, I still remember the origin of “January” — the two-faced mythical being that looked back into last year and forward to the coming one. IBM is putting on a similar set of faces now with its networking strategies, appearing to be an industry trend-setter by embracing Java while thriving on mainframe revenue streams.
IBM is one of Java’s — and Sun’s — best friends right now. Acquiring Lotus Development Corporation gave IBM a strong, network-oriented software development business unit, highly aligned with IBM’s public decrees that the company would pursue “network-centric strategies.” Analysts chided Lou Gerstner for not delivering a vision in his first months at the IBM helm, when in fact he’s been steering on Sun’s path of network computing for some time. Lou doesn’t need a vision if he can execute well; the stock market has voted him a winner by acclamation over the past 18 months.
IBM is vigorously supporting Java, producing an MVS port of it to run on mainframes, as well as a much-rumored Windows 3.1 port of the virtual machine. More interesting is the San Francisco project, an entire business framework written in Java. San Francisco reminds me of the Tivoli Management Environment — but it’s a development environment aimed at business systems instead of systems management. (Like Lotus, Tivoli is now an IBM subsidiary.)
But this is also the point at which I get worried. San Francisco, a shared applications framework, starts to sound an awful lot like IBM’s last architecture for the enterprise, the SAA initiative. SAA appeared to be an IBM platform lock, at the wrong time in the market. So now for today’s question: Is IBM’s latest commercial application framework aimed at preserving mainframe systems with new front ends, or is it really a new way to build light, functional business systems out of components?
Java is a perfect tool with which IBM can drive continued mainframe investment. It’s PowerBuilder for those JVM-enabled MVS systems. With new, network-aware front ends, fielded applications may get a new lease on life. Existing mainframe backends remain relatively unchanged; mainframe revenues continue on for the short-term future. The IBM gods smile. The stock market pundits smile as well, until someone asks what will replace a declining mainframe investment pool.
What this May-December technology marriage needs is a bit of invigoration or inspiration. Fortunately, IBM has it in the form of Lotus: a Java development shop that really is driving component-based architectures with the Domino server products. Lotus represents a small part of the IBM revenue stream, but if it can get customers thinking about life after mainframes, IBM and its stockholders have much to look forward to. If IBM’s Java computing strategies turn out to be mainframe business as usual with a prettier forward-looking face, some future January will look back and see San Francisco as yet another example of the bloatware that Java was supposed to combat.
In the low 60s, IBM could be at an unstable price point. Further public attention lavished on its newfound love of the network, coupled with solid earnings from a broader spectrum of business units, and this is a 00 stock. Should IBM fail to deliver, its shares could see 10 again without too much of a push. For now, we’ll sit in that saddle point and watch both paths for some additional buying or selling signs.
