Columns

Open source upsets the apple cart

• By Tony Baer
• May 1, 2003

Microsoft’s recent quarterly SEC 10-Q report cited open source as a potential threat that could force dramatic software price cuts. Pricing, however, is just the tip of the iceberg. Open source will also require vendors to demonstrate their core value propositions.

Software has grown more valuable than hardware, and is dwarfed only by the consulting services necessary to make it work. Credit that to Moore’s Law, which commoditized powerful hardware; the Y2K challenge, which prodded firms to replace aging homegrown systems with expensive enterprise packages; and the emergence of LANs.

But just as Microsoft and enterprise software vendors began accumulating success, open source emerged, initially at the periphery of computing with the Linux OS and Apache Web server. IBM’s 2001 announcement that it would channel at least $1 billion to Linux and related open-source R&D made the notion respectable. That was followed at LinuxWorld 2003 with the first wave of CIO Enterprise Linux strategy announcements. Suddenly, Linux and open source became too looming to ignore.

From a development standpoint, open source turns the outside world into an R&D lab. From a vendor standpoint, open source provides (choose one): a means of making their technology a de facto standard, a way to redirect competition from perceived commodity to unique areas of strength, or an invitation to give away the family jewels to rivals.

The first two strategies are actually intertwined. For instance, through their respective Eclipse and NetBeans open-source projects, IBM and Sun are attempting to make the frameworks surrounding their J2EE app servers de facto standards, pointing to their strengths in platforms, middleware, third-party products and, for IBM, services.

Although open source has morphed from communitarian ideal to “Realpolitik” market strategy, most vendors are uncomfortable with the idea of prying open their source code. Many have experimented with politically correct-sounding variants such as “shared source” or “community source” that preserve underlying technology ownership.

Beyond all the spinning and positioning, open source is forcing vendors to revisit the question, “Where’s the beef?” Traditionally, value is what the customer buys, and competitive edge represents the way a vendor uniquely delivers that value. It could come through pure technology advantage, domain expertise, delivery excellence, or by virtue of a commanding market presence that implies that the product -- and the support accompanying it -- remains available.

The software market is rife with contradictions as to what does and does not add value. The success of Linux proves that operating systems alone don’t carry intrinsic value. Yet, Microsoft Windows illustrates the value of market dominance that draws all the third-party products and assurances of ongoing support.

Conversely, while the emergence of MySQL proves generic relational databases aren’t anything special, it’s hardly dented Oracle’s market share. And while TomCat and JBoss validate the notion that low-end app servers are commodities, J2EE’s usual suspects have responded by differentiating their app servers at the high end through enterprise integration, content management and portal capabilities. BEA is fighting fire with fire by offering a stripped-down, $699 servlet-only version of WebLogic.

But just because Linux or MySQL is successful doesn’t mean the market expects vendors to expose their goodies. Veritas, whose backup file system dominates the market, provides a case in point. On its own, a file system adds no value. But Veritas’ ease of use, versatility, support for third-party data sources, and the reality of its market presence does add value. While the market might not demand open-source file systems, would such a strategy provide a means for Veritas to enshrine its dominance, or simply give rivals a free ride?

Ranajit Nevatia, who directs Veritas’ Linux strategy, has another answer -- there aren’t enough developers out there to make open source worthwhile.

But as Enterprise Linux becomes the poster child for open source and IT budgets remain tight, the question of where real value lies will remain. Customers may not necessarily demand open source, but neither will they pay for functionality that’s a commodity.