04.11.06
Posted in Linux, Middleware, Open Source, Technology Market Trends at 10:11 am by Tony Baer
After a much publicized courting, JBoss has become engaged, but to someone else. Nonetheless the chase was a bit fun while it lasted. In fact, just about anything about JBoss president Marc Fleury is amusing because of the way he has personalized his venture’s David and Goliath saga.
But make no doubt about it, this is serious business. As we noted several months back, although sentiment is hardly unanimous, open source is a business, not a Robin Hood crusade.
While Fleury went on baiting folks like IBM in public, his quest to make JBoss the next dominant Java platform was deadly serious. A couple months back, the rumored suitor was Oracle, which has started its own open (or to borrow a term from BEA, blended) source strategy. While we could imagine Fleury et al wanting to cash out, we had a hard time imagining how JBoss would have stayed relevant inside Oracle, where it would have been another side show.
Although JBoss and Red Hat follow different open source business models, we think this one’s a much better fit because JBoss becomes Red Hat’s de facto middleware stack, rather than just another piece.
For JBoss, this is of course the natural exit strategy for any up and coming startup that’s hitting the wall. For Red Hat, it’s another piece in the puzzle to become the de facto platform alternative to Microsoft. We wouldn’t be surprised if database was next on Red Hat’s list, as analyst Brenda M. Michelson of Patricia Seybold Group ventured.
Or as fellow analyst Keith Harrison-Broninski suggested, it’s another step in the Microsoft alternative building itself into another Microsoft. He suggested that in the short run, developers would embrace this deal, but he’s worried about the implications for the long run. We agree with part of his sentiments.
As we’ve noted previously, the popularity of open source isn’t because it’s open source, but because it’s a way for customers to acquire commodity technology for a commodity price. Admittedly, as open source powers like Red Hat bulk up their spread and penetration, there’s always the question of vendor lock-in, which in turn leads to price escalation.
But somehow we doubt that here. Recalling the argument of Sun’s Jonathan Schwartz about Red Hat forking Linux and other open source technology, the open source community pretty well debunked that myth. The binaries might be protected, but not the source code. OK, that’s a geek argument. But at a more important level, if Red Hat got too big for its britches, we’d expect IBM to up its ante in Novell or buy SuSE outright to keep Linux commodity -– which is exactly what customers want. Heck, it’s already done that with Gluecode, built by several former JBoss developers, although for now IBM has largely kept it caged for fear of undercutting their WebSphere franchise.
Red Hat has had a checkered history when it’s come to acquisitions. One of its best was the acquisition of Netscape’s LDAP directory when AOL was divesting the technology in the late 90s. But it also wrote off four other acquisitions in the post 9/11 funk of 2001.
But we think this one’s more in the Netscape Directory mold, in that it adds an obvious piece to the platform. And it bulks up what will hopefully remain a commodity alternative for enterprise customers.
Permalink
04.05.06
Posted in Outsourcing, Technology Market Trends at 10:10 am by Tony Baer
Every few years, a new tome becomes the bible of business. Back in the
80s, it was Tom Peters’ “In Search of Excellence.” In the IT field, NY
Times columnist Tom Friedman’s tome “The World Is Flat” has become the
latest conventional wisdom.
Friedman contends that, in a world awash with bandwidth, barriers to entry
have virtually disappeared. Witnessing the explosion of the IT sector in
places like India or China, it’s hard to argue. Last year alone, India
graduated 10x the number of computer scientists as the U.S.
Yet, viewing this as a zero sum game ignores some basic realities. For
instance, when you outsource offshore, you must increase management
overhead because you are running a remote project that is hampered by lack
of accessibility, 12 hours of time zone difference, and culture gaps.
Conversely, when you “insource” a project, you may also have to bulk up on
management to compensate for poor communications that have long existed
between IT and the business.
Yet, at Sand Hill Group’s Software 2006 conference, we were intrigued by
an idea cited by University of Michigan business professor Dr. C K
Prahalad that offshoring provides a multiplier effect. Applying principles
resembling Metcalfe’s Law (which stipulated the value of the network
growing exponentially with the number of nodes), Prahalad said that when
you have more software brains, you gain more chance of getting innovation.
And innovation, of course, creates value and new opportunity.
Arguably, nobody looks at the growth of SAP, Mercury, or Fuego (recently
acquired by BEA) as threats to domestic IT employment. Yet each of those
companies capitalized on development conducted in Germany, Israel, and
Argentina, respectively.
Prahalad made another interesting point: regardless of how fast India is
growing, it remains highly interdependent on western talent that not only
has state of the art expertise, but intimate knowledge of the customer and
market.
If you look at how manufacturing globalized, you might get an idea of
what’s in store for IT (we’ll also credit Erik Keller and Brian Turchin,
who have come to similar conclusions). A generation ago, Japan applied
American doctrines of total quality management to produce cars faster,
cheaper, better.
Yet, in the interim, American manufacturing didn’t disappear. Instead, it
reconfigured its role in two pivotal points of the value chain: ideation
and final configuration and service for the end customer.
Nonetheless, you can’t ignore the fact that IT employment in the western
world is nothing compared to the dot com peak. For instance, according to
Federal Reserve Bank figures, “non agricultural” (the bucket under which
they put IT) employment in the Bay Area dropped 9.5% between 2000 – 2003.
But we contend that the dot com bubble was an historic aberration. Taking
the Fed’s own figures, if you factor out the step gains of 1995 – 2000,
what remains is a more sustainable 10 – 15% overall growth in jobs from
1995 – 2003. Consequently, if you look at IT as a more ”normal” business,
job trends don’t look that bad, the emergence of offshore notwithstanding.
Instead, the real issue for the software industry remains finding a viable
business model. You’d think that after 30 – 40 years of existence, that
we’d finally get it right. Yet, according to numbers cited by Kleiner
Perkins principal Ray Lane, three quarters of all profitability is
accounted for by Microsoft, SAP, and Oracle. Now that’s something to get
worried about.
Permalink