Let's play musical chairs
by CBDi Forums Limited
The word on the street is that HP will divest its Web services software stack. The application server, dev and integration tools will go. Instead they will create stronger partnerships that provide their customers with choice and best of breed products, depending on the chosen platform.
HP and BEA?
When HP announced its merger back in September last year (tempus fugit), we commented that HP would have been better off merging with BEA. Our rationale was that the weak link in the HP armor was its software, and that Compaq, once a strong middleware player, had nothing to offer in this department.
However given the decision to merge with Compaq, HP actually has little choice other than to go with an open, best of breed software strategy. More than ever it now needs to span both the Microsoft and Java platforms in an industry neutral manner.
The problem is that the industry is consolidating and there's a game of musical chairs being played out. In our report on Sun Microsystems which we are publishing in this month's CBDi Journal, we report on the ever tighter integration between the Sun application server and Solaris, and the forces and strategies that are driving Sun to narrow down its partnerships. Whilst the Sun stack remains "integratable", the number of partnerships will inevitably be reduced in order to increase performance, differentiation and particularly quality.
Sun needs to increase its market share in the application server and software markets. It is going about this by embedding for performance and reducing price. Whilst Sun is denying any weakening in the relationship with BEA, we would now expect BEA to be considering very carefully how long it can pursue its cross Java platform strategy.
Industy consolidation now at advanced stage
This game is being played out against a background of gradual consolidation that started over two years ago. IBM WebSphere was not just about naming and branding, it was a deliberate strategy to provide a comprehensive, vertically integrated Web services platform that was more innovative and more comprehensive than Sun's. However, Sun's recent response is actually no knee jerk reaction, they have also been moving in this direction for some time now.
The issue is that within any one layer the products are all becoming commoditized and undifferentiated. This applies to hardware, operating systems (Linux), middleware (Java) and infrastructure software in application servers. So if it's hard to compete at any one layer, a predictable response is to create a comprehensive, vertically integrated stack including hardware, software and services. Sun might have lost the app server battle, but they stand a better chance of winning the platform war.
At times methinks Scott McNealy protests too much, because he is doing pretty much what he accuses Microsoft of doing -- creating a well- integrated stack that is attractive to the consumer. Perhaps only Oracle is really honest in this area; they have been saying for years that it pays to deal with one supplier because their products are designed to work together, and integration costs are probably lower.
So at last we start to see our predictions come true. For years we have been forecasting that the future of software must follow the course of other mature industries, such as automotive, electronics etc. And consolidation is an economic necessity that cannot be avoided, and indeed a compelling response to increasing customer service demands.
So instead of Ford, DaimlerChrysler, GM, Volkswagen, Toyota and Honda, we can clearly see Microsoft, IBM, Sun, Oracle and possibly BEA with deep and well established software platform stacks and a significant level of loyalty in their respective customer base.
And then there were six
In another significant announcement Novell announced they are acquiring SilverStream, a well regarded second rank application server player. This move has been well received by analysts, commentators and media alike, who all nod wisely that Novell had a huge gaping hole in their strategy, and what was needed was an application server. In fact Novell has acquired a quite comprehensive Web services stack that comprises in addition to the application server, an integration and development platform.
However it does seem a little strange that it took Novell so long to see this requirement. Further we might observe that Microsoft, IBM, Sun and others have been building their competencies for many years, and that this is not just about software. By acquiring SilverStream Novell has acquired a product, but not a significant market share. Their post merger revenue projections indicate they understand the magnitude of the task.
Let's play musical chairs
Although the big six platform vendors would like to see homogenous stacks implemented, most customers will actually reject this notion, not necessarily because of the potential lock-in, but because the nature of the enterprise marketplace is highly diverse and it is extremely unlikely that one size will fit all.
Whilst the stacks all seem to cover a very similar functional footprint, the reality is that they do this in quite different ways with very different approaches to functional coverage, componentization and commercial arrangements. Sun and IBM have continually increasing depth in their stacks, which include hardware and services. MSFT already has a deep, hugely rich and comprehensive stack and is rapidly increasing the vertical coverage with forays into applications and devices.
In contrast BEA and Oracle look like niche players. A further concern for Oracle is that they are now being overtaken in their core database market. So for Oracle and BEA the question is whose stack are they going to be part of? Clearly HP looks like a strong prospect for a best of breed stack. And we note in this week's announcement on HP-UX for Itanium, HP features the BEA relationship and WebLogic app server, at the expense of its own Bluestone product.
But the music has not yet stopped for the last time. Over the past couple of years there has been feverish activity in the Web services support and development tools areas. Many start-ups as well as existing companies have brought really imaginative products to market that extend conventional thinking on platform functionality, and interestingly all the major platform vendors are lagging in this area.
A good example is in the area of Web services management where companies such as Amberpoint, Avinon, Blue Titan, Infravio, Systinet, Talking Blocks and more, are implementing functionality that to all intents and purposes looks like platform functionality. It looks like a smaller version of musical chairs will be starting in this area very soon.
In the development tools area it looks like the game has already commenced. Last week The Register reported that WebGain is about to announce the sale of its TopLink ObjectRelational tools business to Oracle along with the exit of WebGain from the IDE business by putting its dev products into the open source domain. Oracle, actually one of the less acquisitive of the big vendors, will add an important component to its stack, and WebGain investors will realize a return on their investment. Of course Web Gain customers will be wondering how this will affect them.
And that's the issue. When you work with a vendor you are placing a great deal of dependence upon them. If they disappear, it costs you time and money, or worse. So the big three, five or six have a strong case that they are a safe decision. And right now, everyone is highly risk averse. So in many ways the case for consolidation is even stronger than ever and is likely to happen even faster because it will be seen as a lower risk solution.
For start-ups the issue is how to provide guarantees to potential customers. How to defray the risk of working with a small, independent company? And here partnerships with the major stack owners is the key, and in the context of musical chairs, those who do not carry enough weight to have their own chair, need choose whose lap they are going to sit upon.
And that is the major issue for Novell. They are not yet part of the core platform vendor set. It will take them a few years to prove that they can stay the course. One swallow doesn't make a summer, and one defensive product acquisition does not make a platform vendor part of the big six. We wonder whether Novell would have been better off making strong partnerships with the likes of Oracle and BEA, rather than acquiring a small player that brings them no significant market share. Is it any harder to make partnerships work, than it is to make acquisitions work? The same applies to Oracle and BEA. Can they survive as generalized platform vendors? How long have they all got before the music stops?
Of course there will be new games starting soon. For example if Microsoft, already with Great Plains under its belt, succeeds in acquiring Navision, we can expect a new game in the applications arena. And the same applies to the mobile arena, a market which is currently under massive pressure and set for consolidation.
Take your chairs, and place your bets.
Feedback and comments are welcomed: firstname.lastname@example.org
Copyright CBDi Forum Limited 2002. The CBDi Forum is an analysis firm and think tank, providing insight on component and web service technologies, processes and practices for the software industry and its customers. To register for the weekly newswire click here.
For More Information:
- For insightful opinion and commentary, read our Guest Commentary columns.
- Tired of technospeak? The Web Services Advisor column provides a clear understanding of Web services.
- Looking for shortcuts and helpful developer tips? Visit our Tip Exchange for time-saving XML and .NET tips.
- Visit our huge Best Web Links for Web Services for hand-picked resources by our editors.
- Discuss this article, voice your opinion or talk with your peers in our Discussion Forums.
- Visit Ask the Experts for Web services, SOAP, WSDL, XML, .NET, Java and EAI answers.