Market Roundup

October 3, 2003

IBM Broadens Virtual Offerings

IBM and SCO: The Latest Chapter

 Taxing Times for Sun

 BEA Launches Program Targeted at ISVs


IBM Broadens Virtual Offerings

By Jim Balderston

IBM has announced that it is expanding its Virtual Server Services to include hosted offerings running on its eServer xSeries, pSeries, and iSeries systems. The company had previously only offered such hosted services on its mainframe ZSeries computers, which it began offering in July 2002. IBM officials said its hosted applications offering could offer customers savings of 15-30% when compared to running comparable solutions inhouse. The IBM Virtual Server offering allows customers to access virtual, not physical, servers hosted by IBM, paying only for the processing power they need.

In one view, this announcement is yet another example of the downward transfer of technology from the high-end to the mid-range environment. As such, it is a move we believe will bring more interest to hosted services, especially among small and medium tier businesses, which in most cases do not need the raw processing horsepower that mainframes provide. By offering similar virtual server options at lower speeds and feeds, IBM is expanding its market opportunity by saying to customers that they can purchase what they need when they need it. Such arrangements are very appealing, in our minds, to those seeking to outsource the hassles of maintaining IT currency or those that have cyclical businesses with notable spikes in business activity during the calendar year, such as catalogs or retailers who make the lion’s share of their sales in the fourth quarter during the holiday shopping season. For these people, buying processing power as needed makes real sense, and by bringing the offering down to mid-range servers, such an offering may well become much more compelling to these users.

That said, there are still obstacles in the way of wholesale adoption of outsourced services. Customers concerns over such things as security — whether based on reality or not — are part of a larger phenomena known as “server hugging,” in which enterprises that could gain real benefits from outsourcing forego the opportunity because of the reluctance to let their IT infrastructure move out of their physical control. That reluctance is well-entrenched and has been for some time, as the once vaunted but largely failed ASP fad demonstrates so clearly. Yet times are changing, and if vendors like IBM can  demonstrate greater security, lower costs, and less muss and fuss for customers, this model could gain more traction in the coming years, especially if IT prejudices (or superstitions) against outsourcing begin to dwindle in the face of both broader acceptance and increased knowledge. By building these data centers IBM has an opportunity to address some of those concerns and can show off what amounts to a “model home.” Offering walk-throughs of such a data center may prove a compelling sales tool to many IT managers, especially when they consider the condition of their own IT budgets and deployments. 

IBM and SCO: The Latest Chapter

By AJ Dennis

IBM has expanded its Linux-related countersuit against SCO Group by accusing SCO of infringing IBM's copyrights by “copying, modifying, sublicensing, and/or distributing IBM’s copyrighted contributions to Linux on the terms set out in the General Public License (GPL) and only on the terms set out in the GPL.” IBM came under attack from SCO early this year when in March SCO Group filed a multi-billion lawsuit (amended in June) accusing IBM of breach of contract and of illegally incorporating SCO-controlled Unix code into Linux software distributed by IBM. IBM filed counterclaims against SCO in August, charging SCO with violating IBM patents, engaging in unfair trade practices, and violating the terms under which SCO distributed software. In this expanded counterclaim, filed in U.S. District Court for Utah, IBM added the charges of copyright violation. The claim cites seven pieces of copyrighted software IBM contributed to Linux under the GPL. By violating the terms of the GPL, IBM states, SCO violated IBM's copyrights.

Given the nature of judicial resolution of such a case, with the courts likely to take months, if not a year, to hear this case, IBM’s strategy appears consistent: the case has no merit; look like an industry leader; win in court, in the press, and around the water-cooler. SCO, with a less-than-satisfying history with IBM (remember Project Monterey), promotes a Chinese menu of assurances as to the merits of their case but with no discernable strategy, seemingly more interested in attention than resolution. In the six months since SCO picked this contractual battle with IBM, there has been little beyond posturing and positioning.

We understand and support the need to protect intellectual property and corporate assets. We also see that such a lawsuit was likely inevitable given the evolution of open-source and proprietary-source software models in a post-2000 tech industry, now fraught with change and challenge. However, we wonder if some of the means and methods SCO has deployed (promising to bill Linux users and threatening legal action against companies and individuals who don't pay up) best serves their professed goal of settling their contract dispute with IBM… or does SCO have another agenda? It has been apparent from the onset that one possible alternative goal would be to force IBM (or any other deep-pocketed organization) to simply purchase SCO in order to stifle their demands and remove the threat of extortion now hanging over the open-source community. It seems they picked the wrong target or used the wrong strategy with IBM if that was their intent. Time and evidentiary discovery will tell but with the action by IBM last week, we believe IBM will continue with its strategic premise: the case has no merit.

 Taxing Times for Sun

By Joyce Tompsett Becknell

This week, Sun Microsystems announced that it will have a $1,051 million non-cash charge in its fourth quarter for fiscal year 2003 in order to increase a valuation allowance for its net deferred tax assets. This amounts to the required warning that Sun foresees a difficult quarter for Q1 FY2004 which they have not yet finished. The official reasons cited were an intense market and competitive dynamics. The change means that Sun will have a net loss for Q4 FY2003 with the forthcoming quarter not looking prettier. The market responded by dropping Sun’s shares close to 15% on the first day of trading after the announcement.

There is good news for Sun amongst all the sturm und drang swirling about them. Sun has a history of fiscal caution: They have always been very careful with their books, and as a result, they are also a cash-rich company. Despite its reputation for fiduciary caution, the stock market has concerns that Sun is moving a large ship through stormy waters and this news seems to indicate potential compass problems. While Sun has made its success by filling the niche at the top end of the enterprise computing food chain, they seem to realize that they must expand their offerings to encompass a full range of business solutions. This means embracing the low end of the market to eliminate the risk that Sun customers will be enticed away when they are sent to Sun’s competitors to satisfy their low-end computing needs. Although Sun has done very well in the midrange to high end with Sparc and Solaris, they have not been able to bring the cost points of Sparc down to fit into the low-end space. Solaris is less of a problem because as McNealy has often proclaimed, software is free. The failure of Sparc to fit into a low-end box with a low-end price has meant that Sun has had to embrace one of its old adversaries, Intel, as well as AMD, who is probably easier for them to stomach. This of course conflicts with another of Scott’s timeless mottos (with apologies to the state of New Hampshire) – live free with Sparc/Solaris or die. The love/hate relationship that Sun has had with the Intel platform is well documented and demonstrates a history of hesitating, surging forward, and retreating in the face of customer ennui.  However, the continuing growth of Linux has given Sun the opportunity to embrace the IA-32 market without resorting to Microsoft products, which probably would result in McNealy’s spontaneous combustion.

This will come down to Sun’s ability to execute. The market is still watching as HP wrestles with its expanding volume platform base and the resulting carnage done to its home grown products. (Just where did Alpha, Tru64, PA-RISC, and MPE go anyhow?) Sun is an R&D company, not a volume distributor. They are good at making really big, really fast machines. Unfortunately the market has different tastes right now and Sun is working to make its systems more like current fashion. And it is simultaneously focusing on its software stack as well as making services a larger part of the system. The question is: Will it have enough time? Will it be allowed to make the inevitable mistakes? Will the customers stick around and wait when Dell, IBM, and HP have figured out so many of the answers already? Does the company that said “the network is the computer” have enough bandwidth? The really hard part, of course, is that it needs the right leadership, and Zander is gone. Joy is gone. McNealy is running solo at a time when Sun needs all the leadership it can get. McNealy is also a bit of a personality, and tough times and big companies don’t mix well with personalities. Neither Michael Dell nor Sam Palmisano cultivates a large public persona, and it hasn’t hurt either company and has probably helped at times. At the same time that Sun is changing direction McNealy needs to be upgrading his persona from captain to admiral and providing customers and investors with the reassurance they need that Sun is not just a survivor but also a winner. Sun will have to prove that it is more than just a fiscally conservative company on the leading edge of techno-geekdom but is also a serious contender for broad based enterprise business solutions. Sun’s stock holders will ultimately have to decide whether its current leadership is up to the task.

 BEA Launches Program Targeted at ISVs

By Tracy Corbo

BEA Systems recently announced a new program for ISVs called BEA Controls and Extensibility Program. There are currently thirty-three ISVs who have joined the program representing a variety of application segments ranging from application management, business intelligence, and content management to packaged applications. A partial list of vendors includes: Attachmate, Cognos, Computer Associates International, Confluent, Documentum, E.piphany, FileNet, Interwoven, MobileAware, NEON Systems,, and Yahoo!. The program is designed to simplify the integration of third-party ISV applications with BEA’s application server platform WebLogic. The ISVs will be able to create custom extensions and controls for WebLogic Workshop, the companion development environment for the WebLogic application server.

The application server market is a mature, highly fragmented one. BEA and IBM remain the major contenders with many other smaller players vying for mindshare if not market share. Despite the fact that the application server is not front page news, it does play dual role impacting both software and hardware markets. On the software side, especially for second- and third-tier ISVs, it provides a common infrastructure to develop and distribute their applications. This in turn will enable them to reach new customers that might otherwise have been inaccessible. Secondly, the application server platform provides an opportunity to foster a broad and diverse development community since the platform is not tightly wedded to a specific hardware platform or operating system.

On the hardware side, the impact is more direct, without viable software solutions any hardware platform is doomed. Remember Alpha? Consequently, hardware vendors such as IBM, Sun and HP depend on ISVs and the development community to keep them in the game. IBM has done a much better job of building a solid story around their application server platform WebSphere than either Sun or HP who have fumbled the ball making BEA an ideal partner for both vendors. The relationship is beneficial to all parties and the ultimate winner ends up the customer. The application server is about providing a common platform for application development and deployment across a wide variety of underlying architectures providing opportunity for multiple segments of the IT community.

The Sageza Group, Inc.

836 W El Camino Real

Mountain View, CA 94040-2512

650·390·0700     fax 650·649·2302

London +44 (0) 20·7900·2819

Munich +49 (0) 89·4201·7144

Milan +39 02·954·41·646


Copyright © 2003 The Sageza Group, Inc. May not be duplicated or retransmitted without written permission.