July 23, 2004
The Liberty Alliance, an industry trade group focusing on problem of network-wide single sign-on, has announced several new members. The three-year-old organization now counts more than 150 members, with some of the technology industry's top vendors including Oracle, Intel, Computer Associates, and Sharp Laboratories signing on for full participation in recent months. The Liberty Alliance develops and promotes open specifications for interoperability and management of user identities within the LAN, through the firewall, and across the Internet. The group's goal is to enable secure interoperability across heterogeneous systems. To illustrate its progress, the Liberty Alliance is organizing several member demonstrations later this week at a conference in San Diego. Among the demonstrations, Novell will demo its Liberty-based Identity Federation Solution code-named “Odyssey.” This technology, expected to ship in the first half of 2005, seeks to make it easier for organizations and their employees to conduct business electronically within a company and with its partners, through centralized authentication, policy management, and single sign-on. In addition, Odyssey will be able to store identity attributes locally plus connect to existing LDAP directories, extending the utility of the solution to individuals and small organizations that do not have network-based directory systems.
In examining the available information on Odyssey a bit more in depth, we were encouraged to see significant Novell value-add while maintaining rigorous conformance to the Liberty Alliance 1.2 specification and protocol. While current Liberty-enabled federated identity projects focus on sharing identity information and require that users have accounts on each federated system, Odyssey goes several steps further to support role-based policies for controlling access to Web sites and other applications either within an organization or with its partners. When one accesses a partner site where one does not have an account, those policies could allow a user the option to automatically request an account using information in the corporate directory relevant to that partner. For example, business partners like airlines, hotels, and rental car companies could be allowed, by the customer, to integrate business processes that share relevant identity attributes between systems for more efficient and effective transactions. Odyssey would then dynamically provision a user account and enable single sign-on to the partner system. Additionally, Odyssey will provide proxy functionality so that when it is set up in front of another server, it will dynamically Liberty- and/or SAML-enable (Secure Assertions Markup Language) virtually any application or service. Throughout this automated process, users control how their information will be shared with other systems to maintain a level of control that is comfortable for each individual user.
Overall, we are encouraged by Novell’s offering in this area. The offering reflects Novell’s extensive knowledge and experience with directories and their relatively recent appreciation for delivering leading products by following the market’s penchant for standards-based computing
Microsoft and Apple have been sued in U.S. Federal Court for patent infringement by a British company, BTG plc, which claims that the two companies are using Web-based software updates without paying licensing fees to BTG. BTG charges that the two companies are in violation of U.S. Patent 6,557,054, which they say covers online software updates. BTG said it filed the suit because Apple and Microsoft had not entered into licensing agreements quickly enough, and said that it was actively negotiating with unnamed other software vendors. The patent was applied for in April 2000 and granted three years later, in April 2003. The patent’s description does not offer technical details of a specific process but only a conceptual description of software updates.
Remember “push” technology? Remember all the hype surrounding it? The fundamental idea was that information or application code could be delivered proactively to the user, without the user seeking out the information manually. Most companies that built their entire business model around push went belly up, as they were disintermediated by both content and application vendors that applied forms of push to their own offerings. In other words, long before this conceptual patent was applied for, automated software updates were being discussed if not applied. In fact, some of the supporting documentation for this patent application included articles dating back to the 1970s. A new concept? Hardly.
This legal action highlights an ongoing and at times serious problem within the U.S. Patent Office, especially when it concerns itself with these conceptual patents. Essentially, the Patent Office is endorsing concept squatting, something akin to buying up high-profile URLs and forcing companies associated with them to cough up big bucks to gain possession of the URL that is most closely associated with their particular brand. What seems clear to us is that the Patent Office is unable to keep pace with existing technology innovations. Even if software updates were not ubiquitous in the year 2000, there was enough evidence to suggest they would be in the near future. By the year 2003, such technology was no more unique and unusual than mobile phones. Surely some review process of the patent application would indicate that its conceptual nature was more an attempt to garner the stuff of legal blackmail than to actually protect a notable piece of intellectual property or process. Simply stated, this is patently absurd.
On Tuesday, the U.S. House of Representatives passed by a 321-111 vote a measure that blocked a Federal Accounting Standards Board (FASB) regulation that would have forced companies to expense the value of stock options on their income statements. The House bill stipulates that the only options that must be counted against revenues are those granted to the five top executives in companies with annual revenues of more than $25 million. A similar Senate measure has been blocked by Banking Committee Chairman Richard Shelby (R-AL), but House supporters said their bill might be attached to must-pass legislation that would force a floor vote later this summer. Numerous IT vendors including Intel, Sun Microsystems, Cisco, AutoDesk, and Apple, and technology trade groups TechNet and AeA, had lobbied aggressively for the House measure. SEC Chairman William Donaldson, Fed Chairman Alan Greenspan, Berkshire Hathaway chairman Warren Buffett, and the Big Four accounting firms have all vigorously supported the FASB regulation. In a separate event, Microsoft announced that its board of directors had approved a quarterly $.08 per share dividend, a special one-time $3 per share dividend, and a plan to buy back $30 billion of the company’s stock over the next four years. All told, the plan is expected to provide up to $75 billion in value to shareholders over the next four years.
In the Roaring 90s leading up to the dotcom bust, stock options acquired a semi-mythic status in Silicon Valley. Options were the nectar that attracted high-value employees to underfunded startups or the glue that bound them to companies looking for an edge, and most everyone knew or knew of someone who had cashed out their vested options and moved on to a life of ease, bliss pursuit, and esoteric charity work. While Valley mavens insisted then and now that options are the key to IT companies being and remaining competitive and innovative, the ugly little secret is that while they may be classified as “contingent obligations” options look, smell, and taste exactly like debt. At least that was the conclusion the FASB came to in 2002, after the Enron and WorldCom debacles had revealed options as perks that could be easily transformed into balance sheet scams. Since then, a host of IT companies and their Congressional lackeys have fought to undermine or overturn the FASB. Their logic may be flawed, but their motivations are clear enough. A Bear Sterns report estimates that expensing options could cut technology profits by up to 60% on average. For an industry with many members who continue to totter along the edge of disaster from quarter to quarter, such a financial reality check could literally spell doom. However, while undermining the FASB may be understandable, it does not address the fact that companies that do not honestly expense options are, by result, overstating their profits. This is a simple lesson in arithmetic investors would be well advised to remember.
Understated profits are not one of Microsoft’s problems. In fact, the reverse has been the case with the company coming under increasing fire for sitting on a growing cash hoard that makes one wonder if Rumpelstiltskin is directing Microsoft’s R&D efforts. Nevertheless, after months of rampant speculation about what the company would do with all that filthy lucre, Microsoft’s decision to put aside childish things and finally assume corporate adulthood came as a significant surprise. However, given the company’s decision to begin expensing options last September, this move makes eminent sense. The fact of the matter is that companies issue stock to acquire money needed for growing or improving their position. Microsoft is clearly beyond this stage of development, generating, according some estimates, more than $10 billion in excess cash annually. Mature enterprises understand that once they go public, their fiduciary responsibility extends to the shareholders who own sizable portions of their organizations. Microsoft’s decision to eat at the grownup table is a mark of just how far the company and the IT industry have come. One only wishes that the IT vendors opposing the FASB were ready to abandon their own adolescent excesses.
IBM has announced new On Demand pricing options for its eServer i5 solutions that the company claimed would deliver unprecedented flexibility for clients who are simplifying or consolidating their IT infrastructures. The new pricing models target clients deploying new workloads on IBM i5/OS, and offer enhanced options for clients who want to mix and match multiple applications and operating systems on single servers. IBM stated that the new options allow customers to purchase the appropriate amount and type of processing power for current needs, while maintaining the ability to expand their i5 infrastructures to accommodate future requirements. The new options will be implemented by the end of the month. No pricing details were included in the announcement.
IBM’s eServer iSeries may qualify as the best loved, least known server solution in the IT world. The SMBs who make the jump to the iSeries tend to never look back, and are usually the platform’s biggest boosters. What do they like most? Probably the combination of system flexibility, manageability, and reliability that is, quite simply, not available in any other server aimed at this marketplace. Toss in the iSeries’ historic ability to simultaneously run multiple operating environments i5/OS (OS/400), Linux, and Windows and you have a platform not quite like any other: this is part of the problem. Enterprise IT tends to consist of solutions with pricing, performance, and capabilities that are roughly equal to others in their class. As a result, many treat products that do not fit these well worn molds with suspicion or disregard. More fools they. The new POWER5-based i5 delivers even more performance, adding AIX to the OS mix, and notably expanding system flexibility via the IBM Virtualization Engine and micro-partitioning capabilities that allow each i5 processor to support up to ten virtual servers.
If this all seems a bit familiar, it is; at least to those who have some knowledge of IBM’s zSeries mainframe offerings. Since 2000, IBM has been notably successful at migrating the mainframe’s self-managing and self-healing capabilities across its other server platforms. Indeed, the company’s On Demand strategy focuses on the goal of making IT infrastructures as scalable, resilient, and dependable as a mainframe. But with the POWER5’s new features, tangible mainframe-style computing just became available to entirely new classes of IBM iSeries customers. The overt flexibility of company’s new eServer i5 pricing options are significantly different than what enterprise customers would expect from other vendors, but at the end of the day they simply reflect the unique capabilities of which existing iSeries customers are well aware.
Duke University announced this week that it would provide its 1,650 incoming freshman 20GB iPods this August 19 when the new freshman class arrives on campus. The iPods will contain university content, such as orientation materials and the school year calendar. The university will also provide a Web site where students can download course content such as language lessons and recorded lectures as well as audio books. The university is also working with faculty to develop new services and content for the university’s network and the iPod experiment. Duke will spend approximately $500,000 on the iPod experiment.
Good for Duke University. The institution recognizes that they have an opportunity to apply a highly familiar technology to their student population in a way that has notable impact on the university and the students themselves. Students are no longer directed to Web sites or student handbooks for academic and social behavior rules and expectations; instead they are handed them on the first day. The iPod now becomes a much more generic personal storage device (PSD), and the university now has new means to put critical information into the hands of the students. In doing so, the university has placed in each student’s hands campus rules and regulations, allowing them to say with much greater confidence that a legally defensible effort has been made to inform students or those regulations. Just as the iPod allowed users to take control of what music they listen to, now the campus iPod gives student complete control over the information they need to know. Whether they read it or not does not affect the basic premise; the university has transferred much more than just data to the students, it has transferred responsibility.
For the enterprise, personal storage devices are already becoming an indispensable tool for many workers, whether in the form of iPods or like devices or simple USB memory sticks. Many enterprises in turn are attempting to ban their use, citing security concerns. This panicked response to personal storage devices tells us that such devices have an irresistible future of ubiquity within the enterprise walls. The adoption of such devices does not bode well for PDAs, which are losing their appeal in the face of greater functionally in mobile phones; personal storage devices and these new phones can offer most of if not more than most PDAs. That said, there are legitimate security issues associated with PSDs, and enterprise networks will have to adapt to their use, just as they did when desktop Internet access became an irresistible demand within the enterprise. Comprehensive and granular security in the form of access control, management of rules and permissions, as well as defenses against viruses and intrusions must come from the enterprise network itself, as trusting users to meet these requirements would be pure folly. iPods and every other form of PSD are here to stay, forcing new opportunities and responsibilities on both enterprise networks and users themselves.
During its quarterly conference call to discuss earnings with the financial community, Sun Microsystems’ president Jonathan Schwartz stated that Sun had begun looking at delivering Solaris on POWER, as well as Solaris on Itanium, as ways of driving incremental volume. Schwartz was interrupted by Sun's CEO, Scott McNealy, who said the comment was not a product announcement. After the call, Sun representatives refused to comment on when or if such products would ever be released.
For those of us who have been following Sun for years and were mighty fond of the company back in the days when they loved to engage analysts in intense discussions of technology and markets, Schwartz’s comment was a notable milestone and a pause for reflection. Like a teenager struggling with hormonal difficulties, Sun has gone through some awkward growth stages. On more than one occasion industry pundits have sketched Sun’s epitaph on the belief that this rebel without a cause was doomed to crash and burn. However, Sun always managed to scrape through and frequently seemed to come out better for the journey despite the resulting scratches and bruises. However, it is hard to say whether Sun will be able to survive another similar crash. This company carved its very existence out of a religious notion that UNIX (particularly Solaris) on RISC (particularly UltraSPARC) was the one and only answer to high-end computing. Sun regularly ridiculed and mocked Microsoft and Redmond’s growing hoard of merry men and by dint of technical excellence and market savvy raised a generation of acolytes and true believers within both the employee and the customer base. But things change inevitably. Sun did not ride out the Internet bust as well as IBM, Dell, or the rascally Microsoft, nor has it enjoyed the massive restructuring excuses HP has fallen back on time and again. More importantly, a host of the company’s expensive strategic moves, such as the dreadful Cobalt acquisition, qualified as outright disasters.
Sun has suffered losses and layoffs, and has restructured its organization and product offerings. But the most dangerous change the company has embraced is to step back from their religious promotion of Solaris+SPARC, a move that became glaringly evident during the company’s stunning make-nice performance of Kumbayah with Microsoft. Part of what has ensured Sun’s success and the devotion of their followers is that they remained zealous and committed themselves 100%+ to their vision. But the company’s SPARC development roadmap has been plagued with potholes of late, resulting in Sun’s recent strategic partnership with Fujitsu. Sun has finally committed to Solaris on the x86 architecture for the volume space, an effort that should be bolstered by its Opteron plans. And now, at least according to a “slip” of the tongue (which had all the earmarks of a trial balloon) by Jonathan Schwartz, the company is seriously considering porting Solaris to IBM’s POWER and Intel’s Itanium. There are some good reasons for this move. POWER has become the volume leader for 64-bit computing and Itanium, at least according to Intel and HP’s marketing literature, is an industry standard platform. Embracing either platform or even both should allow Sun to establish a presence in its competitors’ datacenters and expand the options of its enterprise customers. But the real risk here is not technological so much as it is evangelical. Sun’s fortunes have tended to rise and set according to customers’ trust in the company’s vision, and porting Solaris to POWER and/or Itanium would be a bitter pill for many inside and outside of Sun to swallow. Such a move may be logical from both business and technical standpoints, but logic seldom plays a significant role in issues of faith.