By Timothy Prickett Morgan
IBM has been Sun’s most vocal champion for its Java programming standard, and it probably came as no surprise to many in the industry last week when Patricia Sueltz, the IBM executive in charge of Big Blue’s Java efforts, was offered and accepted the position at the head of Sun Microsystems’ Software Products and Platforms division. Sueltz replaces Alan Baratz, a long-time Sun executive who left the company in August to work as a venture capitalist at EM Warbug Pincus.
Sueltz, who has been at IBM for two decades and who is 46 years old and therefore probably too young to take over IBM’s Software Division any time soon, has seen plenty of action and seems well suited for her new job. In 1991, she directed an 800-person development and marketing team that put together improved versions of IBM’s CICS transaction monitoring software and did the initial work on its MQSeries of messaging middleware. She has also been a vice president of Internet Software at IBM’s Internet Division and was a technical advisor to chairman Louis Gerstner from 1993 to 1995. In that high-level position, Sueltz got a view of IBM’s vast software and server offerings. More recently, Sueltz has worked as the general manager of Java and OS/2 Software at IBM’s Network Computing Software Division. Her responsibilities in that general manager position were essentially three-fold: put in IBM’s two cents about how the Java standard should be enhanced; see that IBM’s various hardware and software divisions got the Java religion; and tweak OS/2 enough to keep its vast installed base of customers happy, but encourage them to move to more modern platforms and use Java.
Sueltz has been instrumental in forging all of IBM’s Java alliances, including those with Sun, Oracle, Netscape, Novell and AOL, and has been particularly keen on making sure that Java and XML play nicely together. At IBM, she was in charge of 3,000 developers and marketeers in IBM locations located in the US and the UK. While neither Java nor OS/2 have directly brought IBM much money – at least not in the IBM scheme of things — her Java work and its long-term affect on IBM has been dramatic.
What we have been looking for since this summer when Alan departed is a seasoned manager, says Ed Zander, president and chief operating at Sun. We wanted someone that could develop an integrated software strategy inside of Sun and build a major presence for Sun in the marketplace. For her part, Sueltz is thrilled to have a chance to run what is effectively a software company, and has picked up the Sun marketing speak pretty fast. I believe that the network is the computer, she explained in a teleconference last week announcing her new job. The vision and focus of Sun towards open market technologies is one that is going to move the entire industry forward. I’ve committed to extending Java, Jini, Solaris, iPlanet and other software technologies which are good for the industry, good for partners and good for customers.
In taking over the helm at Sun’s software business, Sueltz has plenty of work cut out for her, and the short-term stakes are a lot higher than pushing IBM’s long term Java agenda has been. The software business that she is taking over comprises three areas: Java and its related technologies such as Jini; the Solaris operating systems for Sparc and Intel architectures; and software technologies based on these two areas for embedded and consumer applications. Sueltz also inherits the new Star Division office application suite and any potential headaches that will come from trying to take market share away from Microsoft on the desktop with the Windows and Linux versions of the StarOffice suite and from trying to launch a Java-based network computing version of the StarOffice suite, called StarPortal.
Both IBM’s Lotus division and Corel’s WordPerfect group failed miserably to get Java implementations of their office suites out the door, although IBM did offer the Kona eSuite dumbed-down versions of Lotus SmartSuite in Java – and has this month killed them off. StarOffice would appear be the least of Sueltz’s problems, although the press, analysts and Wall Street will be quick to jump on her back if Sun doesn’t make some quick hay with StarOffice and get the Java versions out in a reasonable form and soon. This will happen more because the issues in the Unix and Java markets are more complex, and people seize on the easy things when they want to attack.
While Sueltz is unquestionably more qualified than any candidate Sun could round up on such short notice to take over its software business, it is the fact that she has not been in charge of a major software platform at IBM, such as OS/400 or AIX, that could give Sun and its customers some concern. Sun’s aggregate software business (not counting the iPlanet partnership with AOL and Netscape) will probably represent roughly 9% of Sun’s $12bn in revenues this year. Between $800m and $850m of that will come from Solaris and related products, with Java and related products only getting about $150m to $200m. Java is only incidental to Sun’s $4bn server business and its related $3bn storage business until more customers and independent software developers start using Java rather than other languages such as C++, Cobol, RPG, and Visual Basic to code their applications. Moreover, the vast majority of another $2bn in services revenues and a good portion of its $1.2bn in microprocessor revenues can also be attributed to Solaris servers and workstations, either directly or indirectly. So it looks like more than $10bn out of $12bn in revenues at Sun are really being driven by Solaris, and no one in the trade press or in the analyst community is talking about that because they are more interested in what Sueltz will do about Java (see the separate story).
Java makes lots of headlines, but it is Solaris that makes all the money. So it is probably a good guess that Sueltz is not going to use her OS/2 marketing skills and will somehow figure out other ways to program new features for Solaris that give it even more leverage against IBM mainframes and AS/400s, HP 9000s and Windows NT servers. Just by way of comparison, IBM’s Software Division is about the same size as all of Sun, and standing by itself, IBM’s software business is many times more profitable (in terms of profits as a percentage of revenue) than either IBM or Sun at the corporate level.
Sueltz is also going to have to contend with Linux in a more meaningful way than porting Linux to Sun iron and seeing what develops, or adding Linux APIs and ABIs to Solaris so that people who outgrow PC iron can move their Linux applications up to a real server like a Sun UltraSparc Enterprise 6500 or 10000. While these approaches, which are common among the major server operating system vendors, look fine on paper, very few customers who start out on a PC server or workstation running Linux are going to grow so fast that Intel and its partners can’t supply iron for their needs. While no one has any better answers than this when it comes to taking the wind out of Linux’s sails, supporting it within Solaris or any other Unix or NT will not be sufficient to negate the force of Linux, any more than Sun’s ignoring and berating of NT has made it go away despite the fact that NT is a dreadfully buggy and unreliable piece of bloatware compared to Unix or Linux.
