Prognosticating about Oracle’s Sun Takeover

Monday’s announcement of Oracle’s takeover of Sun Microsystems was nothing short of a mild earthquake in the world of computing. Our hometown of Boston was celebrating Patriots’ Day and the Boston Marathon was keeping us all riveted. So after three days of mulling, thinking and tossing around ideas of what this all means, here’s my take of where this may lead.

Until Monday, Oracle and Sun manifested very different business philosophies. Oracle is a company that sells database server software. Love it or hate it, it is the dominant player in this key sector and luckily for Oracle, its strategy remained relatively simple:

1. Sell licenses for database software
2. Create products complementing the database
3. Sell support
4. Buy products and rivals that are in a weakened position
5. Foster the creation of a thriving ecosystem around product lines
6. Buy said ecosystem products when times are rough (like now)

Of course a lot of business smarts did not hurt along the way and thus far. No matter what Microsoft ,IBM and many in the open source community throw at it – Oracle placidly remains a number one. IT folks will grumble at costs, but will go on betting the farm on Oracle. Oracle remains the choice when storing mission critical data, and to a lesser degree helps the CIO sleep quite well at night, thank you. Oracle is helped by the reality that once you have huge amounts of data in a database and the investment is in place, it is quite difficult to switch to say, cheaper alternatives. Oh, and Oracle runs on Windows, Linux, AIX, anything so whenever you need them, they can sell that license. “We’ll play with whatever you got”, they say.

Sun is another story.

As many have noted too frequently, Sun never recovered from the dot com bust earlier this decade. The Java language may be the most important programming language out there today, and Java-based technologies do power myriad enterprises, but this gift to mankind never really made Sun much money. Java was supposed to drive folks to buy Sun’s expensive and possibly superior servers. Sadly for Sun, the fact that it was an open technology running on every platform made it popular as much as it made Sun servers not required. Then there was Linux that was easier to use and cheaper (eh, free), unlike Sun’s Solaris Unix. operating system That hurt too. Fewer licenses, less money.

Seeking fresh thinking, founder Scott Mcnealy left the Sun CEO post for Jonathan Schwartz, who took openness and freedom to new levels. Sun embraced everything and anything open source. Sun even moved to commodity architectures to try and sell more servers. Instead of sticking with its fast, small, green, smart Sparc processors. Sun, gulp, even embraced AMD processors. Pitting itself against commodity Dell, IBM and HP servers was expectedly difficult. Many remained loyal to its industrial-strength Sparc servers, but growth was not happening there either. To help Solaris gain or regain adoption, the operating system was made open source and free. Concurrently there was an attempt to show the world Sun loved Linux. Sorta. Schwartz did not stop there and pushed on making Java itself open source. Defining the language is a process open to virtually anyone to influence.

And then, last year, Sun bought MySQL. The wildly popular free, open source database now had a home and a parent. A parent who could provide the enterprise-grade support that enterprise-grade clients with deep pockets seek. Legitimacy was finally within reach for an application that less than a decade ago was regarded as a hobbyist toy. The master plan was unfolding:. Sun was going to be a services company (oh yeah, they sold servers) supporting free, open source technologies (and some proprietary add-ons made by Sun that made them sweeter). Put together, Sun really held a lot of the open source in its portfolio. Now Oracle, a company that free is not really a big priority for, a company that open and nice rarely appear in its vocabulary (unless they help move database licenses) – holds that striking portfolio.

So what is Larry Ellisson, Oracle’s iconic leader, to do with all of this Sun goodness?

If the past helps us predict anything about Oracle, the company lets many of its takeover target go on as they would have been, tweaking management, learning the business, getting the clients, (selling database licenses, did we say they are good doing that?) and often, making them more successful. Clients of competing products that were taken over by Oracle were given some relief but gradually moved to the core Oracle technologies if they wanted to remain friends.

The most influential far-reaching item in the grab bag is Java. Java has been good to Oracle and the world as a whole. Larry LOVES Java. Oracle at some point even baked Java into the insides of its namesake database. Not an awesome idea, but it moved licenses. Oracle owns a Java software powerhouse in BEA (maker of the popular Weblogic Java Enterprise Edition application server) and was not a slouch by any stretch of the imagination even before that takeover last year. Being invested in the technology, it is nice and safe to own it. Seems like Java works for Oracle and for now, Oracle will keep it going with mostly hands off.

Sun’s services organization and still considerable client base are also in a good place. Lots of clients run Oracle on Sun hardware (the companies were chummy like, forever) and adding boots on the ground will help Oracle pose more of a threat to IBM’s army of consultants. So big win and major synergy there.

MySQL is where I am getting quite confused. Never mind Sun, MySQL AB, the Swedish company Sun took over last year was manned by developers and staffers quite avidly anti-Oracle. MySQL was becoming THE open source alternative to established (read Oracle) database software, with a variety of large and larger companies betting on MySQL to avoid the Oracle license toll. No question Larry Ellison likes owning the main open source adversary; question is whether MySQL’s development team will like it too. With this economy, and after the hard years MySQL had to prove itself, I would say many MySQL developers will just deal.

Happy so far. Now it gets a bit foggy: Solaris, Sun hardware and within hardware, the Sparc architecture.

Is Solaris necessary in a world that moves faster towards free Linux (and FreeBSD) operating systems? There is a considerable install base paying handsome support fees, often because they WANT a company-driven operating system. IBM’s AIX looks OK but off the mainframe it is not very shiny. HP-UX? who? So Solaris is an alternative; but development is not thriving and maintaining an OS is not cheap. The future is iffy for Solaris.

When it comes to hardware, does Oracle want to go head to head against HP, Dell, IBM, new entrant Cisco and others in the server market? Oracle is extremely cosy with hardware makers – Oracle needs horsepower and disk space – they make it. Even IBM does not mind being in bed with Oracle. Running SAP on Oracle-owned Sun should feel precarious, but that’s like punishment for not choosing Oracle for ERP, and so be it.

This position of power may just allow Oracle to compete with those companies that need its software, at least in the short term. And saying that ‘our servers work best with our software’ is not a promise companies scoff at. Now if Oracle could do something about those pesky high prices. Hardware may stay, but what about its central component – the highly regarded Sparc processor?

Sparc was always ahead of its time. When the world (aka Intel) was just thinking of moving to 32-bit processing, Sun was like, ages into 64-bit realms. They are still ahead of the curve, but Intel processors and the world of 64-bits seem to have people quite content mostly. Making and designing haute couture CPUs is very expensive. You look super cool using it, but without vast production to defray associated costs make it difficult to compete on merits alone. Hence Sun’s AMD fling and $1,000 servers. Sun is emotional about Sparc. I smell spin-off, just like Motorola’s creation of FreeScale.

Finally, Oracle does not mind software duplication as long as it, well, moves database licenses. Other licenses are nice too. So adding Sun’s open source GlassFish and other software servers may be just OK for a while at least. Getting some creds from the open source community is nothing Oracle cares for at all, but keeping free initiatives going does not hurt much.

Oracle is a brilliantly-run company. Swooping in after IBM did not pull the trigger seems like just another brilliant move. Naturally, one with plenty of challenges, but Oracle knows how to throw its weight around just right, and yeah, sell licenses. Not that Sun was necessary, but Oracle now owns an even more massive swath of the software landscape than ever before. IBM more than any company should feel more threatened than ever before. Uncle Larry’s coming after you.