Why the Oracle Acquisition of Sun Gives Me a Java Headache

And why the DOJ is taking a hard second look before approving

In my 25 years in the technology industry, I have seen many once-important companies disappear. Remember Bull or Honeywell or Digital Equipment? Now the sun is setting on Sun. What makes this more remarkable is that Sun wasn't stuck in an old-technology paradigm like minicomputers, when the world all moved to servers and personal computers. It remained relevant from a technology standpoint. Sun pioneered Java, the open source software that is embedded in everything from our cell phones, DVD players and cars to weapons systems. Yet, on April 20, 2009, Oracle announced its plan to purchase Sun for approximately $7.4 billion. It is inevitable that Sun will be bought; the current management has not been able to find a profitable path forward for it to remain viable and independent. But the sale to Oracle raises all sorts of issues that are being talked about at the margin but not in the mainstream. At this moment, the proposed transaction is being quietly reviewed by antitrust regulators around the world. They are taking a look at the issues that surround this combination, such as raising prices and reducing innovation for database software and maintenance services, limiting choice for customers via a lock-in to Oracle's stack of products and expensive maintenance programs, and, most importantly, putting the future of Java at risk.

In the mid-1990s, I worked with the Sun team on the Java launch. It was a skunk works project that we leaked to the San Jose Mercury News in an effort to get the Sun management to keep it alive. When they woke up to read about it on the front page of their morning paper with their cup of java, Java was here to stay. And stay it did, quickly becoming a developer's paradise for creating applications for the emerging Internet and mobile and video device market and then just about everything else possible. The beauty of Java was that it was open, controlled not by Sun, but by a standards committee of industry interests. Sun never found a way to make money on Java itself, but it enhanced its stock price considerably -- giving it the balance sheet to grow its traditional business -- and today, it certainly is the crown jewel that Larry Ellison so desires.

In Ellison's hands, Java may slowly move from its open source position to become more and more a part of Oracle's software stack -- the suite of software programs it sells. Oracle has invested a great deal of time and money to bring Java into its products. But that pales compared to the $7.4 billion it will invest in the Sun deal-- what will Oracle want out of Java to justify that level of investment? Well, it could want to modify Java to work better with its own products, for a start. Or it could pursue a strategy of selling Java like iTunes, one download at a time. The millions of Java developers globally should be more than a little worried about this particular combination of Sun with Oracle. This is their potential livelihood at stake.

More directly, the livelihoods of 10,000-15,000 Sun employees are in jeopardy with this deal, as Oracle had promised that it would ensure this deal is accretive, which from their past acquisitions, means steep job cuts (a relatively recent example is PeopleSoft, where Oracle laid off nearly 50% of their employees post-merger). One thing Oracle has shown is that it is the master of the deal -- it has acquired 73 companies, 52 since 2005. It knows how to buy, gut, integrate and maximize. Oracle is able to consolidate its market position this way and quickly increase prices. When it purchased BEA, it turned around and raised prices for BEA products nearly 50%. Oracle can do this because through all these acquisitions it has quickly become the world's most dominant enterprise software company. It has built a stack of products that enable it to offer customers everything they want, but also no choice at all as they lock customers into their full product line through maintenance contracts. Where Oracle really makes its money is on maintenance contracts.

Sun may be just the latest, but it is the largest acquisition Oracle has pursued. The new administration has signaled through its Assistant Attorney General for Antitrust that it will take a more aggressive stance on regulating competition. On Friday, the DOJ submitted a second request on the Oracle/Sun acquisition, which suggests that there is concern this with this deal merits a strong second look. Oracle has gone unchecked in amassing its dominant position to date through 73 acquisitions. But with Sun -- and the concerns it raises about the future of Java, and general market dominance-- it may finally be time to take a hard look at Oracle. Java must remain open and interoperable. The DOJ seems to understand that Java is a unique and critical part of the technology engine that runs our lives, and its future must be ensured. Is this acquisition in the best interest of Sun, its shareholders, customers, the Java community, and the industry at large?

Comments

[roadrider]

Well, Oracle, Sun and Java ALL give me headaches! A pox on all of them.

I only feel bad for the workers who may lose their jobs.

Java is a nice but highly overrated tool. It's basically an amalgam of concepts that existed well before its inception, a syntax borrowed from C++ and a lot of hubris among its evangelists. Even as a developer with lengthy experience in other object-oriented languages (most of which the Java inventors borrowed from) I have found it near impossible to be considered for Java positions because of the silly attitude that "if you ain't done it with Java, you ain't done it". Excuse me if I don't mourn the possible demise of Java and it's smug adherents (it's a programming tool guys, not a religion).

[KillTheMessenger]

Sadly, Java is the modern equivalent of Cobol. It is huge, poorly designed and anything but portable. Launching a Java application on the most ubiquitous platform in the known universe still feels like loading a deck of cards into a card reader.

Why? Why did nobody at Sun develop a clean Java launcher/environment in almost 20 years of R&D? What's so hard about that?

And what's up with the GUI? AWT/Swing weren't good for much when they hit the light, and they haven't come close to being competitive with other GUIs since.

There are a lot of small languages out there that have a much higher productivity than Sun's Java. Python comes to mind but that's just my personal reference. I am sure Lua and Ruby fans will say the very same about their language of choice.

If Java were the only thing that is important about Sun (and it might very well be), then the company has no right to be alive.

To be honest... Oracle can't do anything to Sun that Sun hasn't done (at least technologically) to itself, already.

[DennisByron]

This Huffington Post article glosses over some information technology (IT) history and misses the point on a lot of present-day IT trends as well. How about some fact checking, Huffington Post? . Re-assure your readers that the Sun Microsystems IT product called Java does not run all of the world’s coffee machines, and that Oracle’s Larry Ellison will not ruin their Sunday mornings--as they wait for their espresso to steam before reading the NYT. And the Obama administration Department of Justice (DoJ) doesn’t need to aggressively intervene in Oracle’s planned acquisition of Sun (and despite what this post says, is probably not doing so)..

Despite the article's revisionism vis a vis Digital Equipment, Sun was also stuck in an old paradigm. The only difference between the two is that Sun chose servers and DECstuck with minis at the point in time when the IT world, led by Microsoft, Oracle and SAP, moved past both.

And Java was not open source during the mid 1990s as the author implies. Java was not open sourced until a late-in-the-game Hail Mary pass thrown by new Sun management in 2006-2008. Perhaps someone from old Sun management can weigh in on the Skunkswork story. It's probably as misleading as the rest of the above facts and opinion..

[heidisin]

I am having a hard time tying any of your criticisms back to what was actually written in the blog. There is nothing about coffee machines. Nor does it say that Java was open source from the beginning. Nor is there anything about DEC. So before you lambast on fact checking and suggest somehting is misleading, you might want to read the piece carefully and get your own facts correct.

[KillTheMessenger]

Java should have been open source from the start. That it wasn't basically killed it. Java as a business never took off and Java as a technology is stuck in the very same dusty drawer it was conceived in.

Please compare the available libraries for e.g. Python with those you can get for Java. Unless you are doing business logic, the Java universe is a close to perfect vacuum.

[gfcorbett]

Great analysis! And it is very true. Ellison bought Sun precisely for Java. And Sun has already initiated the process of dumping employees.

Time will tell. Let's hope other software companies can provide a level playing field for customers, though I have my doubts.