Oracle’s Gravity Pulls in Sun

A little more than a year ago, when Sun acquired MySQL for US$1 billion, no one would have thought that one of the biggest and the most innovative companies would itself be bought out within a year. Things did not build up in just those 12 months. Sun had not been doing too well since the dotcom burst. The company kept operating due to its deep pockets, which were being drained steadily due to recurring losses. But then came a time when this could not continue. The company started looking for suitors.

“I have known for quite some time that Sun’s position was unsustainable and untenable. Somebody was going to buy them. The fact that none of the companies I could imagine buying Sun would find Sun a particularly good fit, did not trump the fact that Sun’s shrinking revenues and relevance meant that the sooner they sold out the better. When the deal was announced, the acquiring company might have been a surprise, but the fact of the acquisition was inevitable,” says Michael Tiemann, co-founder of Cygnus Solutions, a company founded in 1989 to provide commercial support for free software that was later merged with Red Hat in 1999.

“Oracle as the acquirer was unexpected though. There were a lot of different feelings about the acquisition. Overall it did not thrill anybody. It was a cold calculating move from Oracle that did not add much to the industry as a whole. If IBM had been the acquirer, people would have been less surprised, but an IBM acquisition would not have been exciting news too,” notes Laurent Lachal, director, open source research, Ovum, a software and IT services consulting firm.

But how different would it have been had IBM acquired Sun? “IBM has a systems and services culture that is a better fit for selling systems, while Oracle is strictly a software culture based strongly on the sale of software licences and maintenance services. I would personally be surprised if Oracle keeps the Sun hardware business,” says Tony Wasserman, director, software management program, Carnegie Mellon University. Tiemann echoes this sentiment, “It is impossible to say for sure [what would have happened if IBM instead of Oracle had acquired Sun], but I think that IBM has demonstrated a number of strengths where Sun was weak, and vice versa. So it would have been a more logical and complementary marriage.”

“From the Linux Foundation’s view, either Oracle or IBM represented positive outcomes for Linux. Both these companies have robust pan-organisational Linux initiatives,” noted Jim Zemlin, executive director, Linux Foundation.

Since the IBM-Sun deal did not materialise, this is now an irrelevant point. Oracle has changed the game, and now, there is an equally-matched heavyweight that can compete with IBM in almost all the market segments where the Big Blue dominates. Well, competition and choice are ultimately beneficial from the viewpoint of the customer, who’s more interested in the free market dynamics.

On the other hand, HP may be a bit worried, as it will now have to deal with Oracle as a partner for database businesses, as well as a competitor in the hardware and operating system space.

The hardware game

The core factor of this deal is: while Oracle is a software company, Sun was primarily a hardware company. Hardware is a low-margin business, whereas Oracle’s turf till now has only been the fat-margin software business. There are speculations that Oracle will sell off Sun’s hardware business, which will accomplish two things for Oracle: end the worries of its OEM partners, and get rid of the investment-hungry and R&D-intensive hardware business. But Sun’s balance sheets show that it was the hardware business that was making money for the company. In the first quarter of 2009, Sun reported an 83 per cent year-on-year billings growth in its Solaris-based units. Sun reported a 12 per cent year-over-year revenue growth in the emerging markets region, with India, Latin America, a combined Russia, Middle East and Africa growing by double digits, year-on-year.

Now, who would want to let go of such a profitable business? Certainly not Larry Ellison. The Oracle chief recently ended the speculation of a sell-off. He was quoted in a Reuters interview as saying, “We are definitely not going to exit the hardware business. If a company designs both hardware and software, it can build much better systems than if they only design the software. That’s why Apple’s iPhone is so much better than Microsoft phones. We think designing our own chips is very, very important. Even Apple is designing its own chips these days.” [Guess he can’t resist having a dig at his favourite rival, Microsoft.]

But Oracle has no prior experience in hardware except for Network Computers (NC), which could be called a failure. “Fortunately for them, NC was so small an effort that writing it off cost them almost nothing. Sun was very much a hardware company, and it’s a surprise to see Oracle wanting to make such a bold move,” says Tiemann of Cygnus Solutions.
But sticking with the hardware business would mean walking a tight rope. HP, Dell and IBM may be wary of the situation. Well, Oracle might be happy to cause some worry to IBM, which has always been a competitor with an extra edge. Ownership of Java was one of the reasons Oracle acquired Sun, as Ellison claimed. IBM’s stack uses a lot of Java, and Oracle just can’t let it go to IBM so easily. It seems like Ellison made the right move at the right time.

Only time will tell what strategy Oracle will adopt in terms of the hardware business—I’m sure it would not want HP, Dell and the rest to go and smoke in Microsoft’s camp. The fact is that today’s customers are smart. They know what serves their interests in the long term. Customers will now see more value in going to Oracle, which will force others to chalk out their strategy very carefully. But, nothing can be said unless the final details of the deal are out. I guess the whole industry is awaiting that moment.

All eggs in the same basket

Apart from hardware, a lot of equations are going to be recalculated in the field of software. One of the most interesting facts is that Oracle now owns one of the most popular operating systems its databases were running on—Sun Solaris. But at the same time, it also has its own Oracle Enterprise Linux (OEL), which the database giant proudly admits is virtually a copy of Red Hat Enterprise Linux (RHEL), minus the RH trademark.

How will the two competing operating systems live and grow within the same company? What kind of commitment does Oracle have towards Linux (besides its own offering, OEL) and it being the preferred OS to run its database?

The fact is that Oracle is one of the biggest customers of GNU/Linux, as well as one of the biggest contributors to the development of the Linux kernel and other GNU tools. According to Zemlin, “…they are one of the largest consumers of Linux in the world in their data centres; this is a company that is very pro-Linux. Purchasing one of Linux’s competitors (Solaris) certainly bodes well for Linux.”

Oracle may be the best friend open source and Linux will ever have. Shane Owenby, director, Linux & Open Source, Asia Pacific, Oracle, earlier told us in an interview, “Our developers internally develop on Linux, so there’s quite a lot of momentum there. There is something called ‘Oracle on Demand’, where Oracle runs an Oracle customer’s Oracle software for them—we run that on Linux. So, we trust Linux that much.”

Greg Kroah-Hartman, the current Linux kernel maintainer for the stable branch, admits, “They have contributed a lot. See the articles about Linux kernel development on LWN.net for every kernel release, for the details on exactly how many contributions they have made, and where they are.”

But will the acquisition of Solaris affect the development of Linux? Chris Mason, a developer from Oracle’s Linux kernel team and the primary author of Btrfs (the filesystem poised to replace ext3 as the preferred filesystem on Linux), notes, “Oracle has been heavily using and contributing to Linux for a long time, and all of our public statements around Sun have talked about how Linux is still important to us. In terms of projects done by people on my Linux Kernel team, we don’t expect any decrease in our contributions.”

“I expect Oracle’s work on Btrfs to remain unchanged. The work the Oracle kernel team does is top-notch and is likely to expand,” Zemlin hopes.

There is another possibility. Since Oracle now owns Solaris it can change
its licence to make it compatible with Linux, allowing code to be shared between the two projects. This way it will be able to port missing features of one operating system to another which, in turn, will eventually improve both the products. Zemlin’s statement seems to echo this, “We would love to see Solaris under a GPLv2 licence… This would be a win for users of both platforms.” [The open source Solaris project, OpenSolaris, is licensed under CDDL, a licence which is incompatible with the GPLv2 licence that Linux uses.]

Whether Oracle relicenses Solaris or not, either way it’s a win-win for Oracle, at least. There seems to be no valid reason for Oracle to decrease its contribution to Linux, which has become very powerful and popular in the server segment.

So, what now remains unanswered is the future of the core open source technologies that Sun had developed or acquired over time. Some of the technologies seem to compete with Oracle’s line of products. Will Oracle sell them off, kill them or polish them to enter the markets it is absent in?

The day after tomorrow for MySQL

MySQL is one product that almost everyone is talking about because it seems to compete with Oracle’s database business. MySQL is arguably the most widely-used database for Web development and Web apps, and therefore has kept a sizable chunk of the market away from Oracle. This means, the acquisition could either spell the slow death of the open source database or fill its lungs with fresh air.

No doubt it’s been anything but a bed of roses for MySQL after being earlier acquired by Sun. Most of the key developers, including MySQL founders David Axmark and Michael “Monty” Widenius, left. Tiemann has the same view, “A year after Sun’s acquisition of MySQL, the founders, CEO, and several lead developers had all left without any good things to say about Sun. To me, this hurt the otherwise stellar brand of MySQL. Fortunately, the open source community has many fine open source database projects, and whatever damage may have been done to the MySQL brand, an equal amount of good came to the other projects. Thus, there was no net loss to the open source community, though there was a competitive shift among the various projects.”

Monty came out with a new project and is currently working on Maria DB. However, he has no doubt about the quality of MySQL, and even has a few words of praise for Sun. He also says that Maria DB is less of a fork and more of a branch. He says that the core competency of MySQL is “…the people [behind it]! Sun has excellent support, sales and training people in the MySQL area. In the development part, they used to have excellent developers but a large portion have now left or are about to leave. Sun has also a great website with a lot of information about MySQL, and is the main source for the MySQL source code and binaries.”

This sort of praise for Sun is surprising, coming from someone who once wrote on his blog, “The main reason for leaving was that I am not satisfied with the way the MySQL server has been developed… In particular, I would have liked to see the server development moved to a true open development environment that would encourage outside participation and without any need of differentiation on the source code. Sun has been considering opening up the server development, but the pace has been too slow.”

But things have changed now, or rather those at the helm have changed. MySQL now belongs to a company that is the world’s largest database vendor. Will things change for the good? “Not really,” believes Axmark, “Since the suffering was mostly from being integrated in a big company. And this means another integration.”

Another factor that will decide the fate of the popular open source database is what MySQL means to Oracle—is it perceived as a threat or a benefit? Axmark thinks that MySQL is both a threat as well as a benefit. “But mostly a complement, since Oracle is not that big in the online/Web market. Oracle is aimed at OLTP [online transaction processing], and MySQL at Web-related stuff.”

Monty agrees, “MySQL is more suitable for Web developers or for applications where the demands change rapidly. It’s also more suitable for new applications where you want to quickly get things going without having to pay a lot (of time or money) to get the database running.”

All this is from the point of view of the MySQL founders. No one really knows what those at Oracle are thinking. There’s been no word about MySQL’s future in Oracle’s statements. There are several possibilities, which may determine what Oracle could do with MySQL [see the side box on this page].

If Oracle does put MySQL in the ‘benefit’ category, it’s good news. Oracle is a profit-driven, high-speed train, and can achieve what Sun couldn’t—monetise MySQL. This will happen only if the company sees some benefit from MySQL. Given the installation base MySQL has, there is no doubt Oracle will get access to that market. Also, it will now have the entire developer base working around MySQL. Monty points out the benefits Oracle will get by merely owning the MySQL brand: “Owning the trademark of the most used open source client server database gives a lot of value. Oracle can use the MySQL website to spread their message.”

But will the core developers return to their much loved project and join Oracle? “We would be happy to cooperate very closely with Oracle but we are not likely to join. After working for a big company like Sun, we enjoy the extra freedom and benefits we get in a smaller company. Another thing is that I don’t think Oracle can match the benefits of the hacking business model we have been using. The main problem is that the critical people needed to drive the MySQL project may not be staying around to give Oracle a chance to improve. I see it as my main job to ensure that the developer community is not split up and we keep the core engineers working on MySQL. If Oracle plays nice with these people, then things will improve dramatically with MySQL development,” notes Monty.

So that’s back to Square One! Everything will depend on what Oracle does with MySQL. In a scenario like this, MySQL users must be sweating over concerns about the future of a product their businesses depend on.

However, this is where the beauty of open source comes into the picture. No one owns the project, but only the brand. “The whole idea with FLOSS software is that it doesn’t matter who owns it even remotely as much as it does for closed-source software. So they [customers and users] can continue to use MySQL whatever happens. And there are still tons of good MySQL support people at Sun, so there is no problem buying support from Sun,” says Axmark.

And just in case Oracle doesn’t play nice, you can always fork the project. “With the current licence, there is no problem to fork. There are already several existing forks,” says Monty. “The newest ‘branch’ is MariaDB that I am working on. We already have many of the core MySQL developers working on this so we have the competence to do this properly.”

However, many feel that though forking might be easy, maintaining the fork and ensuring its popularity is the biggest challenge. Also, a customer might be wary of going for an unknown project instead of a well-known one. So, how can a fork be successful? Monty argues, “The brand popularity gives some protection for Oracle, but it has been shown before in the open source world that name changes of products can be done quickly. Just look at how quickly Firefox was adopted. The reason for being able to change quickly is that MySQL is a product mainly used by developers. Developers are quick to find and go with new trends, and if one fork gets a lot of attraction from the community it can take over very quickly.”

If Oracle realises the benefits of MySQL and boosts its development to capture new markets, it would be the best for the community and the customers. In case it fails to do so, customers don’t have to worry; there will always be a fork ready.

OpenOffice.org: The road to independence

OpenOffice.org (OOo) is one product that may not seem to fit in Oracle’s game plan, and thus might have a hard time under the database giant’s umbrella. John McCreesh, marketing project lead, OpenOffice.org, has this to say, “IBM already has its own commercial OpenOffice.org derivative, so it has a clear financial interest in keeping the project going as long as there is an active body of developers working on it. IBM is better known than Oracle in OpenOffice.org circles.”

But, an office suite is a product that is an absolute, basic requirement for any enterprise. It also ensures penetration into new segments. Heavy deployments at big enterprises as well as government agencies mean huge revenues. This has been Microsoft’s bread and butter for ages. IBM and Novell both realise this and that’s why they have their own OOo derivatives—Lotus Symphony and GO-oo, respectively.

Oracle knows this only too well. An office suite was one important component missing from making Oracle the Apple of the enterprise world. So, the estimation is that Oracle might try to optimise OOo for its Oracle Collaborative and e-Business suites. The level of integration could go as high as pulling data from the Oracle backend database into OOo components, thus offering newer solutions to its customers.

This may seem a bit out of focus, but Microsoft is Oracle’s archrival and an office suite is an important part of Microsoft’s backbone. Oracle may not let go of an opportunity to annoy Microsoft and dig into its market. Apple also depends heavily on third-party office suite applications, so a polished version will only mean constant
flow of money for Oracle. If Oracle makes a mistake here, OOo will be forked by the community and be adopted by either IBM or Novell, and the potential market will slip from Oracle’s hands. But, considering the support Oracle offered for ODF (Open Document Format—an ISO standard for office document formats, and the default used in OOo), it is hard to believe that Oracle will let go of a potentially lucrative market.
“Oracle is a big player in the enterprise software market, and just like Microsoft, once it gets one product to a customer, it will try and use that installation as a lever to get other products in. Up till now, those at Oracle haven’t made a serious attack on the desktop space. Maybe OpenOffice.org will give them an incentive to do so,” says McCreesh.

He also warns that if something goes wrong at Oracle, the OOo community as a whole will not hesitate in making the office suite an independent entity. He points out, “There have been discussions around creating a foundation for the independent development of OOo. I think that’s a real possibility if Oracle is not interested in OpenOffice.org or withdraws the sponsorship that Sun Microsystems used to provide.”

Well, to twist around an old fashioned greeting a little, “May the fork be with you!”

Hitting at Microsoft

Customers are looking for heterogeneous environments, and they find Red Hat and Oracle to be a better choice than Microsoft. While Red Hat is the leading open source company, Oracle offers a mix. The latter will be able to address any query that a customer might have regarding the entire stack—from the operating system layer to the application layer.

The buyout of Sun will put Oracle in a much stronger position against its ‘favourite’ rival Microsoft. Let’s face it—Ellison never misses an opportunity to punch Microsoft. This is the first time he will have all the required arrows in his quiver to make a dent in Microsoft’s armour.

The company is now a complete vendor—it has its own hardware, operating systems to run on those hardware, middleware, the mother Java, the most popular databases (whether open source or proprietary), an office suite, and many more technologies. Seems like it’s becoming a multi-polar world, with Microsoft, IBM and Oracle in the ring.

And let’s face it, most customers would prefer going for Oracle’s stack than Microsoft’s. The fact is that many companies are migrating to GNU/Linux and Microsoft has been doing nothing but token lip service regarding its uncertain and dubious interoperability efforts. In the case of Oracle, everything will be well-optimised and well supported. If customers face any problems—from operating system to database, or office suite to middleware, all they have to do is dial Oracle.

This seems to be an end to the dominant, dictatorial Microsoft era. IBM, on the other hand, will also have a giant to compete with. The three players will keep a check on each other. The result? The community and the customer will win.