Oracle's Open Source Spin -- Redmond Developer News

More Redmond Developer Network Sites >>> VisualStudioMagazine.com | ADTmag.com | Events

Data Driver

Oracle's Open Source Spin

I attended the Linux/Open Source on Wall Street conference yesterday in New York and was intrigued to hear Monica Kumar, a senior director for open source product marketing at Oracle, talk up her company's competence in helping organizations move to Linux and work in development environments such as Eclipse. "Our customers are adopting open source and are demanding our technologies support open source," Kumar said during the opening panel session.

No surprise there, after all Oracle has championed Linux for about a decade and the company typically releases new software first on Linux, then other platforms. But I couldn't help but notice the irony in her remarks, when it came to the obvious omission of open source databases. After her session I asked her about that, particularly in the wake of Sun Microsystems' acquisition of open source database provider MySQL.

"We haven't seen our customers asking for open source databases," she told me. "Not many customers are interested in looking into the code and mucking around with it, and making changes to it. All they care about is 'give me the best support, give me the lowest price of entry'." For that Kumar pointed to Oracle Express.

While MySQL accounts for a small sliver of the overall enterprise database market and poses little short-term threat to the established database platforms, it's also subtraction by addition, since Oracle until now was Sun's preferred database when going to market with its solutions.

For now, Oracle's database business is holding its own. While Oracle's overall revenues for the quarter that ended March 15 failed to meet Wall Street expectations, it is interesting to note that its database business for the period actually grew 20 percent, up from 17 percent a year ago. However that didn't do much for the company's stock price last week, when the company reported earnings were up 30 percent on lower than anticipated revenue, which sent shares tumbling 8 percent. Its stock has since started to bounce back.

So what's your take? Do enterprise developers want to muck inside the code of a database or will the DBMS remain the platform of choice? Send me e-mail or post a comment to this blog.

Posted by Jeffrey Schwartz on 04/02/2008 at 4:01 PM

Share this Page

Reader Comments:

Sat, Apr 5, 2008 Andrey Russia

What "customers interested in looking into the code"? Code is not for "customers", period. Monica is telling the truth, but the truth similar to "even the most safety concerned auto buyers seldom crash test their newly acquired cars". Also true, but smells idiotic. Why does Monica, as well as lots of other people, risk to look like idiots while writing about anything open source?

I guess there are 2 reasons for that. First, they want to exercise their NLP skills and link something negative to something open. Second, they want to hide the real picture.

The fundamental feature of open source is that it is THE best tool for a cooperative effort, at least in the software domain. Cooperation with contracts, outsourcing and licenses is also possible, just not so efficient. It turns out that one can get more with cooperation than alone, trivial, right?

All OSS projects evolve according to THE pattern. They are started by people who cooperate solving their own problems. Next, other people join and package for "users". They also do OSS solving their, very different, problems. Finally, areas where cooperation is not necessary or inefficient are discovered, most often in packaging, and closed solutions emerge. If a project originates differently, it drifts to the pattern, like Qt or Java.

While this is happening, lots of mis leaded "users" are attracted by the zero price tag and generate lots of bla-bla, since they are divided into 2 groups doomed to never understand each other: those, who, by share luck or limited expectations, are happy and those who are not.

A perfect example of that is "installing software in Linux is hard". If you cannot compile, you belong to neither of the two cooperating groups and should be happy with that which your distribution vendor has packaged for you.

So, there are 5 distinct groups of people: developers (including, as it is much more common with databases, testers), packagers, paying users, happy nonpaying users, and suffering nonpaying users. Great confusion happens when a statement lacking explicit reference to the group in question is made.

For example, there is a recent finding that OSS database adoption is "wide but shallow". Most likely, this means that there are many happy nonpaying customers. Never expected anything different. How much deeper are the projects supported by the database vendor? Interesting but unknown.

Jim from NY sees himself in the suffering nonpaying group. This is easy to understand since OSS databases still lack essential enterprise features. However, this does not mean that "The DBMS will remain the platform of choice". The mature ecosystem will be like any other OSS ecosystem. Oracle will offer a free OSS database just like Red Hat offers Fedora. Other companies and consultants will sell support for it, but Oracle will be the brand, exactly like Red Hat. Oracle will have less margin in the high end, but more users overall.

This will not happen tomorrow, but Express, the first step in this direction, is already here. I guess Monica is not sure that the net effect will mean more profit for Oracle and just wanted to postpone the inevitable. Thus, no mentioning of the OSS databases.

Why did I write that? I feel that the "Do enterprise developers want to muck inside the code of a database or will the DBMS remain the platform of choice?" question could be formulated better.

Thu, Apr 3, 2008 Jim NY

The DBMS will remain the platform of choice.

Mucking inside the code of a database might be fine for open source supporters and hobbyists (and potentially good for your resume), but deploying software to end users demands a mission critical DBMS, not just a "stable" release from open source.

We use MySQL internally, but not for mission critical operations. The number of features (important to us) that are missing in MySQL is staggering. We don't have the time to develop the missing features in MySQL (or wait until someone else does); we devote our time and effort to our customers and what they want today. If we wanted to deploy our products against MySQL, we'd be constantly programming "around" missing features that have been available in Oracle/DB2/SQL Server for quite some time. We have far better things to do.

We have deployed our product in hundreds of locations supporting thousands of users -- number of database problems? ZERO. Our customers have more problems patching their OS (Windows or Linux) than with their DBMS of choice and our software.