db4o Developer Community

What a beautiful documentation of "Innovator's Dilemma" from Pointbase, unsuccessful makers of an embeddable Java relational database. In an executive whitepaper they try to spread FUD about open source databases.

This is really the best marketing for us, because it is so false that I can only ask any user and customer to read it and laugh about it as much as I did.

Here are the main 2 arguments:

Argument #1 - "Open Source is Technologically Behind"

Open source databases have not yet reached the level of sophistication of other open source technologies and developers using them must be aware of certain inherent weaknesses when assessing their usability.

First, the nature of open source databases is that they cannot exceed a certain level of complexity. To do so would require an organized, concentrated effort that cannot be carried out via the uncoordinated work of disparate open source technology contributors, with each rewriting source code to achieve different personal goals. An open source technology must by definition accommodate the lowest common denominator of its contributors. In fact, many open source contributors are not highly qualified; almost none are experienced database experts who have worked for years on advanced relational databases. Typically, open source contributors are recent college graduates who work on open source database technologies in their spare time. Unlike commercial database developers, they have no requirements or incentives to respond to the specific needs of paying customers.

FUD: open source developers are uncoordinated.

Reality: most open source developers are on the payroll or contractors of IBM, Novell, RedHat, JBoss, MySQL and db4objects. They hire the very best people.

And because that's not true, they can as well continue to make false claims:

The laws are so hazy that some companies, such as MySQL™ AB, have even been able to set up licensing programs that allow them to resell work contributed for free by the open source community.

Fact is: MySQL like Sleepycat like db4objects have individual contracts with every contributor of their respective product's source code which transfers rights to these companies, usually for financial compensation or a voluntary free ceding of rights. Everyhting else that the community contributes (Q&A, peer support in the forums, add ons etc.) happens on a voluntary basis and is nothing that these companies charge for. So the above claim is simply false.

Second, the pace of new development for open source databases is much slower than for commercial projects sometimes twice as long. For this reason,open source technologies in general are still much less advanced than their commercial counterparts.

FUD: Open source is not innovative

Reality: Oracle bought Sleepycat and attempts to buy JBoss because their respective, closed-source products Oracle Light and Toplink were outcompeted by "uncoordinated college graduates" such as Dr. Margo Seltzer, CTO of Sleepycat, or Gavin King, one of the best experts on O/R mapping in the world today.

And why would college graduates - even if it was true - not be innovative? I remember when I was a college graduate, that was probably the time, when I was most open to experimenting. By nature, you get less risk prone when you grow older, when you have a family to care for and a mortgage. I also cannot see Pointbase as an innovator. I see a complacent closed-source vendor that sees its revenues melt away, trying to lock people into their technology while spreading FUD about the attractive alternatives.

Of course, db4o, being the only native object database not only to Java, but also to .NET, experiencing rapid adoption by the likes of Boeing, IBM and Seagate is not innovative either... Just the fact, that db4o introduced Native Queries in version 5.0 in November 2005, what Microsoft plans to release as LINQ sometime in 2007 and 2008, should speak volumes.

Given that, how does this read:

Open source Java technologies are even further behind, with no technical support in existence. Based on the newest development language, it will be some years before open source Java technologies are able to catch up to even the C-based open source database products that have been in existence for a significant number of years.

Ha ha ha ha ha!

Argument #2 "The Open Source Licensing Boondoggle"

The authors state, that the GPL is viral. That is true.

But for exactly this case all 3 successful open source database vendors, MySQL, Sleepycat and db4objects adopted the dual license model: Instead of choosing the viral GPL, enterprises can alternatively choose a commercial license, which is exactly like the one that closed-source vendors offer, i.e. indemnified, safe, non-viral. Except for the price: It is up to 80x more affordable.

Can it be a disadvantage to have two licenses to choose from rather than one? A free GPL license to evaluate and do open source projects and education, while commercial customers choose the non-viral commercial license, which is much more affordable than those of closed source vendors due to the low cost open source business model? Of course, not.

And then they quote Oracle as their chief witness. Too bad, that Larry Ellison just changed Oracle's strategy with respect to open source: "We are moving aggressively into open source. We are embracing it. We are not going to fight this trend."

The Boondoggle only happens in the sales statistics of vendors like PointBase: They just loose too many deals against those vendors that offer better software at lower prices with more choices.

The Bloodbath

Stay tuned and look at the consolidation bloodbath that will happen, when all these overpriced niche vendors see their business implode. It will be like United Airlines (or for PointBase: Olympic Airways) against Southwest Airlines. And who doesn't like the lower prices of Southwest (no matter whether you fly Southwest or not - you benefit from their competition by seeing the old guys lowering their prices, too)?