In-Depth

Will Open Source Be One of Your BI Sources?

Are you open to open-source? It’s out there, but look twice.

• By Mike Schiff
• 08/30/2006

The Open-Source movement has not been ignored by data warehouse and business intelligence vendors. While the Linux operating system and the Apache Web server have become the poster children for the open source community, the range of open-source products available today includes database and business intelligence software as well.

For example, MySQL AB has two open-source database products: MySQL, which the company says has more than 10 million installations, and MaxDB, which it obtained from SAP (who in turn obtained it from Software AG where it was known as Adabas D). Computer Associates has spun off its Ingres database to the open source community. PostgreSQL is another open-source database offering.

Actuate, primary driver of the Eclipse Foundation’s open-source Business Intelligence Reporting Tools (BIRT) project, offers a supported version branded as Actuate BIRT. JasperSoft, meanwhile, says its open-source reporting product, JasperReports, is currently being used in over 10 thousand corporate deployments, and boasts that its Web site is experiencing 100,000 downloads per month. JasperIntelligence, with additional features that include OLAP functionality, extends the company’s BI capabilities; and the company has touted open-source ETL technology in its product roadmap. Pentaho Corporation offers a suite of open source BI technology as well, and says it experienced 55,000 downloads of its software this past June.

The financial question involved in all these endeavors, of course: Where is the money? If the software is free, how can these companies generate revenue? Most commercial open-source product vendors derive revenue from services that include training, product support, documentation, and professional services consulting. Many of these vendors also offer the open-source technology through a fee-based commercial license.

While variations exist, open-source products are typically licensed under two general models: public (as typified by the General Public License or GPL) or commercial. With a public license, developers of an application created using open-source technology must make public any changes or extensions they make to the software. Further, if the application is distributed outside of their organization, they could be required to distribute the source code to the application as well. Under a commercial license, modifications to the open-source software usually do not need to be made public, and the application source code need not be distributed; the developers’ extensions and application code remain proprietary, which is typically viewed as a competitive advantage.

(As an aside, commercial licenses typically include an indemnification clause. Are we really at the point today where we must pay vendors for the right to use this software without running the risk of being sued? Will the indemnification clause ultimately be the SCO Group’s (the company claiming that some of its Unix code was used without permission in Linux distributions) legacy to the world?)

While it is somewhat obvious that developers seeking to market proprietary applications and products would likely seek a commercial license for any open-source technology that they utilize, this would also likely apply to internal IT shops using open-source software for the development of their own solutions. If they were to utilize a public license and distribute their software to a non-employee (e.g., partner or customer) they could potentially have to make the source code for the application publicly available as well—even to their competition! At the very least, they should have any open-source software license agreements reviewed by legal council so that they are aware of any obligations they have incurred as a result of using the software.

Note that organizations sponsoring open-source projects and vendors offering commercial licenses have introduced variations to these general categories that may be more or even less restrictive. As not every licensing option is available for every open-source offering, users should consider the ramifications of the available licensing options when evaluating open-source software. Any evaluation should also encompass product support options: public licenses often rely on Web-based bulletin boards and knowledge bases while commercial licenses often include more extensive “call center” support.

IT organizations that have yet to experience open-source BI and database technology would do well to download some of the software and try it out. While it may not have all the features and functionality of mature proprietary commercial offerings, it is in many cases a low-cost technology that can be expected to rapidly improve as the open-source community collaborates to enhance the technology. Open-source BI software is probably in your future; the real issue is not whether, but when. Even if you don’t expect to deploy open-source software in the near term, the mere fact that an organization is trying it can have a dramatic effect when negotiating with commercial database and BI vendors. These vendors want to expand the breadth and depth of their deployments in customer/prospect accounts and the perceived threat of competition from open-source software can lead to pricing and support concessions.

The source is out there; open up your horizons and give it a try.