Column: Mozilla gambit reveals risks of open sourcing

There has been much ado recently about the open-source movement following the success of the Linux operating system and the Apache Web server.

However, Netscape's sorry experience with open-source software development - it has yet to finish the next version of its browser - reveals how messy this approach can be for a public company.

Netscape jumped into the open-source model last year by giving away its browser source code (the "Mozilla release"). Netscape hoped that software developers around the world would suggest improvements, helping it compete with Microsoft. Initially, outside developers responded enthusiastically and downloaded hundreds of thousands of copies in a few weeks. Netscape received bug fixes and some feature suggestions, leading managers to predict that open source would speed up development of Navigator.

Yet problems emerged, delaying development at a time when Netscape's browser share has fallen below 40 pe rcent. Open sourcing isn't the only cause: Since America Online's acquisition of Netscape, the departures of key Mozilla team members have contributed to product delays. However, Mozilla's development process has created difficulties that the truly open software initiatives, Linux and Apache, haven't experienced.

First, because Netscape remains the corporate owner of the source code, what Netscape does is not really "open." Netscape insists on owning and distributing the software on its own terms. As a result, few outside developers have been willing to spend a lot of time writing code for Mozilla because they can't control its distribution or usage.

Outside developers have also found it difficult to understand the complex Navigator code base. The product grew from a few tens of thousands of lines to more than 3 million lines by Version 4. Netscape rushed those versions to market and lacked time to adequately modularise the code. Much of the code still resembles a pile of spaghetti - tough for anyone to unravel.

A more fundamental problem: Netscape can't control people who aren't employees or subcontractors. Any features that Netscape people want in the new version within a particular time frame, they need to develop themselves. They can't wait for the outside world to produce them spontaneously. So most developers working on Mozilla still are Netscape employees. In addition, the Mozilla team has had to throw out tentative schedules whenever it decided to accommodate time-consuming suggestions from the outside.

The unpredictability of open-source development matters little for Linux and Apache because these are the products of international movements, not for-profit corporations. (Netscape and Microsoft differ from companies like Red Hat, which make money by offering versions of open software with easy setup and fee-based technical support.) Linux and Apache have committees and a process to introduce changes to their code bases. But no for-profit entity owns the Linux or Apache code or has to schedule releases that go head-to-head with Microsoft in a market struggle.

The bottom line: Open-source development hasn't helped Netscape keep the Navigator browser alive, and it may not be an effective way for any commercial company to manage product development. It is similar to a free lunch, and we know what those are worth.

• Bookmark this page
• Share this article
  •  
  • facebook
  • slashdot
  • digg
  • Reddit
  • stumbleupon
  • linkedin
  • twitter
• Got more on this story? Email Computerworld
• Follow Computerworld on twitter

• Share
  Share this article

  •  google+
  • facebookfacebook
  • slashdotslashdot
  • diggdigg
  • RedditReddit
  • stumbleuponstumbleupon
  • linkedinlinkedin
  • twittertwitter
• print
• email
• Bookmark