SOFTWARE REVIEW

Succeeding with Open Source review

8.0

Editors' Rating

Wendy M Grossman ZDNet.co.uk

Published: 28 Oct 2004

The concept of open-source software is familiar enough by now: software whose source code is openly published and may be modified or redistributed by anyone who cares to do so. But many businesses are still getting used to the idea of actually deploying software that's written by communities of volunteers -- even though a lot of open-source software is actually written by programmers paid by such well-known companies as IBM.

In Succeeding with Open Source, Bernard Golden, CEO of the California consultancy Navica, offers a formal guide to evaluating, selecting and using open-source software to fulfil business needs. Golden calls his method the 'Open Source Maturity Model', a framework he uses for determining how mature a particular open-source product is; he demonstrates it in practice by using it to analyse the application server JBoss.

One of the key issues with open-source software that exercises many IT managers is its licensing. Golden devotes a section of his chapter on risk to explaining the different types of licences and the consequences. In addition, he tackles the risk of intellectual property liability, concluding that this is more of a theoretical risk than a real-world one. Methods are also provided for assessing the quality and availability of technical support and training.

Golden also covers the additional resources that are available. Many -- such as checking the bug lists on Sourceforge, looking for test suites and other such metrics -- may be familiar to those who are already comfortable and experienced with open-source, but may not be obvious to the nervous newcomer. Also aimed at newcomers is the detailed advice on how to post to technical support mailing lists (reminiscent of the advice we used to give newcomers to Usenet). Perhaps most important, Golden tackles integration, which he describes as the Achilles heel of open-source. Commercial companies do not, for example, always consider it worth porting their products to open-source; however, new computing standards are beginning to alleviate this sort of problem.

What's astonishing in a way is that the publication of a book like this should be thought necessary when publishing a similar book about assessing commercial software is not. When you think of all the failed commercial projects that make the trade press, it's even more surprising. Yet commercial software is the devil we know. It's not perfect, but it seems to be proven. If open-source software seems to you like a foreign land with strange customs, Golden's book is a good guide.