Title: The Business and Economics of Linux and Open SourceAuthor: Martin FinkPublisher: Prentice Hall PTRISBN: 0130476773

The Business and Economics of Linux and Open Source is written for executives whose companies produce software and for IT managers who must choose and/or deploy this software within their companies. It introduces both free and open-source software (OSS), but predictably, the book focuses mostly on the latter. In spite of this, actually, for these reasons, I'd also recommend the book to hackers, for reasons that will be clear later in this review.

Part 1 covers the basics of what a kernel is and how Linux is developed, provides definitions of free software and OSS, explains how they deal with intellectual property and copyright, and discusses which licenses are available.

Part 2, titled "Operational Linux", discusses Linux and OSS deployment in a company, making clear when and why this is a good thing. The book presents all options, from installing an existing distribution to creating a custom one. All phases of deployment are covered, from migration to procurement, and all support options are examined in detail, from commercial to internal.

The last part of the book, "Open Source in Business", describes how a company could make money producing OSS or what a company should learn from OSS in order to make better proprietary software. Chapter 9 answers the question, "Since the 'bazaar' model produces better software, what kind of org chart and processes should a company have to fully apply it internally?"

The discussion of the real value of software--and how one can make money from producing, packaging or selling support for OSS--goes into great detail. Special attention is also paid to legal and intellectual property issues, from closed hardware drivers to training engineers to not blindly paste GPL code in a closed-source product.

The last chapter of the book is titled "Human Resources - Getting Top Talent". At the same level with traditional selection criteria, such as the "ability to focus on the project and its ROI", are traits like "(OSS/Free SW) community visibility and respect" and "actual contributions to other open projects". It was about time, wasn't it?

The Business and Economics of Linux and Open Source is readable for both hackers and managers. Buzzwords and technical jargon are kept to the bare minimum of what is required by the argument. Balance and concreteness abound; OSS is not presented as the answer to every problem. Instead, solid business reasons are provided for using or not using it, and good advice is given for both long-term strategies and daily operations. For example, we are told that buyers should be trained to pay only what must be paid. (I did hear about companies merrily paying X for support, Y for five Linux licenses").

The whole book also provides insight about what corporations can expect from both Linux and from individual hackers in the medium and long term. In the first case, the message is "No more UNIX wars, follow the LSB". In parallel, recruiters are told to:

Time to grow up, huh? Also, any manager reading "many, many new graduates are fully trained in Linux" might conclude "Great, lower salaries!". On the plus side, the author finally puts great emphasis on finding the right, competent engineers for the job and letting them be as free as possible.

A couple of important things are missing, however. First of all, mention should have been made that open data formats are even more important than open software for avoiding expensive, unnecessary upgrades. In the second place, the author, while demonstrating that "patent attacks on the OSS community" are a bad idea, does say that objections to software patents "stem from the fact that the community freely gives away its IP and expects the same in return". In truth, the community is not that childish. I would have really expected a pointer to any of these documents, which explain how serious and damaging for a market economy the whole issue is:

lpf.ai.mit.edu/Patents/patents.html

lpf.ai.mit.edu/Patents/AgainstSP/asp-toc.html

lpf.ai.mit.edu/Links/prep.ai.mit.edu/issues.article

swpat.ffii.org/vreji/papri/eubsa-swpat0202/

lpf.ai.mit.edu/Patents/danger-of-software-patents.txt

Even considering the two issues I mention above, the book can be quite useful for its target audience. Furthermore, it also can make easier something that is badly needed these days: a greater reciprocal understanding between hackers and executives.

Marco Fioretti is a hardware systems engineer interested in free software both as an EDA platform and (as the current leader of the RULE project) as an efficient desktop. Marco lives with his family in Rome, Italy.