IT managers who are about to reach for the aspirin as they try to figure out their open source corporate position should grab this book with both hands instead. As the title suggests, it's pragmatic and deals in realism, not ideology.

The first major benefit is the book's centering on case studies of real implementations, including quantities and dollars. The second is that this case experience is then organized into a plan and a check list to guide any IT manager into exploring what part of and how much of the corporate system can profitably be migrated to open source software and commodity hardware.

Profitably is the operative word, and the theme of the book is the return on investment (ROI) method as a means of evaluating migration potential and of persuading management to consider the merits of and need for the migration. The book examines and puts aside total cost of ownership (TCO) studies not only because they're often silent about their methods and the firms actually sampled; the TCO studies also reflect the motivations of their sponsors and deal in opinions, not facts, and prognostications, not experiences. The chief failing of TCO is that in adding up all present and future costs it omits any benefits, so that a cost/benefits assessment is impossible. The book, then, is a framework and method for conducting IT cost/benefit analysis. As Chad Robinson, senior analyst at the Robert Frances Group (RFG), puts it, "TCO doesn't examine what you could save, what flexibility it will provide you, how much money you could make. Those are the reasons you deploy something."

Despite the TCO focus of the 2002 RFG report, "Total Cost of Ownership for Linux Web Servers in the Enterprise," the book praises the study for its emphasis on the results of actual implementations. It likewise focuses on actual implementations in government and industry, and calculates the benefits from them. One benefit is a perpetual drop in licensing costs for machines switched from proprietary to open source software; another is lowered support costs resulting from having a choice of support vendors. An interesting fact from the RFG study is that "Average salaries for Linux administrators were about 4% higher than for Windows administrators, and about 20% lower than for Solaris administrators." No wonder the largest group of migrations so far could be expressed Slashdot-style as: (1) Drop old Unix hardware; (2) Buy Intel/AMD hardware; (3) Install Linux; (4) Profit. (Even Sun is taking this route with its new Opteron servers, with the hope that users will choose soon-to-be open source Solaris instead.)

The book begins with an introduction to Linux and a general background on how IT departments ended up having to evaluate platform-migration strategies. Rather than bash Microsoft as an evil empire, the author simply sees Windows as an increasingly expensive lock-in that managers must deal with, and shows how to minimize or eliminate the disruption of users from migration, including a good section on how to run Windows applications on Linux (on the other hand, there is scarcely a mention of synchronizing PDAs with Linux). There is important open source licensing information and a discussion of open source benefits.

There are explanations of ROI (including how to calculate benefits), detailed evalu-ations of current open source server and desktop software for corporate use, and a host of tips from the migration case studies. The most interesting is a discussion of how to host Microsoft Office on a Linux server and pipe it out to a number of concurrent users limited to the number of paid Office licenses in the company. The author is an advocate of remotely administered thin-client Linux for savings and efficiency (users can no longer load their computers with junk from home). The growing popularity of thin client is obvious from the recent crowding at the LWCE booth of the Linux Terminal Server Project (www.ltsp.org), which provides free resources for such projects.

While the book may end with rosy predictions of many support organizations and choices becoming available, its blunt talk ranges from schedules (manufacturer Ernie Ball - smarting from the BSA pillory - moved the entire company to open source in four months), to prices (Oracle will be embarrassed), to personnel (since Visual Basic programmers are tied to the Windows platform, they will need to be sent to reeducation camps or liquidated - well, not quite that blunt).

Busy managers should turn to the final chapter for a fine summary of the book's contents and methods before jumping into the details. Remember, benefits should outweigh costs (ROI) and reduced costs enable benefits, such as having more custom software written. Don't forget to look at the long term too, not least because of the competitive advantage the early adopters will have over the late ones. This book says it all.

To purchase this book, go to www.lulu.com/winslow.