Precisely what we needed

At our company, we write Java applications. Soon, we got to the point that we needed a more formal way to read/write data than merely an ad hoc approach. We use XML. The obvious approach is to use a well tested relational database, like those supplied by IBM, Oracle or Microsoft. A problem was getting detailed, objective explanations of what would be involved with each choice. Each vendor is perfectly willing to be our "friend" and supply us with reams of documentation. But still...

The chapters in this book that describe how to hook up XML to those 3 vendors' databases were excellent and clear.

But what we ended up doing was going with something suggested in ANOTHER chapter - building an embedded XML database. You will not see this advocated by a vendor; there is no sale for them here. Other than this book, we found it tough to get lucid explanations of the pros and cons of this route. It will take more work, but we hope it will give better performance - no interprocess communication, for one thing. Plus of course no licence fees, and easier installation and management, since we will have access/own all the source code. This was not our original intention, by any means. But the book's comparative analysis was so persuasive that we ended up taking this road. (Hopefully, it will not be a dead end.)

That one chapter on embedded XML databases was, to us, the most precious thing in the entire book!

Other Reviews

• Well overview of available products and strategies
• Precisely what we needed