The Open Source Business Intelligence Book, as originally conceived, is dead. Despite the transparency of the blogosphere, we don't feel the need to go into the details as to why. The Open Source Business Intelligence research project, blog, wiki and lens, will continue, and improve, without the distractions and roadblocks posed by the book project.

We do want to share some of the things that we have learned during our year long attempt to have our book published.

The traditional process of publishing a book doesn't serve the needs of the technology industry. Joe Wikert, in his "A Book Publisher's Blog", discusses many of the issues. A simple summary would be that it simply takes too long; by the time a book is accepted and published, it is, in many cases, out of date.

One lesson that keeps coming up over the years, is that every partnership, every business arrangement must be in writing, with as many details and contingencies thought out as is humanly possible. Until everyone is willing to put things in writing, you aren't doing business.

Some large publishing houses are hopeful that eBooks will be the answer for their technical book sales woes. But as long as the model is the traditional publishing model, the problems aren't addressed. Whether the end product is a static eBook or a static physical book, books sales will continue to slide.

Our frustrations with attempting to use traditional methods, and our learning experience with online publishing platforms, as well as our use of various eBook systems is causing us to reconsider our formulation of the TeleInterActive Press, with its blogs, podcasts, wikis, surveys, CMS and document management systems. What would other authors want? What do consumers need? How must the publishing industry change? How will all of this come together into a viable new publishing model to make technical publications relevant, immediate, easily updated and convenient?

We'll see as we continue to develop the OSBI wiki, this blog and related lens, surveys, and total content/document management, all in the context of researching other book topics.

The following are taken from an email exchnage with the publisher currently looking at our Open Source Business Intelligence book proposal.

'm sorry, but things are going more slowly due to the holidays. Now, I've really only gotten one review back (it was positive). My other reviewers appear not to be coming through.

"Can you suggest a couple of reviewers? Ideally somebody from the target audience?"
-- First email

Nevermind on the reviewers...my reviewers seem to have emerged from their
holiday slumber...
-- Second email

Dang! I hope it was the holiday cheer that put them out, and not the proposal itself. At least the one reviewer who was awake liked the idea.

We (Clarise, Bernard and I) recently submitted two article ideas, and one was rejected becuase

The second one is product-specific, thus unsuitable for publication. If you wanted to reshape it into a product review, which are written by end users to describe their experience with a product, please contact [a sister publication]... must emphasize that you avoid mentioning product names as well as methodologies specific to certain vendors.

Here's the proposal for that second article, which will now be writen on this blog.

Mondrian is an open source OLAP engine that is very mature, having been in use since 2001. It is of interest, not only for its own capabilities, but for the fact that it is included in or required for nearly every other open source BI project that includes OLAP capability, from simple tools such as jPivot to full BI suites such as Pentaho. This article provides details about Mondrian and discusses its use in and importance to open source BI. The article will also discuss how to incorporate Mondrian into an organization's BI project.

As far as we know, Mondrian doesn't have a commercial arm, though its recent relationship with Pentaho may change that. It is difficult for me to think of various open source solutions as "vendors". That's why I tend to refer to them as "projects" rather than "products". Granted, some open source projects are dual-licensed, or have a commercial arm, like Green Plum for Bizgres, and Kinetic Network for KETL, MySQL AB for MySQL. So, is this rejection an indication of a telling lack of awareness about open source, or am I wrong in my thinking? Are open source licensed software packages projects or products? Is discussing an open source project, a discussion about a "specific vendor"? Food for thought.

Open Source Business Intelligence is a technology with a ready market. Oopen source RDBMS have matured to the point where they can be used reliably as a data warehouse to support business intelligence solutions. Many open source projects are being introduced to expand the capabilities of open source solutions beyond reporting, and even simple OLAP, into complete business intelligence suites. Just take a look at the LinkBlog in the sidebar.

There are still some rough spots, however. So, in an effort to help data warehousing experts and open source afficianados along the path to making effective use of open source software for BI, we're introducing this blog, and a wiki, as companions to our effort to guide our forthcoming book with the working title of Open Source Business Intelligence.