"We would like to do agile but we cannot go because we need documentation for the FDA/SOX/..." is one of the statements I frequently hear after talks, panels, or in client conversations. This is a popular misconception about agile, which I'd like to clarify in this Advisor.

The concept of uncertainty is another way of defining project risk (including IT), insofar as risk is defined as those uncertain factors that can measurably and negatively impact project performance and undermine the outcomes sought. Historically, uncertainty has been identified and analyzed by its source: technical, people, financial, and so on. A completely alternative approach exists in thinking about uncertainty as the central characteristic of risk, an approach involving the foreseeability of risks.

Most IT practitioners today are well acquainted with the concepts, benefits, and pitfalls of offshore outsourcing. Furthermore, most companies, IT or non-IT, have used some form of sourcing since the early days of the information age.

A recent development that should help accelerate the adoption of service-oriented architectures (SOA) is the introduction of repositories and registries for cataloging and managing services and their associated artifacts. Over the past few months, several vendors have announced such offerings, including IBM, BEA Systems, and WebMethods.

Team captains in sports arise from natural leadership skills that key players use to bring the entire team up to a new level of play. A team captain is not necessarily the quarterback and is certainly not the coach or the general manager. The captain rallies the team in tough times. During some particularly tough times at Pages Software, Bruce Henderson, one of the senior developers, roamed the halls, wise-cracking and generally encouraging the rest of the team.

In the first two parts of this Advisor series (see "Errors, Mistakes, and Awareness, Part 1: Errors and Mistakes," 16 August 2006 and "Errors, Mistakes, and Awareness, Part 2: The Cost of Failures," 30 August 2006), I defined terms to clearly distinguish the causes and effects of human error in creating software. In this part, I shall examine how and when we notice errors.

In June of this year, Toyota and its Lexus brand took the top spot in 11 out of 19 vehicle categories in the J.D. Power and Associates' automotive quality survey. Yet less than a month later, Toyota President Katsuaki Watanabe bowed deeply in front of the world's press, publicly apologizing for the numerous quality problems that recently have been plaguing Toyota automotive products.

Last week, Business Objects announced plans to acquire Armstrong Lang Limited Software (ALG), a vendor of profitability management and activity-based costing solutions, for approximately US $56 million. The acquisition is an all-cash transaction for all outstanding shares of ALG by Business Objects (UK) Limited, a Business Objects wholly owned UK subsidiary. Business Objects plans to use ALG's software and consulting expertise to boost its own line of business performance management offerings.

Over the past year, I've seen a lot in business technology discipline (or not). I spent a day at a large enterprise where they explained how they had implemented 11 ERP systems and multiple instances of them. When I got up off the floor and promised them huge savings by reducing the variation in their back-office environment, they told me that they really weren't interested in the savings because they would require too many meetings, too many arguments, and -- well -- just too much discipline.

In this Advisor on agile integration, I return to the organization; specifically to improving organizational performance. One question that comes up often, particularly with large IT organizations is, "How do agile methods fit with our CMMi initiative?" In answering this question, we first need to look at the larger issue of improving overall organizational or enterprise performance.

Two tragedies struck in recent weeks: the loss of 49 lives at a Kentucky airfield and the loss of one life off the coast of Australia. Both episodes made the news as evidence of risks taken, perils faced, and lives lost. For those who missed the events, the 27 August crash of a Comair Canadair Regional Jet at the Lexington, Kentucky, USA, airfield claimed 49 lives. The jet was on the wrong runway and literally ran out of road.

In my previous Advisor for Cutter Consortium's Enterprise Architecture advisory service (see "User's Needs Analysis or Requirements Analysis?" 7 June 2006), I argued that it's not always safe or wise to entrust all requirements analysis to the contractor.

In my last Advisor (see "Domain-Specific Languages," 30 August 2006), I talked about domain-specific languages (DSL), where a DSL defines the design elements, or abstractions, for building solutions in a particular domain, using the concepts, terminology, and notation of that domain.

Over the past few years in the system integration arena, I have seen that the concept of earned value management (EVM) offers a widely accepted project management technique for many US federal government agencies, including the Department of Defense (DoD) and the National Aeronautics and Space Administration (NASA).

Several years ago, I attended a meeting in an unfamiliar building with a group of unfamiliar people. I sat at the meeting table, and the person to my right began talking. It was obvious that this person was in charge by his demeanor. He seemed to know most if not all of the people in the room and all about the subject of the meeting.

I was lost. As I said earlier, I didn't know who this person was, I didn't know who the other people were, and I was ignorant of the subject of the meeting.

For years now, CIOs have contended with business management's questions about the value of IT. But this is only half of the issue: CIOs also have to contend with the cost of IT. In our experience, the ways in which IT cost is managed -- how IT's cost affects the business units -- is a critical element of IT management and governance. For example, it's hugely different whether business units pay for IT (with fungible money) or whether the costs are assigned to them. In either case, knowing what the costs are is critical.

The purpose of enterprise decision management (EDM) is to automate (and impart consistency to) the decisions associated with such activities as marketing, product recommendation (i.e., personalization), pricing, workflow management, and compliance. To date, most of EDM coverage has focused on the use of business rules management systems (BRMS) for implementing the decision processing functionality underlying EDM applications.

In my last Advisor (see "Working Around the Delete Key, Part 1," 24 August 2006), I told the story of how I had spent a year or more getting around a malfunctioning delete key on my laptop. The article pointed out the lengths to which I, a fairly savvy computer user, was willing to go to avoid giving up my principal computer. This article is about some of the thoughts this experience brought to mind.

A recent discussion on NewGrange (a project management discussion list I moderate) had to do with the role of planning in ensuring project success. In my mind, planning is a perfectly good word that has gotten hijacked.

In June of this year, Toyota and its Lexus brand took the top spot in 11 out of 19 vehicle categories in the J.D. Power and Associates' automotive quality survey. Yet less than a month later, Toyota President Katsuaki Watanabe bowed deeply in front of the world's press, publicly apologizing for the numerous quality problems that recently have been plaguing Toyota automotive products.

Readers of last week's Advisor (see "Verifying the Value of Offshoring, Part 1," 31 August 2006) were introduced to the idea that the exploration of value created from an offshoring initiative is the last but not least important component of total management in the offshoring lifecycle.

When implementing a SOA, one of the biggest challenges is what service gets implemented first. Merely putting an SOA wrapper in front of an existing application runs the risk of tying the functions of your SOA architecture to the stove-piped legacy systems that you currently have. To identify the first SOA application, you really need a clear understanding of your overall SOA strategy -- and that takes time.

Conjure the image of the UN chambers, where delegates from around the world, their ears hidden behind headphones, listen with rapt attention as a colleague delivers an oration from the podium. The only thing permitting members from vastly different cultures and languages to work together is the army of translators who take the speaker's utterances and convert them into the native tongues of the audience. Would that this were so in the multilingual inner sanctum of information technology.

I'd like to take a moment to introduce the concept of virtually private networks. No, I don't mean virtual private networks (VPNs). I mean publicly available networks that contain information that is limited in its scope to a discrete class of user. I mean networks protected by the language barrier.

SeeWhy Software is offering a version of its real-time BI platform free for download. With SeeWhy Community Edition, SeeWhy Software is stealing a page from the open source community's strategy in order to accelerate interest in its products and services.