Applied Architecture Patterns on the Microsoft Platform Book Review

This is a book that would be good for anyone that wants to get a snapshot of the current Microsoft technology stack. It gives decent primers on Window Communication Foundation (WCF) 4.0 and Windows Workflow (WF) 4.0, Windows Server AppFabric, BizTalk, SQL Server, and Windows Azure. The primers are thorough enough to give you a decent understanding of each technology. The book then covers common scenarios found in most enterprise level applications. The scenarios include, Simple Workflow, Content-based Routing, Publish-Subscribe, Repair/Resubmit with Human Workflow, Remote Message Broadcasting, Debatching Bulk Data, Complex Event Processing, Cross-Organizational Supply Chain, Multiple Master Synchronization, Rapid Flexible Scalability, Low-Latency Request-Reply, Handling Large Session and Reference Data, and Website Load Burst and Failover. Each topic has a complete chapter dedicated to it. The chapters start off by describing the requirements, presenting a pattern, and then they list several candidate architectures giving the good and bad aspects of each. They then pick the best one and implement a solution. Most of the chapters were pretty good. The only one I found really lacking was the one on Multiple Master Synchronization. It included SSIS, Search Server Express, and Microsoft's Master Data Services as part of the solution. The details of the solutions were way to vague to give any real insight into how to implement the suggested architecture. All in all I found this book a very interesting read. It gave some great insight into Microsoft's current technology stack. I definitely recommend grabbing a copy.

MASTER DATA MANAGEMENT AND DATA GOVERNANCE, 2/E Book Review

This book is great for beginners just starting out with MDM as well as seasoned professionals. I was introduced to MDM with the first version of this book. That was not my copy, so I bought the second edition for myself. This book covers everything MDM from architectural considerations to governance and market trends. The book offers some really sound advice on building the business case for MDM and how to communicate that business case effectively. Listen to their advice!!!! I am an architect, purely a technical guy, but I can tell you from experience without the backing of the business your project will be squashed. The book covers the technical aspects of implementation in great detail. This book will give you the insight you need to fully understand the complexities of an MDM project. There are a ton of vendors out their claiming to have a magic install of MDM available for purchase, but the truth is they just do not exist. MDM is a process with many steps involved, not a product. All in all if you are involved or getting involved with MDM this is the book to read.

Requirements Engineering: Fundamentals, Principles, and Techniques Book Review

It is not common to see a book like this come out in today’s technical book market, or for that matter in any era of the technical book market. Today most books are rushed to market, and publishers seem desperate for authors, because there is some real junk being published. But I am not comparing this book to the junk. I am comparing it to the classics like Design Patterns: Elements of Reusable Object-Oriented Software, Patterns of Enterprise Application Architecture, and Code Complete: A Practical Handbook of Software Construction, Refactoring: Improving the Design of Existing Code . This book is going to be a classic. It is the best book I have seen on Requirements Engineering, by far!!!! It is organized very well. The logical flow and organization of the book helps to make it a great read from front to back, as well as a great reference. Some, I repeat "some", of what it covers includes System and Context Boundaries, Fundamentals of Goal Orientation, Documenting Goals, Fundamentals of Scenarios, Scenario Types, Documenting Scenarios, Solution-Oriented Requirements, Documenting Requirements in the Data Perspective, Documenting Requirements in the Functional Perspective, Documenting Requirements in the Behavioural Perspective, Documenting Quality Requirements in the Three Perspectives, Using UML2 and SysML, Natural Language Documentation, Structuring Natural Language Requirements, Fundamentals of Conceptual Modelling, Elicitation Techniques, Requirements Negotiation, Conflict Management, Requirements Validation, Requirements Management, Requirements Traceability, Prioritising Requirements, Change Management for Requirements, Co-development of Requirements and Architectural Artefacts, Requirements Engineering for Software Product Lines. The topics are in depth and concise. No filler chatter to make the book bigger. The way it covers the topics is through the use of a visual framework which breaks the topics down into different topics and activities. The framework allows you to see how all the topics and activities are related. The book contains a nice glossary, index and a well organized section on the references and other reading material. The book is big, which is a welcome change from some of the small books being printed today. Some of the diagrams in the smaller books can hardly be read. That is not the case with this book. This book does a wonderful job of making a very intense and difficult job understandable. If you have anything to do with software development this book is a must read. If you buy just one book on software development in this decade, make this the book. You won't regret it.

Agile Development != Chaos

The most agile project teams I have seen are those that do not claim to be agile or lean. They have a solid well documented architecture in place as well as designs of the modules being built. They have separated the responsibilities amongst the team members according to the team member's skill set. They don't try to pretend everyone has the experience levels that would allow them to contribute to all aspects of the development process. Requirements, Architecture, Analysis and design, and Proof of Concepts take up 80% of the projects resources of time and money, and coding takes up 20%. As the team's process becomes repeatable and reuse starts to be capitalized on, project's time to production is shortened, estimates are actually accurate, and budgets are met.

In my career I have come across approximately 20 teams claiming to be agile, one actually was. The rest of them used agile as an excuse to remove internal documentation and use refactoring as the excuse for horrible coding practices and coding efforts put into requirements that were not fully understood. Refactoring exists to improve code design, not fix broken code. So if your team is spending a ton of time refactoring they are wasting your money trying to fix their broken code, not improve their code design. Making changes to fix bugs or correct requirements that were missed to due lack of discipline or poor requirement elicitation practices is simply wasted money. Wasted money because your team does not have the skills to properly execute a proper software development process.

If your team is:
- Missing Dates
- Comes in over budget
- Suffers from unpredicted team member levels (adding unexpected man hours to complete a project)
- Bugs in production
- Development team needs to be part of the maintenance team in order for the maintenance team to understand the application's architecture and module design
- Doing one-off development (cut, paste, and modify the previous project's code base)
- You need to eliminate promised features to hit your dates

Then you are simply failing to executing whatever process you are claiming to have whether that be agile or not. Claiming to be agile is how most teams mask their incompetence. It is a shame because I know agile and lean processes work. What most of the IT industry does not understand is that executing an agile process correctly and delivering a solid product is much more difficult (requires more experience) than executing a traditional process like the Unified Process (UP). Read Agile Principles, Patterns, and Practices in C# to find out what your team should know and be doing to be able to implement an agile process. Very few teams I have come across have the skill level required to do anything with an agile process except create chaos.

One of my experiences turning this around involved removing what a company had deemed to be an agile process and implement an instance of the UP. Please don't get me wrong here, without the proper experience on your team, no process is going to be executed correctly. UP is just a process that includes iteration and milestones deliverables. Executed correctly these deliverables provide confidence as they start being delivered on time and start being used as communication tools during meetings and in other conversations.

The biggest hurdle to get over was the minds of the business owners who had been sold on the idea that agile = chaos (not literally). It doesn't. After a company spends a lot of time in the chaos mode (putting out the hottest production fire, missing dates, going over budget) it is hard to break them out of it. The worthless motion is perceived as productivity, and the internal teams never get the time to step back and take a look at the rat race they are running. Many times no one in the company has ever experienced any other way of doing things. You may have the same difficulty convincing the development teams as well.

Sometimes the only way I have found to break the cycle is to ask for a small project you can execute on your own from scratch to finish, or take on all the jobs considered drudgery yourself on a project team. Doing all the boring documentation, sitting through the long requirements elicitation meetings, simply doing all the stuff that allows a project to succeed but no one wants to do, or they don't know how. The concept is simply know as mentoring. Getting in at every level to help accomplish the tasks that would not be done, or would be done with the attitude of "this is stupid". It is the only way to show the value of the things that make a software development project succeed.

If you can get the development team to play ball, the business will warm up to the new process when dates are hit, they are within budget, and software starts going to production without bugs.

When the fires start going out and the chaos is tamed, the business can actually start thinking about their business.

Some other things Agile does not equal...
Agile Development != Low Ceremony
Agile != Ignoring Requirements