Metapattern: Context and Time in Information Models [Paperback]

5+ stars for the author's depth of thinking
1- star for requiring the same depth of thinking from the reader
4 stars

A quote from an appendix: "Multicontextualism offers a powerful synthesis, a blend of analytical philosophy and structuralism. Or, as a friend remarked, it unites Wittgenstein I and II". If the name Wittgenstein means anything to you, let this be your warning about how abstract this book is. The author had provided an abundance of examples. In the spirit of the title, I would have to label them as metaexamples. They are too abstract. While reading this book I felt a state of weightlessness, struggling to locate any firm ground to support my understanding.

To get the full value out of this book may require more than one reading. After the first reading I am left with a valuable inspiration. What attracted me to this book in the first place? Simple and consistent treatment of time as a fundamental attribute of information objects. Treatment of the classification of information into multiple subjects.

See the publisher's web page for the detailed table of contents. A large part of the book consists of the reviews of several other books. This makes this text a review of a review, or a metareview. It may be helpful to read those books first. I did not. Here is the list of those books....

Chapter 6-9, 13:
[Advanced Object-Oriented Analysis and Design Using Uml] [by James J. Odell]

Chapter 10, 11:
[Business Process Engineering : Reference Models for Industrial Enterprises] [by August Wilhelm Scheer]

Chapter 12, 13:
[Data Model Patterns : Conventions of Thought] [by David C. Hay]

Chapter 14, 15:
[Analysis Patterns : Reusable Object Models] [by Martin Fowler]

Chapter 9:
[Framing Software Reuse: Lessons From the Real World] [by Paul G. Bassett]
[Business Specifications: The Key to Successful Software Engineering] [by Haim Kilov]

I purchased this book about 4 years ago and did
not understand it on the first read.

After studying functional languages such as O'Caml,
SML/NJ, Haskell, and Common Lisp specifically
CLOS along with ontological editors such as Protege
the lights started coming on. This book is years
ahead of its time and will one day be recognized
for the master piece that it is. That day will
come as XML B2B integration efforts turn towards
ontological mappings as the means of engineering
the flow and configuration of content in time and
context. Much of the theoretical foundations presented
here will one day be applied to data warehousing
also. When that day comes this book being resold
now for pocket change will likely become a
collector's item. I hope Pieter goes ahead and
publishes his next title Semiosis & Sign Exchange,
Conceptual Grounds of Information Modeling. Pieter
is a true visionary.

A lot of Microsoft-tool obsessed programmers I know toss of the phrase "it's academic" like rice at a wedding to deride technical books/solutions they cannot buy for $XXX.99 from their favorite software vendor.

Consequently, they do not get to benefit from books like this. Books that take some things that are not instinctively simple to solve elegantly, and give you pat solutions to these problems that fit well with today's object-oriented information systems.

This book is not for people who buy solutions from other people. This is for people who create solutions - but don't mind learning from others who have gone before them.

You will learn how to model real world objects in practical object-oriented terms.

The book aims to make you a better object-oriented analyst/achitect/programmer. In my opinion, it will do that if you are already a good one. If you know nothing of programming or OOP then I would not bother starting with this book. Learn the basics first, then grab this one.

It's not that this book is hard to read - it is not.

It is just that you will not have anything to apply this knowledge with if you don't have some concrete programming skills and knowledge of how to take an object-oriented design and transform it into object-oriented code.

If you do have those skills, this book is a superior tool for problem solvers. It can sharpen those skills. Whether you are designing an enterprise IT system or a sophisticated web site that models some complex objects, you will be glad you have this book to reach for on your bookshelf or in your memory.

The book teaches you how to put not just some abstract representations of physical or abstract objects into a model - but how to do it in context.

The context could be geographical, temporal (having to do with absolute date/time), or chronological (ordered by time). It could have to do with authorization or authority. It might have to do with hierarchical categorizations of topics, subtopics, and subjects - and inherited characteristics.

The context things occur in are often what make them significant. And often, it is the context itself which has significance. Grab this book and you can do significant work.

Subjects include: contexts, types, documents, publishing, accounting, etc. Pat, standard ways to model these things. Diagrams are included with each model so you have a visual picture to stick in your head, along with the descriptions and advice.

It is not strictly limited to time and chronology, though the subtitle tends to imply that. It covers that subject aptly but it covers all manor of relationships of objects and types of objects to each other.

There is more to this world than arrays, vectors, trees, lists, sets, records, tables, rows, and vanilla business objects. There is what you do with them and the target you aim to hit with groups of them.

This work gives you the tools to do implementations of business objects and domain objects that suit the real world. That, my friend, is your job.

This book is a tool that will sharpen your ability to address complex information system design problems with considerably more powerful idioms at your disposal.

It will help you see your target and hit it.