data storage and management

I approached this book with an open mind, but after I stumbled upon a couple of obviously wrong and some nonsensical statements, I started reading it much more skeptically, and finding more and more problems with it. (However, it did provide more fun ;))

The book has a good premise, trying to explain information system with the factory metaphor. Although authors give some good insight in the way IS should or could be thought of and modeled, there are many instances in the text where you read something and say to yourself "what where these good people thinking". This then undermines your confidence in their vision and full understanding of the matter. And although I think this is a matter of personal preference, authors sometimes seem to be in love with their style, producing some beautiful nonsense like this: "The legacy environment is only a very small vestige of its former invincible self." (pg. 42)

Let me give you some more examples of what I'm talking about:
Authors create metaphors of user classes, calling them "tourist, farmer, explorer and miner", which in itself is not a bad idea, but then they go on to say "...farmers found at the ODS environment are quite different from the farmers found at the data mart". So why did you create the single metaphor then?

Also, check this out: "A miner will typically look over many, many rows of data...". As opposed to what, just a "many rows of data"? Whence some people might need "not so many rows of data"? Like I'm reading a book for my eight-year-old, for goodness sakes!

Then there is this graph showing the directional flow of data, but then it reads: by the way, in all this streams, data can sometimes ("in 1-5% of the cases" - authors never say how they got these numbers, it is all a slight of hand) flow in the opposite direction!!?(Pg. 24) And they go about giving 5 examples, 4 of which are wrong - there is no data flow in them at all!
Example:"...sales dept. notes that loans are slowing down. The decision is made to reduce home loan rates by 1/2 percent."
This is not the back flow of data, as authors assert, it is an information feedback loop that involves people (management), and their decisions. Data (loan rates) is not coming back from Data Mart. User is somehow entering it into Operational system (application). His decision is influenced by data analysis, but it does not reverse the data flow. The fundamental issue here is that authors ignored the fact that information processes in the company involve people as well as the data and systems, and should be modeled as such. To use their metaphor, users should be a part of the information ecosystem. Hence it is not true that, as the book claims, corporate information factory embodies the information ecosystem. (Pg.7)

"...'event' date ...reflects the moment in time when the data in the record was accurate". (Pg. 96) This is incorrect. 'event' date is just recording the time of that single event. Record is always accurate after that, it does not 'age' with time.

At pg. 191 it is asserted that Data Warehouse provides "depth" to the data. That is true only if it is built (modeled) with "depth" requirements in mind. Before I can get "deep" information from the DW, I must build it with my questions in mind, otherwise, it will not give me data. An abstractly deep DW does not exist. It is always an answer to a particular question, or number of questions.

"The Dimension of History" (pg. 193-194) is just plain good old nonsense, with example (life stages of an individual) being completely off the mark. Reminds me of the student who does not know the correct answer to the question so he tries to invent some plausible response, letting his imagination fly ... Sad.

And so on, and so on...

I saved the best ones for the end:
"...the external world is full of normal occurrences and normal events. The very ordinary nature of the external world makes us take it for granted." (Pg. 49) Very philosophically deep, indeed ;)
"The emergence of the integrated applications comes slowly and, in many cases, imperceptibly" (pg.42) Yeah, it just creeps on you when you're not looking... ;)
".. the back flow of the data is minuscule to the point that in some cases it is so small as to be unmeasurable". (Pg. 23) :0 Beg your pardon? This is not quantum physics, guys, this is computer science. Anything can be observed, perceived and measured to the level of a single bit. Or are we talking bit-quarks here? Informational principle of uncertainty?

It is disappointing to have this book co-authored by the "father of the data warehouse".

To the (prospective) readers: This is a fun book if you are an experienced data architect, bad if you wish to use it as a blueprint for your work, and dangerous if you are an IT manager and impose it on your staff.

To the authors: Give us a break, please go back and re-make a decent book around the good basic idea. Less poetic style would also be appreciated. Forget about quantum physics. And give it to some unbiased reviewers first. Remember, only the real friends will tell you the unpopular truth.

The very idea that all the technologies serve the purpose of businesses gets lost here in this book.

If one wants to look for help for their work, look somewhere else.
If one wants to learn some jargons to impress his/her date, this one is the ideal one.

Evaluating "overview" books such as these is difficult. If this is a new subject to you or you are a manager needing an overview, you'll probably find it a good introduction to the Inmon-style Data Warehouse (the other being Kimball-style). But after giving the book a "once-over", that's it. Nothing is handled in enough detail to make this any kind of reference or guide. So from the novice perspective this is a "must read" and as good an introduction to CIF as you will get, from the people that define the standards - give it 4 stars (why not 5 stars will be discussed later). If you have worked on Data Warehousing projects in the past but your environment was not particularly designed with much forethought, then by all means whiz through this book and you'll see that all the pieces can be part of a cohesive strategy - 3 stars. If you are experienced then you'll probably see this as a waste of time, too high level. You can get everything in here from various web sites, whitepapers or other books - 1 star. Regardless of where you're coming from though this is one of the few books that actually use pictures in a way that is detrimental to clarity. Some are repeated so many times, so inane or so large that you have to surmise that they are mostly there to thicken the book. The appendix, CIF Architecture Guidelines, is also the largest section of the book yet I have no idea what the point is. I think it would be more appropriately titled - Incomplete Ruminations on Random Topics.