Customer Reviews

Enterprise Application Architecture with VB, ASP and MTS

5 star:		(6)
4 star:		(4)
3 star:		(3)
2 star:		(2)
1 star:		(8)

 

 

In a nutshell Moniz puts forth a great concept, but plan on using your own implementation. Read on for the details. When I first read this book I couldn't turn the pages fast enough. I was elated that someone had an architecture that supported just about everything my users were asking for. Then we implemented, or tried to anyway. The code generated by his 'Object Factory' was poorly commented and dismally formated. It uses older ODBC, and improperly uses CreateObject() when the components are supposed to be built to take advantage of MTS. (You must use CreateInstance() to keep your components in the same context) He is also passing whole user-defined objects across process boundaries instead of serializing the data. Incredible performance hit! Still enamored with the concept I converted it to ADO and fixed the MTS errors, thinking that I would just copy this cleaned-up project over and over and edit it to support new objects. Man, was that ever complicated! So, now I'm in my third iteration (and last) of trying to implement this architecture by partitioning the functionality into separate components. In theory, this redisign should work better, and be much simpler than his implementation. (Maybe I'll write a book with my version.) ;^)>

This book not only wasted my time but resulted in utter frustration. I read over half of the book before returning it for a refund. This book is a bad mark in the Wrox name.

The book it self is bloated repeating things like `what we are going to cover in this chapter' multiple times in the chapter. The easy issues are covered in wasteful detail and the difficult issues that you would expect to find in this book are not covered at all. He spent an entire chapter on how to use an object that exposes properties and methods. Hello! VB 101?

The technical content also leaves much to be desired. I believe that he must not have spent much time developing in the real world because he suggests functionality like an undelete as part of every object. Most objects I know of rarely use delete let alone a thick client concept like an undelete. I believe that his architecture would not scale and perform like a hog. I think he would hit the ASP threading limit with 50 customers on one of his applications. (Not that I saw that issue mentioned in the book). He uses collections in a distributed application and seems to load all of the data for every search. When he got to the point to explain how to throttle the data for a search he skipped it indicating it was an easy exercise.

I am not usually this critical of any book and I respect Mr. Moniz for his work and ideas but I think this one will lead developers astray.

This book does a good job in the data layer portion of an N-Tier framework.

The author shows how to create classes that can create their own database objects such as the tables, views and stored procedures (although I don't know if I would ever use this technique - our DBAs would have a heart attack).

These classes also provide good functionality for Inserting, Updating, Fetching, Deleting, Auditing, and Un-doing changes to a table.

However, there are numerous problems with the book that prevent me from using the framework such as:

1: Very poor discussion/handling of business logic in the Business Layer. Every application I've written has complex logic that results from interactions between various Objects in the Business Layer. This framework does not discuss how to handle these interactions.

2: Non-intuitive way of handling Parent-Child relationships. For example, a Person object and Phone Number object would normally be modelled with a Person table, and a Phone Number table (that has a foreign key to the Person table). The author instead creates a separate join table represented by a "Connector" object.

3: Very little discussion of security. The objects in the book all have a Security property that can be set, but the objects don't ever do anything with this value. The author just briefly mentions this property and all the great things you can do with it, but never shows you how.

4: Potential performance issues - For functions in the data layer, all the properties are passed as arguments for Saving, Fetching, etc. This will cause a huge overhead hit when your objects have more than the 10-15 properties that the author uses. (Some of my objects have 75+ properties).

I knew there were going to be problems when I read "I trained as a chemical engineer and not as a computer scientist. Although I MIGHT have read a few books on the subject, I've never had any formal training in OOP, data access etc. What this means, for want of a better phrase, I don't know any better. Whereas many developers are constrained by what they think they know,...,I have nothing to hold me back".

How right Mr Moniz is. Chapter after chapter show a complete lack of understanding of basic computer science concepts. The book is overly complex, with simple issues such as handling data collisions given chapters for convoluted solutions.

To wear ignorance as a badge of honor is something I don't repect.

Give me my money back.

The book does a very poor job of explaining the higher level concepts that the author used to develop his "methodology". To make matters worse, the book is just difficult to read and I felt explained things out of order. I also had trouble with the author's methods. They seem more complex than they need to be. I also had trouble running the sample code downloaded from WROX. All in all I must say I'm very disappointed.

The content of this book definitely does NOT live up to the title "...ASP and MTS...". The chapter on ASP is very light. There are hundereds of pages TALKING about distributed components in the n-tier architecture, yet there is no mention of DCOM and the descussion of COM and MTS is almost useless. The author never really describes how to connect the component tiers with MTS.

I bought this book with very high hopes and was really disappointed when I couldn't read it - literally. This book is written in quasi-english, for which the editor should be well reprimanded. It's enough that it is horribly complex to understand some of the concepts, but to mix in unreadability makes it a dust-collector on my bookshelf. As an aside, I never got the download sample code to run properly either. I now work with someone who worked with the author, and have been told he was under enormous schedule pressure. It shows. I want my money back.

I purchased this book with high hopes, but it became immediately obvious that it was a rambling self glorification of one persons approach to Enterprise development (sort of "my way is the coolest").

The code is not available on a CD you must download it from Wrox.

And the code makes references to "custom" dlls but then does not provide the code for those dlls (as far as I could find.....finding anything in this book was difficult).

This may be a good book for theory, but for practicality.....nah.

I've read both editions of Rockford Lhotka's Visual Basic Business Objects and have been directing my development team in the use of these objects. The results were very good and we achieved interface independence. However, when we tried to reuse these objects in other applications, all we could was reuse was the database tables. Not what OOP promises at all. This is not to imply that the techniques are bad, wrong, or that Lhotka's books are bad. Just that from OOP, I've always been left wanting more.

Enterprise Application Architecture takes VB OOP to the level it needs to be at. The ability to reuse large portions of code, compilied or not, across multiple applications is addressed and handled very nicely. Joseph Moniz is to be commended on an writing an easily understandable book that presents advanced object design patterns that deliver on the promises of OOP about as much as I think I need.

This book does teach you to create reusable components. It misses out on the implementation of Business Logic and searching. The further you get along in the book the more you realize you're not really getting the answer to "How do I make a real application now?" It's very frustrating. I am using these principles and objects in a large scale web app and will report back here in a month or two when it's done whether or not I need a new job... :)