ACCU Home page ACCU Conference Page
Search Contact us ACCU at Flickr ACCU at GitHib ACCU at Facebook ACCU at Linked-in ACCU at Twitter Skip Navigation

Search in Book Reviews

The ACCU passes on review copies of computer books to its members for them to review. The result is a large, high quality collection of book reviews by programmers, for programmers. Currently there are 1949 reviews in the database and more every month.
Search is a simple string search in either book title or book author. The full text search is a search of the text of the review.
    View all alphabetically
Title:
Convergent Architecture
Author:
Richard Hubert
ISBN:
0 471 10560 0
Publisher:
Wiley
Pages:
276pp
Price:
£29-95
Reviewer:
Silvia de Beer
Subject:
process
Appeared in:
16-6
Convergent Architecture is a high ceremony methodology for developing software. The term Convergent Architecture indicates the whole process of development, from business modelling through to code generation, including testing and deployment. The author advocates one stream of development, using (where possible) one tool to refine the models of previous steps. By doing this, no information is lost between the different phases of the development.

The book is well written, but uses too many abbreviations to my taste. Halfway through the book I lost interest a bit, because Convergent Architecture was advocated as the new solution to all the software development problems. The descriptions given were very theoretical and repetitive and, above all, aimed at very large companies. The IT-organisational model discusses all the roles in the organisation, which you will never be able to fill in a small company, but only in a company of over 100 people if you would count the number of roles. Of course, a person can fulfil more thanone role, but still, in my opinion there are too many roles for a small company.

I think that it is a good thing that the author advocates that the whole process of development should become a more continuous flow. Supported with tools during the whole development process, and with as much as possible be automated, i.e. code generation and automatic model validation, to avoid as much as possible the tedious programming and debugging phase, which might generate implementations which do not match completely with the model. The author advocates that we should avoid changing the source code, but rather we should change the model, and regenerate the code. Software development should become more repetitive, to avoid reinventing the wheel for every new project.