Customer Reviews

Beautiful Software

5 star:		(4)
4 star:		(1)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

Connell's book makes a really interesting case for how to think about the problem of designing good software, and why the industry doesn't seem to do a very good job now. The author uses comparisons to the construction industry and other engineering disciplines to good use, showing how in other fields the types of failures found in the software industry would not be tolerated. He also makes cases for both open software, and for inherent patentability of software--no, they're not incompatible! The book is well-written in a style that's not jargon-laden, nor accessible only to those with special knowledge. This makes it of interest to those with merely an interest in the industry, as well as to software developers with years of experience.

This is a very enjoyable and informative series of essays exploring many aspects of software development that deserve more attention. Connell's main theme is that we tolerate substandard development practices in software that we would never tolerate in other fields. He discusses the similarities and differences between software and other engineering disciplines, focussing on why software projects are notoriously hard to schedule and frequently unreliable. The book leads through a series of well-organized issues and addresses them in a very readable and engaging style.

Overall, a very stimulating book about a topic that more of us should be thinking about (professionals and consumers alike). There are a few specific points where I disagree with the author but they are minor; and even then, the essay prompted me to dwell on interesting or unusual issues not normally considered.

Really, my only criticism is that I wish the book was longer and that Connell explored each topic in more depth. When Connell writes a follow up, I'll definitely read it.

Beautiful Software is a readable and enjoyable book. It's short, with several essays. And, some of the essays have been blogs, with replies, so the information is more engaging than a typical book written by one author without responses. In one chapter, the author even let's his critic have the last word. My favorite parts are the metaphors, such as comparing software to a house, requiring both good initial design by an architect and also good construction techniques by the builders. This book will be entertaining and interesting to anyone who has ever written any software, or who wonders why software can't be better, or is in any field of engineering that demands quality delivered to a schedule and a budget. I programmed for a few years in Visual Basic, but mostly have been in electrical engineering and reliability engineering, and I liked this book very much.

The writing style is very friendly, not at all condescending to those of us who are not experts in software engineering, nor is the tone of the writing at all arrogant when the author presents his opinions. And, it's refreshing to hear someone in the field of software explain that software can be better. This book would make a perfect gift--for yourself or someone else.

This is an engaging volume of highly readable essays on the nature of computer software engineering -- what distinguishes that discipline from computer science, and the ways in which it is similar to other engineering fields. Connell writes persuasively that we as a culture have become accustomed to poorly designed software in a way we would never tolerate with respect to, for example, civil engineering projects.

Connell examines the open-source movement as a potential solution, arguing that a publicly viewable code base is akin to visible support structures in a highway overpass. Then, however, in an entertaining back-and-forth with a Linux developer, Connell highlights that community's delusion that complex software can be developed and become mission-critical to a user base without the strong influence of a central architect.

The reason for the present state of affairs is that software engineering is ultimately a human discipline. It suffers from human limitations such as the desire to maintain (an employer's) market share, personal glory, money, etc., even as it requires that practitioners deeply understand user needs, as well as their own capabilities, the capabilities of their development team, and the range of existing tools and methods.

The book closes with a set of essays discussing whether or not it is even possible to develop a rigorous academic theory of software engineering. Taking into consideration the human aspects of the problems, Connell's predictable answer to this question is "partially." Notwithstanding the equivocal answer, however, the discussion of issues is where insights are found, and the book does not disappoint.

Highly recommended for anyone who appreciates good design in general and finds fault with computer systems and their software in particular.

"Beautiful SOFTWARE" is a collection of short, very readable essays on the issues software developers confront on a daily basis. Are you an individual developer who's finally bringing onboard additional help? Or possibly, you need to lead a disparate team towards a single solution? Then have everyone read this book (did I mention it was short?) and you'll find it should spark the kinds of conversations your team needs to have BEFORE they write a single line of code.