The Design of Design: Essays from a Computer Scientist Book

No words I might add to the still-growing tide of reviews of Dr. Brooks' latest book would ever convince anyone to buy and read it. You either already know who Frederick P. Brooks, Ph.D is, or you don't. The sad thing is, like his other popular book, The Mythical Man-Month: Essays on Software Engineering, 20th Anniversary Edition (TMMM), most who will comment on it will never have read it completely, cover-to-cover, and understood much, if anything, in it. In the online world we find ourselves in today, what passes for research (never mind genuine, original research) sometimes only consists of finding some text (quoted elsewhere) using Google, and accepting it as fact. Hence, for example, the frequent misquotation (not to mention, the misrepresentation and misinterpretation) of "Brooks' Law." Sad to say, but this gem -- with an even larger load than TMMM of what in the broadcast world is termed "sound bites" -- will be equally misquoted.

For computer programmers, there is much, but not as much, I think, as the horde seems to have expected. I have already encountered online grumbling that the book is not what they were told to expect. Brooks discloses (as well as can be expected) that the book is about design -- and the care and feeding of what he calls great designers. Only some of his examples are taken from software engineering, and I am afraid the rest are likely to be deemed "too boring" for the impatient, increasingly-web-oriented audience to slog through and thereby learn what Dr. Brooks has to teach.

I think Dr. Brooks anticipated this (as he has many things) because he cautions the reader to treat the book as a series of related, but disconnected essays on the topic at hand. But I think most that would miss the point of many such essays would also ignore that advice to begin with (and then go on to complain).

I recently recommended this book to a well-known blogger who (tongue-in-cheek) complained that he would have to add it to his "already-too-long reading list." I responded that he should move it to the top and that he would be glad he did. I also offered the suggestion to heed Dr. Brooks' advice that it is not necessary to read every chapter. Dr. Brooks even suggests some chapters that can be (initially) skipped. For many, the chapters on the design of the NC beach house and the remodeling and expansion of his Chapel Hill, NC home can be initially omitted. The case studies of System/360 and OS/360 are interesting, but I think they would be obscure to anyone who was not once a programmer on those (or the early System/370) systems.

The bottom line is that you can skip almost half the pages, and not just initially. Brooks' prose is so engaging that I think most will go back and read the rest, anyway, but that's not necessary (nor even particularly valuable) the very first time. Like all of Brooks' written works, more and more value can be gleaned from rereading and reflective study. Except for the fact that it's been this way for 40 years, it would amaze me that, when I need to explain some things to those that are still wet behind their young programmers' ears, I can still find inspiration relevant to certain issues by reference to a book of his which is even older than The Mythical Man-Month.

And so I told the blogger to move it to the top of his reading list - and then put it back in the middle of the stack.

There is much here of interest, regarding both Dr. Brooks' personal and professional life and research, tidbits about the writing of TMMM and his equally great work, Computer Architecture: Concepts and Evolution (with Gerrit A. Blaauw), and his well-known Christian convictions (which do not intrude). Unlike TMMM, this is a good book to give to rising stars in any field of endeavor to encourage them to learn more about shining in their own field, how to go about it, and how to develop humility about their own talents and that of others on whom their efforts will necessarily depend.