I'm eagerly awaiting my copy. Brooks' previous work, 'The Mythical Man-Month', is still one of the best books ever written about building software. I re-read it every few years, just to keep those lessons fresh.
For those of you who aren't familiar with the book, Brooks was the manager of IBM's huge OS/360 project in the 60s. The system eventually shipped, but it was massively overbudget and late. He took the lessons he learned from the debacle, and wrote 'Mythical Man-Month' in 1975.
It's shocking how many of his insights still apply: Adding people slows down projects. Keep teams tight, and structured around your most productive programmers. It often pays to think deeply about a problem before you tackle it. Don't overbuild.
It's true that dynamic programming languages, tools like git, and rapid prototyping have made some of his examples feel dated. However, the underlying principles are still solid.
Hopefully 'The Design of Design' is a similar goldmine. Even if it's only half as good as MMM, it's still earned a slot on my bookshelf.
"For those of you who aren't familiar with the book, ..."
... buy it already. Except for occasional references to kilobytes of RAM (what the heck are those? are they like gigabytes?), you'd think the book was written last week.
I don't think it seems written last week. It slowly becomes out of date -- very slow, much slower than any other book, but still.
Among the things that has significantly changed since the 20th anniversary edition are:
(a) agile/incremental methods in software development are now widespread. In the book, Brooks present himself as a big advocate of what could be called proto-agile, contrary to the waterfall model of his time, and
(b) internet service companies flourishing in the internet. The "release early, release often" model could not be practiced anywhere better than in the internet, and this mindset has a great impact on how we think about the software development in general (note that at the Brook's times there were no internet services at all)
Thanks. I own the 20th anniversary edition and while I knew it had some extra content, including an update to the No Silver Bullet essay, I didn't realise NSB wasn't part of the original edition.
I found it very much worth buying the new MMM edition for his revisiting the "plan to throw one away, because you will" discussion on prototyping (e.g. see Joel's discussion on this). Very useful to me since I'd by then experienced most of the possible outcomes of this question.
He's not one to rest on his laurels and is to be commended for turning an epic disaster into a lifetime of suggesting to us with style and verve how we might manage to avoid such ourselves.
For those of you who aren't familiar with the book, Brooks was the manager of IBM's huge OS/360 project in the 60s. The system eventually shipped, but it was massively overbudget and late. He took the lessons he learned from the debacle, and wrote 'Mythical Man-Month' in 1975.
It's shocking how many of his insights still apply: Adding people slows down projects. Keep teams tight, and structured around your most productive programmers. It often pays to think deeply about a problem before you tackle it. Don't overbuild.
It's true that dynamic programming languages, tools like git, and rapid prototyping have made some of his examples feel dated. However, the underlying principles are still solid.
Hopefully 'The Design of Design' is a similar goldmine. Even if it's only half as good as MMM, it's still earned a slot on my bookshelf.