Mythical Man Month

The MythicalManMonth, Essays on Software Engineering, Anniversary Edition by Frederick P. Brooks, Jr., Published by Addison Wesley 1995

ISBN 0-201-83595-9 If every software manager read this book, we would, within two years, double the number of software projects that delivered production code.

Twenty years ago, FredBrooks looked at a project he had completed and, like so many of us, asked himself what had gone wrong. Unlike most of us, Brooks found answers besides blaming Fate and other people. -- The project was the creation of the system software for the IbmSystemThreeSixty.

There's more to the MythicalManMonth than its most famous contribution, BrooksLaw. The first chapter is a meditation on why creating software is so satisfying; I reread it when I'm discouraged and asking myself why I ever got into this crazy business. Brooks is an amazingly concise and lucid writer; reading this book is a pleasure for the quality of the prose alone.

Even if you own the MythicalManMonth, it's worth investing in the recently reissued 20th Anniversary Edition. This not only contains the valuable NoSilverBullet essay, but also has a tour-de-force: Brooks's own evaluation of where he was wrong (and right) in the original edition. The world would be a far better place if more people had Brooks's combination of humility, candor, and ruthless intellectual honesty.

-- BetsyHanesPerry

MythicalManMonth also coined "SecondSystemEffect" and popularized ConwaysLaw. Later editions of the book also include Brooks' essay NoSilverBullet.

Any software manager who hasn't read this book should be taken out and shot.

This is a bit radical, but you are giving me ideas. A couple of weeks ago we had a discussion with this big "Auditor" (terribly expensive flying over from the States) discussing planning...too much testing, not enough productivity from new teammembers, what is TimeBoxing ("FreeWheeling?")... Never read the book; neither had any of the guys at the company. But they have got their planning method, QA and ISO9000. Offered him the book, but he sleeps on the plane. Any suggestions? -- MartineDevos

Shoot him. -- DonOlson

[So there is a SilverBullet after all :-)]

DeFenestration. Well window-dressed, it looks like an accident. -- KirkBailey

Not only would taking out and shooting those managers improve software quality, it would dramatically increase the sales of The Mythical Man Month.... -- BillTrost

But the lions would starve with no LionFood!

Just fire them... helps to save lots of money. -- GeorgeFankhauser

But then you just end up passing them on to another company that you might end up at later. I once ended up on 3 assignments in a row where there was a person (with a PhD in CS no less) who left messes that I cleaned up. Besides, bullets are pretty cheap, so it would probably cost less to shoot them than waste the time and material for the termination paperwork. -- DarrenRemington?

This simple rule, consistently applied, would, within two years, double the number of software projects that delivered production code. Well, it might double the proportion of software projects that delivered production code. I personally think it would greatly reduce the number of software projects of all kinds. Which might not be such a bad thing.... -- CurtSampson

Thank you, C.S.; I needed that joke. -- DanielKnapp

Actually, as the managers who hadn't read the book were shot, new managers would be promoted to replace them. In time, the percentage of managers who hadn't read the book would approach a steady state. -- JamieNettles

Refactor them. They're pretty much fungible already.

Another book that I consider in similar calibre to Brooks's MythicalManMonth is SteveMcConnell's RapidDevelopment. It's an easy read, and covers pretty much everything that I've seen go wrong in previous projects. That said, there is a timeless elegance to Brooks's work that I doubt will ever be surpassed. I'm not so sure you should shoot the managers who haven't read it though. Drawn and Quartered sounds better. -- DaveBeer

As a manager who has read it, I am in quite favor of shooting those who have not. Gotta like cutting down on the competition. :-)

Anyone who hasn't read both of those books isn't competition. Wasn't that supposed to be a guarded secret?

TomWait? sent me a copy of The DesignOfDesign. Thank you Tom. This is FredBrooks collection of essays on design, published 2010. Does he mention XP, Agile, TDD, EmergentDesign? Not that I can find. I can't tell you how frustrated this makes me. I met Fred soon after NoSilverBullet was published. I showed him Smalltalk. He said, "maybe this is the silver bullet". Decades later is there any mention of Smalltalk or the design thinking it embodied or begot? Apparently not. You can't write 400 pages without getting some things right. But this book is a better history of how we got stuck than how we broke free. -- WardCunningham

CategoryBook