Prentice Hall 1997
Date finished: 2003-06-06
A death march project is defined in this book as one whose parameters exceed the rational value by 50%; you have half as many staff as you think you need, or the schedule is 6 months when you think it will take 12, or the budget is 50% what it should be. Yourdon classifies death march projects into "mission impossible" (successful project, happy team), "ugly" (unhappy but successful team), "kamikaze" (happy failure), and "suicide" (unhappy failure). He suggests thinking carefully before committing to a death march -- changing jobs may be less pain than the stress and family problems that often follows months of 14-hour days -- and then provides various suggestions on surviving a death march and maximizing the slim chance of success.
Death march projects are less likely to succeed by their very nature, so you may as well take some other risks and ignore bureaucratic rules that get in the way. In a foreshadowing of extreme programming, Yourdon also suggests prioritizing features into must-have, should-have, and could-have; could-have features may not get implemented in time for the deadline.
The book doesn't take very long to read, and a number of its observations are worth remembering. If you're on a project that might be a death march, take an evening or two and read it.
%T Death March %S The Complete Software Developer's Guide to Surviving "Mission Impossible" Projects %A Yourdon, Edward %G ISBN 0-13-748310-4 %D 1997 %@ 2003-06-06 %I Prentice Hall %K computing %P 213pp