The Martian Principles for Successful Enterprise Systems: 20 Lessons Learned from NASA?s Mars Exploration Rover Mission

by Ronald Mak

• ISBN13: 9780471789659
• ISBN10: 0471789658
• Format: Paperback
• Copyright: 2006-05-01
• Publisher: Wiley

Summary

For the first time ever, the senior architect and lead developer for a key enterprise system on NASA's ongoing Mars Exploration Rover mission shares the secrets to one of the most difficult technology tasks of all-successful software development Written in a conversational, brief, and to-the-point style, this book presents principles learned from the Mars Rover project that will help ensure the success of software developed for any enterprise system Author Ronald Mak imparts anecdotes from his work on the Mars Rover and offers valuable lessons on software architecture, software engineering, design patterns, code development, and project management for any software, regardless of language or platform

Author Biography

Ronald Mak was a senior computer scientist and software architect at the NASA Ames Research Center. He was the architect and lead developer of the middleware for the Collaborative Information Portal, an important enterprise software system that is a part of NASA’s ongoing and highly successful Mars Exploration Rover mission. Mission managers, scientists, and engineers continue to use CIP—after over two years of continuous operation, it has an uptime record of better than 99.9 percent.
Working as a key member of the CIP development teamvalidated the principles that Ron describes in this book.Ron was also the architect and lead developer of an enterprise class information portal for NASA’s International Space Station and the future Crew Exploration Vehicle.
Prior to joining NASA, Ron had over 15 years of experience designing and developing enterprise systems using several programming languages and technologies on various platforms.
Most of these systems were highly successful, but therewere a few failures, too. The Martian principles are derivedfrom these experiences.
Ron held an academic appointment with the University of California at Santa Cruz, and he worked on contract to NASAAmes. He earned his B.S. degree with distinction in the Mathematical Sciences and his M.S. degree in Computer Science from Stanford University.
He has written three previous books on computer science, Java Number Cruncher, the Java Programmer’s Guide to Numerical Computing (Prentice Hall PTR, 2003), Writing Compilers and Interpreters, C++ Edition (Wiley, 1996), and Writing Compilers and Interpreters, a Practical Approach (Wiley, 1991). He recently wrote several papers about CIP for refereed journals. He continues to hone his exposition of the Martian principles by giving presentations to both industry and academic audiences.
Ron recently co-founded and is the CTO of Willard & Lowe Systems, Inc. (www.willardlowe.com) which develops enterprise systems for information management and collaboration.

Table of Contents

About the Author	iii
Preface	xi
Acknowledgments	xvii
Introduction	xix
Part 1 The Martian Principles	2	(106)
Principle 1 Don't reinvent the wheel.	5	(6)
Someone Else Has Already Solved Your Problem	7	(1)
Understand What Your Added Value Is	7	(1)
Use Commercial Software Whenever Practicable	8	(3)
Principle 2 You won't do better than what's already been done.	11	(6)
Adhere to Industry Standards and Best Practices	12	(1)
Seek User Groups, Chat Forums, Online Documentation, Books, and the Like	13	(1)
Do Not Gum Up the Plumbing	14	(3)
Principle 3 Your customers don't know what they want	17	(6)
Do Not Push Too Hard on the Requirements	18	(1)
Do Rapid Prototyping and Lots of User Testing	19	(1)
Customers Are the Best Testers	20	(3)
Principle 4 Get something working as soon as possible	23	(6)
The First End-to-End Thread Is Critical	24	(2)
Working Code Builds Confidence	26	(1)
Always Build on Top of Working Code	26	(3)
Principle 5 Use sound software engineering practices	29	(6)
Use a Component-Based Architecture	30	(1)
Use Design Patterns	31	(2)
Get All the Development Team Members to Agree	33	(2)
Principle 6 Don't trust the client applications	35	(6)
Be Very Paranoid---They Are Out to Get You	36	(1)
If You Do Not Break It, They Will	37	(4)
Principle 7 Plan to make changes	41	(6)
Do Not Hard-Code Values	42	(1)
Use External Editable Parameter Files	42	(1)
Implement a ``Read Parameters'' Method	43	(1)
Maintain Client Parameter Files Centrally on the Server	44	(1)
Identify the Parameter Values	45	(2)
Principle 8 You can't predict the future.	47	(8)
Make Each Service Dynamically Reconfigurable	48	(1)
Create Field-Replaceable, Plug-and-Play Services	48	(2)
Hot Redeployment Allows Reconfiguring without Rebooting	50	(1)
Make Each Service Loosely Coupled	51	(1)
Take a Peek into the Future	51	(4)
Principle 9 Don't tie your services into knots.	55	(4)
Keep Your Services Independent of Each Other	56	(1)
Services Should Treat Other Services As Equals	56	(3)
Principle 10 Build early, build often!	59	(8)
The Major Challenge Is Not Code Development but Code Integration	60	(2)
Use a Source Code Repository	62	(1)
Maintain a Separate Environment in Which to Build and Deploy	63	(4)
Principle 11 ``What middleware?'' should be your greatest compliment.	67	(6)
The Middleware Should Be Invisible to Users	68	(1)
Good Middleware Creates Ideal Virtual Worlds for the End Users	69	(4)
Principle 12 Expose the invisible.	73	(8)
Put Hooks in the Middleware	74	(1)
Do Runtime, Real-Time Monitoring	75	(6)
Principle 13 Log everything.	81	(8)
Do Not Turn Off Logging in Your Production Code	82	(2)
``Log Mining'' Reveals Usage Patterns	84	(5)
Principle 14 Know the data.	89	(8)
Learn the Data Usage Patterns	90	(1)
Create Appropriate Application and Middleware Data Models	91	(2)
Map to Practical Physical Data Models	93	(1)
Adapt to Third-Party and Legacy Data	93	(4)
Principle 15 Know when it will break.	97	(6)
Do Lots of Stress Testing	98	(3)
If You Do Not Find Out What the Limits Are, Your Users Surely Will	101	(2)
Principle 16 Don't fail due to unexpected success.	103	(5)
Missions May Last Longer Than Expected	104	(1)
Data Repositories May Grow Larger Than Planned	105	(3)
Part 2 Project Management and Software Engineering	108	(26)
Principle 17 Strong leadership drives a project to success.	111	(8)
A Good Architect Must Also Be a Good Leader	112	(1)
Any Architecture Is Only As Good As Its Implementation	113	(1)
Strong Project Management Is Necessary for Success	113	(1)
Project Milestones Are Opportunities for Demos and Rebalancing	114	(1)
The Project Milestones Near the End Allow You to Get Your Project Done on Schedule	115	(4)
Principle 18 Don't ignore people issues.	119	(6)
Software Projects Are Not Democracies	120	(1)
Agree to Disagree, but Then Move On	120	(1)
Scale the Project According to the Team Members' Abilities and Experience	121	(1)
Do Not Be a Slave to the Latest Project Management Methodology	122	(1)
Foster Good Communication Plus Teamwork, Teamwork, Teamwork	122	(1)
Remove Team Members Who Cannot or Will Not Perform	123	(2)
Principle 19 Software engineering is all about the D's.	125	(6)
Discovery	126	(1)
Diplomacy	126	(1)
Definition	127	(1)
Design	127	(1)
Development	128	(1)
Debugging	128	(1)
Documentation	128	(1)
Deployment	129	(1)
Dmaintenance	129	(2)
Principle 20 The formulas for success aren't complicated	131	(3)
Successful Architect = Good Designer + Good Developer + Good Leader	132	(1)
Successful System = Good Architecture + Good Software Engineering	132	(1)
Keep It Simple!	133	(1)
Index	134