Darrell Norton's Blog [MVP]

Fill in description here...

Review of Facts and Fallacies of Software Engineering by Robert Glass

Facts and Fallacies of Software Engineering is an excellent book. Short and concise, the book covers 55 essential facts and 5+5 fallacies that should be common software engineering knowledge by now. Facts are discussed and the controversy over them reviewed. And every fact is backed with the references, studies, and data we need to show non-techies why their proposed action or decision is wrong. The same format is used for the fallacies, but the arguments aren’t as compelling (that would be my complaint about the book).

This book is important because we, as software developers, need to curb the widespread fallacy propagation in our industry. People are still added to late projects, and estimates and project management still suck. When you talk to your doctor about a new medical treatment, he still determines if some new article is based on facts or is just over-hyped drivel. He (or she) makes the medical decisions and you make the personal decisions. In the same manner, when someone non-technical tells us to stick to unreasonable estimates or forces some other inane decision on the team, we need to clearly show why the decision is wrong with facts and stand our ground. Technical decisions should be made by technical people, and business decisions should be made by business people.

Published Jun 07 2004, 11:48 AM by darrell

Comments

Mark Bonafe said:

Well said. But the doctor gets paid the same no matter what the time table. If you tell him he must finish sooner, he can laugh at you and get away with it. He wouldn't do that, to be sure. He would explain why the process takes as long as it does and you have to live with it (pun) or not. Sure, we can explain when our clients' decision is wrong for them and back them up with all the facts in the world. But we aren't doctors. We can't just turn projects down because the client wants something done "wrong." Well, most of us can't, anyway.

Remember the golden rule? They hold the gold, so they make the rules.

# June 7, 2004 9:23 AM

Dave said:

Amen, brother. Amen.

# June 7, 2004 9:29 AM

Darrell said:

Bonafe - to some extent that is true. But we can't continue to allow non-techies to make technical decisions when we are held accountable. Plus whenever a fundamental fact is disobeyed by a customer, either knowingly or not, it is our duty as consultants to correct them. If they wish to continue, we outline how we will deal with it. If they continue to make demands we cannot live by, and the penalties reduce our payment, then there's no reason to take the project at a loss.

# June 7, 2004 9:58 AM

Dave said:

Darrell: For clarification, my comment was for your post, not for Mark's comment :-)

# June 7, 2004 1:29 PM

Darrell said:

Dave - got it! :)

# June 7, 2004 2:59 PM

JosephCooney said:

I have seen this book in the my local technical book-shop and have been meaning to read it for a while. You mention the "mythical man-month" fallacy in passing. What where some of the other facts/fallacies that were given good treatment?

# June 7, 2004 4:18 PM

Enjoy Every Sandwich said:

Take Outs for 7 June 2004

# June 7, 2004 4:29 PM

Darrell said:

Joseph - check out the reviews on Amazon. One guy has a really long review that covers more of the facts. You can also look at the table of contents, all of the facts are listed in it. :)

# June 8, 2004 2:14 AM

Mark Bonafe said:

I understand your point, Darrell. It's just that, for the first time in my long career, I have a client who has done it all - wrong. I have explained in every way possible why the client's demands will cause problems "down the road." Even with some example projects that detail the problems. I have done my best to correct him.

The penalties do not reduce my payment. The penalty is that I'm creating an application that will be hellish to maintain or modify and performance will degrade in a linear fashion with every new dll.

So I find myself being well paid, but producing something I know to be "less than optimal" and I hate doing it. My contract ends in October. The project does not - an extension is all but guaranteed, but I'd rather not continue. So, hurry up October. Anyone need help?

# June 8, 2004 2:31 AM

Darrell said:

Bonafe - ah, now that's a different situation entirely. One day maybe they'll understand, especially once you start giving them obscenely high estimates to make simple modifications. Heh. Good luck on it though.

# June 8, 2004 2:34 AM