Friday, November 25, 2022

R.I.P. Dr. Frederick Brooks

Dwight posts his obit.  Dr. Brooks wrote "The Mythical Man-Month" which is perhaps the seminal computer science text.  I read it 25 or 30 years ago and it's been useful to me over the course of my career.  Dr. Brooks taught that complexity kills projects, and the way to ensure success is not to throw bodies at the problem but rather ruthlessly enforce clarity in your goals.

Essentially, not all teams are created equal, just like not all cars are created equal.  So what's the best car?  It all depends on what you want to do with it.  A car is not a car is not a car.

I hadn't known that Brooks became a devout Christian in his middle age.  This really isn't remarkable - most of my science and engineering professors were regular church attendees.  Only modern intellectuals think there's a tension between science and theology.  Of course the obit writer is an intellectual and so this part of his life is presented as something mysterious and unusual.  Sigh.

6 comments:

Old NFO said...

I used that text as a program manager... it worked!!!

Ed Bonderenka said...

Now I've got something to read.
But KISS always works for me.
"I can program that code in three lines...."

Ed Bonderenka said...

Kindle is $19.24?????

Eric Wilner said...

A most useful book, which I first read in college.
The following quarter, I quoted it... in a paper... for a music lit class. While quoting Wagner in a project for a comp sci class. Sometimes, wisdom from one field can be applied in another.

Steve Sky said...

Have you read his 20th anniversary version of the book with the chapter, "no silver bullet"?

It's well worth reading, and I've periodically reread it to remind myself of real life programming. He starts out:
There is no single development, in either technology or management technique, which by itself promises even one order-of-magnitude improvement within a decade in productivity, in reliability, in simplicity.
It's a sobering reminder of reality, when being hit with the newest 'programming productivity tool' and the corresponding 'productivity improvement' expectations.

Eric Wilner said...

I didn't realize there was a new (1995 = "new"; is my age showing?) edition. Ordered! (Gah, $$$$)