An old friend.

It was the Turing Award Lecture in 1972. Printed in "Classics in Software Engineering" by Yourdon Press, 1979, ISBN 0917072146

Here are some of the previous submissions here on HN:

http://news.ycombinator.com/item?id=86288

http://news.ycombinator.com/item?id=109724

http://news.ycombinator.com/item?id=126638

http://news.ycombinator.com/item?id=135111

http://news.ycombinator.com/item?id=156505

http://news.ycombinator.com/item?id=449806

http://news.ycombinator.com/item?id=1179277

http://news.ycombinator.com/item?id=1649246

http://news.ycombinator.com/item?id=1672262

http://news.ycombinator.com/item?id=1799296

http://news.ycombinator.com/item?id=1894784

http://news.ycombinator.com/item?id=2011732

So many postings, so little discussion.

This is the only one with substantial comments:

http://news.ycombinator.com/item?id=1799296

Edit: for some reason missed this: http://news.ycombinator.com/item?id=1894784

Heh, one of my lecturers was ranting about the millions wasted on research grants for testing methodologies when it's proven that they can't formally confirm program correctness. Old Edsger had a lot of foresight in that respect. Still, he's gloriously ambiguous about what he considers a 'modest and elegant programming language'.