I disagree with the author, who says "My code traded the ability to change requirements for reduced duplication, and it was not a good trade." I think this is

I disagree with the author, who says "My code traded the ability to change requirements for reduced duplication, and it was not a good trade." I think this is a great trade, because the number of times I've seen errors resulting from duplicate code is FAR greater than the number of times I've had to kludge an abstraction because of changed requirements.