One night a week I volunteer at a public library. I was in the children's non-fiction shelves when I noticed a recent enhancement: placards with illustrated Dewey Decimal System numbers. These are placed on the shelves amongst the books so they are at a kid's eye level.
They make it easy to spot, for example, that the books on birds start at 598 and that the books on cars are at 629.222. There were at least a couple of dozen placards, for topics such as dinosaurs, scary stories, the middle ages, etc.
What a cool idea. Not earth-shattering, and no doubt not unique, but still a useful enhancement. As I was shelving books I noticed a placard at 973 labeled American Indians. The book that was right next to that placard was about early American colonialists. Hmmmm... I'm no expert on the Dewey Decimal System, but that looked really odd to me.
I asked the branch manager about it and he confirmed that 973 is used for United States history.
So while there might be some books that mention American Indians at 973, the books dedicated to the subject of American Indians are at 970, which is for the History of North America. We removed the placard so that it could be redone. If a librarian had reviewed that placard before it was deployed the error would have been avoided.
An analogy that Jason uses when making the case for code review is that no one expects a writer to publish words without first having an editor apply a second pair of eyes to those words. The incorrect placard at the library provides yet another example of the value of review.
As far as I know, only that one placard was wrong. And it was only off
by 3! But as in computer software, 3 can make a really big difference.