I tried this a few times in the CS sections of a university library, and it was shocking just how ephemeral most of the topics were. There were books that got V

I tried this a few times in the CS sections of a university library, and it was shocking just how ephemeral most of the topics were. There were books that got VERY EXCITED about systolic arrays. There was a small but persistent geologic stratum of books from the 1990s which all had to involve XML, Xpath, XSLT, Xforms, or some other revolutionary new technology beginning with the letter X. There were books with names like "Proceedings of the 89th conference in high-performance algorithms" which would have article after article talking about yet another kind of n-ary tree that performs ever-so-slightly better than another kind on a hypothetical computer without a cache. For some bizarre reason there were some of those "learn how to click the buttons correctly in MS Access 97 in only twenty days!!" books that had been mistakenly shelved in the CS section instead of the dumpster, and had never been checked out since. And very occasionally, once in a long while, there was something that would actually continue being relevant for multiple years after its publication.