Worse yet is when I travel to a city having one of those great massive bookstores. I invariably end up spending the better part of an entire day perusing the bookshelves. The Strand in NYC and Powell's in Portland leap to mind.
Much as I like them, though, books tend to make lousy prior art.
This is not to say that there's not likely some good prior art out there in book form. The hard-copy paper-and-binding medium, though, presents some serious problems for examiners that prevents their widespread use as prior art.
The biggest problem is identifying useful prior art in books in the first place. Since they're not in electronic form, there's obviously no way to perform an automated search on them. An examiner has to first have a strong suspicion that there's something useful in a particular book before they're going to seek it out. Usually this happens when the book has been cited in other relevant prior art.
The examiner then has to find out if the library has got a copy of the book. If not, they can file a request for an inter-library loan (which generally takes about a week or so).
Once (and if) they manage to get a copy of the book (I usually have been able to, but not always), the examiner then needs to read the thing, or at least skim it well enough to determine if it is useful as prior art.
Then, if it is useful, the examiner has to take the time to photocopy the relevant parts of the book, or sometimes the entire thing. (I'm sure somebody has worked out all that silly copyright stuff, right? I've often wondered...) This is truly mind-numbing work, believe me. I've gotten pretty good at doing it efficiently, but a few hundred pages will still take well over an hour.
Before the advent of electronic Office actions, the examiner would be done at this point, but today they need to take the additional step of scanning the photocopied book to create a pdf version that can be submitted with the electronic Office action.
Since examiners are given a limited amount of time to perform their examination, taking the extra time to pursue prior art in book form is not a very attractive option. This is one reason why attorneys might not see books used as prior art all that often.
So when I read stories in the press about Google wanting to digitize every book ever printed, I'm tempted to say more power to them. Of course, digitizing books would introduce a whole new category of prior art that can be easily searchable, resulting in more prior art that needs to be considered, so it would certainly present a mixed blessing.