The buzz at the recent Charleston Conference  (and practically every other recent conference at which academic librarians have gathered) is a combination of new formats and a new collection development philosophy, shifting from print collections with titles chosen by librarians and faculty to making thousands of e-books available and letting the purchasing choices be made by "patrons"--an old-fashioned term for library users of all stripes, a large contingent of which are undergraduates writing "research papers" that are mostly papers synthesizing other people's research.
(Though one might think the "research paper" that has little to do with genuine research should have disappeared sometime after 1982, when Richard Larson famously skewered it as a "non-form of writing " that gives authentic research a bad name, but in fact this kind of expository writing from sources is more prevalent than ever . But I digress.)
This new way of building collections emphasizes speed and choice, things that are popular these days. No need to wait for interlibrary loan; just click on the title in a large shopping mall of e-books and you can have what you want right away. One model that's popular is to enter the e-book options into the library's catalog. Browsing for a short period of time is free; browsing for a longer period is treated as a rental and the library pays a fee; and if a book is "rented" four times, the library automatically purchases the book.
As some have pointed out, librarians don't have a terribly good record of acquiring books that are actually used. Why not let patrons take a whack at it? At our small undergraduate library, a disturbingly high percentage of books have never left the shelf, and decades later are still there, dated and in most cases useless except as historical artifacts. (Yes, we're working on weeding; it's painfully obvious how many mistakes were made.) And faculty selection isn't necessarily better. Most libraries have shelves full of books on a topic that was Professor Poindexter's passion, but which hasn't been taught since he retired in 1978. Other sections are sparse, because the faculty teaching those areas aren't interested in advising the library about acquisitions. Buying what you need when you need it--and being confident that every purchase will be used because it already has been--is tempting. Used as a supplement to traditional collection development (which is already patron-driven, in that we have always tried to match choices to expressed or even inarticulate needs) it makes a certain amount of sense--provided your patrons will use e-books. At our library the jury is out; in fact we haven't really seated the jury yet, since our first ebooks were so roundly spurned we've been reluctant to add more.
But I'm troubled by the enthusiasm with which some librarians are embracing the concept. Collection development as we knew it is so over! No more time wasted selecting books! No more money wasted buying ones that nobody uses! Patrons get what they want instantly, no lag-time waiting for interlibrary loan! (And in fact, the right to loan ebooks to other libraries is typically disabled, so no more time wasted sending books to other libraries.) In an era of declining budgets, it's tempting to think you can offer an all-you-can-read buffet and serve it up fast.
But the recent Project Information Literacy study  clarifies something I've long sensed: undergraduates don't necessarily need a bigger banquet. They need a limited number of good choices, not all-you-can-eat. (A study I worked on with colleagues of a popular and huge all-purpose journal database found that the majority of the full text articles downloaded by students came from a small number of titles, mostly popular magazines; 40% of full text journals didn't have a single article downloaded in two years at 14 institutions--yet librarians surveyed about the database thought the database wasn't too large; in fact, more full text would be better.) I also question whether the fast-food binging and purging that goes on as students feverishly compile patchwork prose masquerading as research is really a useful form of learning and whether libraries should build their programs around that practice.
Another reservation I have is the relative affordances of print versus e-books. So far as I can tell, browsing long-form texts online is harder for undergrads than browsing a shelf of print books. I experienced this myself just yesterday. While preparing for a Friday morning class introducing first year students to the idea of finding and using primary sources to develop a research question, I found myself frustrated by one very small selection of e-books that we have that includes primary sources from the Middle Ages. For the novice, they are virtually unusable. You have to guess whether books are going to be useful, open them one at a time, and scan through the contents to see if any of them might be relevant to a particular topic, which is difficult to determine because the authors and words in the titles of the entries are entirely unfamiliar. Yes, you can search the full text, but novices have a difficult time knowing what words might yield results. Trying to guess what words used in a 12th century text is even trickier. The kind of browsing that works for them--flipping pages, sampling contents, comparing texts side-by-side--seems to be simply easier with printed texts. And to make it even easier, we're putting them all together onto one shelf for the students. After all, the point of the assignment isn't to learn how to find books in the stacks, it's to query primary sources. They'll have an opportunity to find their own books as they develop an answer to their research question.
I'm not faulting the companies that are making these products available. Their customers are librarians, and librarians typically want their patrons to have as much choice as possible. I'm just not as thrilled with the e-book banquet as most academic librarians. Until our students take to e-books, and until I'm convinced that supporting the cut-and-paste research paper is the best use of our resources and their time, I will keep choosing books one at a time, doing my best to guess what students not only might want to use today for that paper due tomorrow, but what they might be able to use to good effect for at least the next few years.
Yes, I'll be wrong much of the time. But the alternative doesn't seem an improvement to me, at least for now.