Continuing on from previously:
Another problem with electronic formats is the ease of “skimming” for research purposes. But this is probably more a matter of interface design that needs to catch up. For those who don’t know: to traverse a lot of literature quickly, you need to be able to, non-linearly but nonetheless highly methodically, work your way through a potential target text, often by just reading say, the first and last paragraph of every chapter, or the first and last sentence of every paragraph. Electronic books kind of assume that you will read the entire thing sequentially; flipping through them can be hard.
Something else related to this is the internal references in the text themselves. I mean god knows why book publishers insist on end notes but their idiocy is massively amplified in electronic editions. It’s even worse in multi-author compilations, when you just don’t want to check the end notes, but the bibliography for each article is off with the fairies somewhere down the back the (750-page) book, and the publisher has provided no internal hyperlinking. This is just stupid, and actually makes using the text for what it’s intended, i.e. scholarly research, almost impossible. Why do they do this, even think that’s a good idea? Publishers, if you’re not going to hyperlink the references then every reference ought to be a footnote, and rendered right on the same page with the text it appears in. Got that?
On the dreaded DRM see this excellent post by the author Charlie Stross.