I've observed, as have others, that there is often a large gap between "digital library research" and "digital library practice" (by some definition of those terms). I got a good taste of this at the Joint Conference on Digital Libraries last week. At one session, an audience member asked the presenter if he had read this:
Nov. 2004, PhD dissertation, Marcos André Gonçalves, "Streams, Structures, Spaces, Scenarios, and Societies (5S): A Formal Digital Library Framework and Its Applications", http://scholar.lib.vt.edu/theses/available/etd-12052004-135923/
... as it related to the topic at hand. The presenter hadn't heard of it, and neither had I. But why hadn't I heard of it?!? This sort of work should absolutely be on any digital library practicioner's reading list, and any researcher in this area, be it computer science (as this one was) or LIS, should have some familiarity and ongoing discourse with practicioners. Both pure research and pure implementations of digital libraries are necessary, but that doesn't mean there is no middle ground, or that the two can't engage each other in a meaningful way. My work will be better for having read this research, and research will be better for having learned about what departments like mine produce.
I think one reason for this gulf is the differing definition of "library" held by different folks. But that's a post for another day.