• How can we achieve technology transfer from digital library research projects to practical applications? • How can we future-proof our digital libraries in the light of constant and accelerating change in hardware, software, formats, standards, and so on? The four-year NSF/ARPA/NASA-funded Digital Libraries Initiative (DLI) based at UIUC was entering its final year. Thus it was a good time to reexamine the development process. The idea was to use the DLI as a detailed case study from which to derive the questions outlined above in order to serve as a resource for future digital library activities. The focus on the DLI was complemented by external invited speakers giving a wider perspective. We were trying to provide valuable insights for those wanting to develop their own digital libraries. As well as issues that might be typified as theoretical, technological and managerial aspects of digital library research, design, development, and implementation, some are aspects imparting "trade craft." In any newly emerging field, the early developers must do much tinkering in order to pursue their larger theoretical goals. In this way the trade craft of the discipline emerges. These can be rules of thumb, hints, or tricks that improve the development, implementation, or acceptance process. These are the result of considerable trial and effort but individually are often not considered profound enough to merit mention in the published research results. Thus newcomers to the field wishing to replicate or build on the efforts of the early pioneers have to reinvent this trade craft-an unnecessary and wasteful process. As well as presentations, demonstrations of various parts of the UIUC project were made on the second evening, and the clinic was concluded with a chaired general discussion of issues and lessons learned from the DLI and other projects. The keynote address was by Stephen Griffin, program director, Digital Libraries Initiative, National Science Foundation (NSF). He summarized the history of the Digital Library Initiative (DLI-Phase 1) and the lessons learned for the second phase including the need for wider access and development in a larger framework which included content, use, and usability of distributed knowledge generally, as well as more interaction between technologists and librarians and a wider international perspective. Ed Fox reported on the Networked Digital Library of Theses and Dissertations. He described the ingenious mechanisms the project provides for addressing the concerns of access relating to copyright, patent processes, and prior publication regulations as thesis results are used as the basis for publication in journals. This is an example of the broader mana