In 1985 I started on a project in the alcohol field concerning population models of alcohol use. Neither the Maastricht University library nor any other Dutch university library had material on this other than the regular scientific journals. I had to go to Amsterdam to lock myself up for a few days in the Stichting voor Wetenschappelijk Onderzoek Alcohol- en Drugs (Foundation for Scientific Research on Alcohol and Drugs; [SWOAD]) library (with doors opening into a lovely garden). In their collection I was able to find all that I was looking for, and much more. As it goes, SWOAD became Nederlands Instituut voor Alcohol en Drugs (Netherlands Institute for Alcohol and Drugs) moved to Utrecht and joined the Trimbos institute, when, in 2002, the alcohol and other drugs (AOD) library was eventually closed to the public. The collection is still there, but no additions have been made since 2005. At the moment, no further plans exist. In the early 1990s I spent a year working at the Alcohol Research Group in Berkeley, CA, USA. I practically lived in the library—the life's work of Andrea Mitchell. Sitting in a library and becoming immersed in a subject is one of the true pleasures of academic life. I recall being surprised many times, finding many things I was not looking for. Libraries are excellent places for immediate experiences of serendipity—a necessary source of inspiration for any science. These days are long gone, and most of our resources are available electronically. Most students and many academics here hardly ever go to the stacks anymore. They consult only digital sources, applying the adage ‘if Google can't find it, it is not worth looking at’. As the editorial rightly acknowledges, this idea will become more true in the years to come. The availability of information has a profound effect on what information will ‘survive’ and be used as base for further development. The internet seems a democratic platform, providing access to knowledge for more people than ever before, but that does not lead automatically to in-depth sources and richness of information. I underwrite the call in the editorial for identifying valuable collections and to start making them available electronically. This is costly, and I guess that research funds may not be willing to spend extra money on things not related directly to projects. Perhaps some work could be funded by joining large-scale heritage projects, currently under way in the Netherlands. Needed for such an exercise is the expertise of the AOD librarians, a dying breed as I interpret the call, who are the only ones able to save what should be saved for later generations of academics in the field 1. Meanwhile, at our library, I have taken out, for an indefinite loan period, a historically valuable report by an advisory committee on a proposal for a Dutch heroin trial in 1979, afraid as I am that the library would discard the report without letting us know. I agree, this is not the way to go forward, but am left with no other choice. none.