A recent essay in Nature on open access possibilities for chemistry data.

Murray-Rust, Peter (2008) “Chemistry for everyone.” Nature 451, 648-651 (7 February 2008)

The author (a chemical informatics specialist at Cambridge) doesn’t seem to see a role for libraries in this new information age, but only references libraries as being part of the old way of doing things:

The issue of open data is particularly problematic. Unlike astronomers, geoscientists and biologists, chemists have no global data-collection projects; their data are usually published in many different online journals and then collated by hand into CAS. In the era of real paper, limited page counts ensured that most chemical data were never published, and so are effectively lost. Even now, most electronic documents still use visual representations of a printed page (such as PDF files), rather than machine-friendly formats that allow data to be shared across different information systems. Moreover, the default business model for chemical publishing is ‘reader pays’. As a result, non-subscribers — that’s most of the world — have no access to a large percentage of chemical data.

But things are changing. The web is an almost infinite, comprehensive source of free data. Young scientists don’t go to libraries and no longer look to traditional sources of information, but to search engines such as Google. They expect to be able to express their questions in natural language and to get instant answers. They have no time to learn proprietary systems with idiosyncratic approaches. For reference, one of the first places to look is Wikipedia.