We need to have a conversation about OpenURL, a close look at a corpus of error reports from the 2013 academic year. Or, How I learned to stop worrying and give up on clean metadata.
OpenURL has been a stalwart in the arsenal of Librarianship for many years now, but it is getting a bit long in the tooth, and cracks in the façade are showing. How well does it still stack up? In the fall of 2013 Brock University Library started looking at the reported errors generated through our OpenURL system to answer this question. By making a clever hack  to the error report mechanism it was possible to log all the malevolent OpenURLs in a database to see what the problem was. The end result is a corpus of around 1000 reported URLs  that were meticulously examined  to find the problem. In the end numerous things were discovered: - Some of the OpenURLs worked just as they were supposed to - Some databases provide really bad OpenURL requests - Some databases don't know how to resolve OpenURL requests properly - Punctuation in metadata often breaks OpenURL - Forget citations to supplementary material - DOI's can't save the day This poster will outline the details of the process and present visualizations of the results of the analysis. OpenURL looks great on paper but there ends up being many obstacles in the actual implementation that both frustrate users and leave Libraries feeling helpless as many of the fixes are out of their hands.  http://elibtronic.ca/content/20130823/tracking-sfx-error-reports-sans-effort  http://hdl.handle.net/10864/10653  https://github.com/elibtronic/metadata-cruncher/tree/master/open_url_breaker