an early 2003 paper on IRs that stood the test of time...

Continuing with the Institute on Scholarly Communication. Over the past couple of days at breakfast, I have been re-reading one of the foundational articles on IRs, "Institutional Repositories: Essential Infrastructure for Scholarship in the Digital Age," written by Clifford Lynch. This article has, in my view, stood the test of time very well. Here are some thoughts from 2006:

• The "Preservable Formats" section provides a nice discussion on preservation issues. In the context of the DSpace software being employed for WSU Research Exchange, Lynch's breakdown of support levels between full and qualified support (or "supported" and "known" in DSpace nomenclature) is readily comprehensible.
  
  In a separate section, Lynch articulates one concern about IRs in general that's particularly relevant to the digital preservation task: "at some institutions, repositories will be offered hastily and without much real institutional commitment." For a long term effort, the costs related to properly classing file formats for preservation purposes, performing format migration, and perhaps paying for support from a digital preservation service all need to be factored into the budget for an IR effort.
  
  
• The "Identifiers" discussion in the article touches on the versioning issue. During the discussion at yesterday's session on repositories, one researcher noted the need for a "versioning server" and the facilitator, Karen Williams, referenced the Fried Foster/Gibbons D-Lib article from last year that noted the importance of versioning support in building an IR service that's relevant to faculty. As noted in this earlier article, DSpace, the tool being used to support WSU Research Exchange, does not currentliy provide seamless support for document versioning. Since DSpace is the most heavily-used IR tool being employed, there is a disconnect between the need for versioning and IR functionality.
  
  
• Lynch's discussion on federated searching is interesting in light of the growing importance of Google access to repositories. (At yesterday's session, I raised the info contained in Michael Organ's current D-Lib article, which documents that for one Australian repository, Google is the access method for 90% of document downloads; the repository's dedicated front-end is being used to support only 10% of downloads.)
  

Continuing with the scholarly communication sessions in a few minutes....