By David DormanAmerican Libraries Columnist
ddorma@ltnet.ltls.org
By David Dorman
American Libraries Columnist
ddorma@ltnet.ltls.org
Library consultant for the Lincoln Trail Libraries System in Champaign, Illinois.
Column for October 2002
The Gileadites in Biblical times had a straightforward way of testing identity and dealing with the results: If you couldn’t correctly pronounce “shibboleth” you would be killed on the spot, otherwise you could continue on your way and cross the Jordan River. The university folks who are bringing us Internet2 are developing, with IBM support, a modern version of the “shibboleth” test in the form of an identity protocol. The bad news is that this modern version is a much more convoluted process of determining identity. The good news is that the consequences of failing the test are much gentler.
Commercial authentication protocols typically require a person seeking access to positively identify him (or her or it)self to the target’s gatekeeper before access is allowed. Privacy in the form of anonymity is just not a consideration in the commercial authentication world. The academic and library communities have different values, however, and they felt that an access authentication protocol was needed that would allow the user’s anonymity to be preserved. The result is Shibboleth which defines an intricate exchange of messages among the user, the server of the organization the user belongs to—what the protocol geeks call the origin server—and the remote server guarding the target data.
One feature that supports privacy is that the origin server can vouch for the identity of the user without giving the target server any identifying information. Another feature enables a user to tell the origin server what information about the user should be released to the various target servers that the user has to deal with.
Shibboleth will use OpenSAML, a set of open-source Java and C++ program libraries that are SAML-compliant. SAML stands for Security Assertion Markup Language. Think of SAML as an XML data type definition (DTD) for exchanging security information.
As the academic community begins to implement Shibboleth, order will be introduced into what has up to now been a very chaotic authentication marketplace. Library systems will be faced with how they want to implement the protocol—as a target, as an origin, or both. Whatever choice is made, it is a good bet that any serious player in the academic ILS marketplace will need to support Shibboleth in some form in the not-too-distant future.
ProQuest has finished digitizing the New York Times, back to its first issue in 1851, and the Wall Street Journal back to its first issue in 1889. Every issue of both newspapers has been digitized from cover to cover, including news, editorials, photos, graphics, and advertisements. Basic and advanced keyword searching is available, or the issues can be browsed page by page.
“This was an unprecedented conversion effort, comprising 3.5 million pages and over 30 million articles,” said ProQuest President and Chief Executive Officer Joe Reynolds. “Newspapers present unique challenges for conversion, given their large page format, multiple article types, varying lengths, and page jumps. To accomplish the conversion, we pioneered new techniques in digitization, zoning (identifying areas of relevant text and relating them to each other), and image quality enhancement. The result is a fully searchable file that allows users to view articles in their original context.”
The two backfiles are available in the company’s ProQuest Historical Newspapers service on an annual subscription basis. To request a free trial, visit the Proquest Web site.
The National Information Standards Organization (NISO) released a new draft standard for library statistics in July. “The revised standard addresses an area critical to the information community: measurement of electronic resources,” said Vivian Campagna, NISO’s program officer.
The body of the standard contains definitions of e-resource formats, collections, and services. Three appendixes are especially valuable for librarians designing or collecting statistics on e-resource use: “Appendix A: Methods of Measurement,” “Appendix B: Measuring the Use of Electronic Library Services,” and “Appendix C: Measuring Public Library Networked Services: Preparing Your Library to Collect Network Statistics.”
The standards document does a commendable job of elucidating all of the factors that go into gathering statistics on e-resources, even if it doesn’t have definitive guidelines. The document is on the NISO Web site. If statistics is not your thing, at least check out Pat Harris’s short but informative article on “Standards: Myths and Realities.”