With support from the Andrew W. Mellon Foundation, seven institutions (Cornell, Harvard, the Massachusetts Institute of Technology, the New York Public Library, the University of Pennsylvania, Stanford University, and Yale University) have engaged in a project to create digital archives of electronically published journals. The planning phase of this initiative has raised a number of key issues that must be addressed, including: the types of e-journal archives and their interaction, the economic models for sustaining such archives, the collaborative agreements needed with publishers, the forms of access that archives will provide users, and the technical architecture of the archives. This session will summarize the key findings of the Mellon-funded projects and outline the directions that the Foundation is contemplating to address these issues.

With support from the Andrew W. Mellon Foundation, the Digital Library Federation and Outsell, Inc. are engaged in a multi-institution study that compares the information-seeking behaviors of more than 3,200 students and faculty across academic disciplines in liberal arts colleges and research/doctoral universities. This session will present preliminary findings and highlights of this research into how student and faculty behaviors and preferences are affecting library use and the demand for information resources.

Middleware is a layer of software services that manages electronic personal identity, security, access, and information exchange. Internet2, EDUCAUSE, and the Southeastern Universities Research Association (SURA) entered a three-year Cooperative Agreement with the National Science Foundation and the GRIDS Center under the NSF Middleware Initiative. As the Enterprise and Desktop Integration Technologies (EDIT) Consortium, the partners are working towards integrating middleware into campus enterprise environments for the purpose of advancing productivity for educators and researchers. This session will discuss the status of the project to date and future plans of the Initiative.

This briefing will describe the various components and building blocks being produced to create a regional digital library in the Glasgow area and how these fit into the emerging information strategy for Scotland.

Librarians at Wesleyan and Carnegie Mellon Universities are experimenting with two innovative ways to meet clients¼ needs for reference assistance. One uses staff members available for extended hours online, and the other uses software to mimic reference interviews and provide answers.

Wesleyan University Library, through a grant from the Davis Educational Foundation, is in the first year of a pilot project to test real-time collaborative reference service on the Internet for a group of liberal arts colleges, including Wesleyan, Connecticut, Smith, Wellesley, and Vassar. The two-year project included installation of a local server in the Wesleyan University Library and hiring staff to provide services to more than one institution during extended hours.

Carnegie Mellon University Libraries' statistics indicate that over 75% of the access to its online resources occurs outside of a library facility. A drawback of this kind of remote access is that users do not have traditional reference librarians to guide them to relevant and reliable material. Carnegie Mellon University Libraries is developing software known as the Automated Reference Assistant (ARA), which will attempt to mimic the reference interview online. It will elicit information about the users' research needs and guide them to pertinent material without human intervention.

Gathering the resources and people needed to create complex digital libraries often requires intimate collaboration between two or more institutions. This session offers briefings on two such collaborations: one creating a digital library of historical maps in Washington state and another developing a research portal for Civil Rights documents in Alabama.

The Early Washington Maps online collection (http://www.wsulibs.wsu.edu/holland/masc/xmaps.html) was built jointly by the Washington State University Libraries and the University of Washington Libraries. The core software package used to support this project, CONTENTdm, enables institutions to develop image, sound, and video collections in a collaborative manner. This session will present additional tools employed to build this digital library and some of the challenges faced by project participants.

The Transforming America Project (http://www.transformingamerica.org/) is a subject specific primary source research portal of Civil Rights collections in the state of Alabama. The project provides a one stop search for primary source Civil Rights documentation by creating a virtual repository of distributed archival finding aids and museum collection guides using hypermedia, XML, and Dublin Core-based descriptive metadata. The collaborative effort includes libraries, archives, and museums throughout the state. Participants include Auburn University Libraries, University of South Alabama, Trenholm State Technical College, Birmingham Civil Rights Institute, University of Alabama Libraries, Troy State University-Dothan, Rosa Parks Library and Museum, and the Birmingham Public Library.

The recent inclusion of seven new generic top-level domains in the Internet Domain Name System provided a basis for proving a number of concepts. One of the purposes underlying the establishment of the .museum TLD was to determine the viability of providing narrowly defined target communities with dedicated domains. A second purpose was to explore the potential of a controlled namespace for enhancing the value of the DNS for the Internet user community.

The protypal impetus that .museum might have towards the future creation of similar TLDs in adjacent communities, for example, .library and .archive, will be discussed from strategic, political and infrastructural perspectives. This will include a review of the activity needed to establish .museum and possible modes for extrapolating from it to enable a less strenuous expansion of the heritage management community's named presence on the Internet. A second discussion will address problems inherent in the conceptualization of a mnemonic device such as the DNS in terms of semantic potential.

ZING, Z39.50-International: Next Generation, brings together a number of current initiatives by Z39.50 implementors looking at making Z39.50 a more mainstream protocol and lowering implementation barriers, while preserving the intellectual contributions of Z39.50 that have accumulated over 20 years. This session will focus on one of these efforts, the development of Search and Retrieval for the Web (SRW), a proof-of- concept initiative to enable development of value-added search and retrieve applications such as the scholar's portal, that will integrate access to various networked resources. SRW, building on Z39.50 along with web technologies -- XML, SOAP/RPC, WSDL and HTTP -- recognizes the importance of Z39.50 for business-to-business communication, and focuses on getting information to the user.