The Navy Lab/Center Coordinating Group's Computing Committee, and the Librarian of the Navy, have tasked LMI with determining the requirements of this community for scientific and technical information discovery and delivery. One of the challenges of providing STI for this community is the fact that the particular service a user might require is differentiated by a combination of factors: the discipline in which the user is working, the stage in the RDT&E process which is being supported, the timeliness required in response, the importance of a definitive answer, and the budget for the required piece of information. There are about 15,000 people in this community that require some level of STI support. Given the extremely variable nature of the STI support required, a system which provides all types of service regardless of user location at a range of service levels and prices, is required.

To provide such a service, an alternative to the current site-specific STI systems is needed. The growing availability of software for, and ubiquity of, the Internet make this a key component of such an alternative STI system. With the ability to distribute information regardless of geographic location, and the buying power of the entire community, a wider variety of information resources can be made available on demand. When the system is designed under a client-server architecture, site-specific or discipline-specific interfaces and/or collections of information resources can be made available through client customization.

The World Wide Web software meets the requirements for the user interface and Internet compatibility, and the COTS client software can be provided to all users at little to no cost. Further development is needed in the area of user identification and security, and interfaces to remote databases. However, current trends in the online information industry favor such development. The next six to twelve months also happens to be the best time to negotiate wide-area user licenses with online database providers, as the industry has been taken by surprise by the surge in Internet popularity, and new pricing models have not yet stabilized.

We recommend that the NLCCG pursue the development of the NDVL using the Internet as a backbone, the World Wide Web as the application environment, and a rapid prototyping development methodology. The first-stage prototype should be deployed in six months with a sample of information resources, while licensing negotiations with online information vendors proceed. The beta version of the NDVL should be tested with users from every site and refined over a second six month period, with full deployment accomplished by the end of calendar year 1997.