data for use in nationwide archival and retrieval systems are increasing.

• Providing standardized and systematic archival services for centrally stored, as opposed to distributed, data is far more effective, efficient, and less expensive.

During 1990, the USGS developed a prototype mass storage system. The design for the mass storage system focuses on a simple, straightforward solution to storing and retrieving very large files in a networked computing environment. The primary requirements of this design include the following: • A storage system that has potential for organizing, storing, and managing any set of data files, regardless of where the client is located geographically.

Virtually unlimited file storage capability (no system-imposed expansion limits). Initially this capability will approximate 1 terabyte (1012 bytes) of storage capacity that can be expanded simply by adding units of storage hardware.

• A very large file server accessible to any local area subnetwork via USGS telecommunications facilities.

Access to the very large file server by using simple file transfer, management, and storage system command structure and syntax. Regardless of the diverse computing environments of users, the connection between client and file server will be explicit. All data file transmissions to and from the server, as well as queries and data histories, will be activated by the user; the user will have full control over the archive process.

• Compatibility with various communication networks and storage hardware.

• Good file security mechanisms that allow the owner to control access to data.

The ability to share data files among diverse computer environments.

A storage server that is constant over evolving generations of client hardware and operating systems.

Accessibility, 24 hours a day and 7 days a week, at a low cost per gigabyte (109 bytes) of storage.

This mass storage system will eventually replace data archiving functions now associated with the existing mainframe computer. The system will embody all the advantages available in the current mainframe storage environment as well as vastly increase file storage capabilities, improve efficiencies in the use of file system hardware, have a new hierarchical library structure, and reduce storage

costs.

For the scientific workstation and distributed data-base activities in nonmainframe environments, such as UNIX, VAX/VMS, Macintosh, and MS/DOS microcomputers, this mass storage system will have filename,

directory structure, and command syntax conventions that are familiar to users of these environments, support common network standards and file transport facilities, allow sharing of data between these environments, and improve file security.

The mass storage system user, whether using UNIX, VAX, Macintosh, MS/DOS, or IBM-like mainframes, will learn one file server interface that will work the same for all systems. Standard commands will be available for connecting to the mass storage system; establishing or listing the contents of the user directory and detailed file descriptions and histories; storing and retrieving file(s) or hierarchical groups of files; adding, changing, and deleting files or directories; and invoking help commands. Through a file importexport capability, archived files can be stored in and retrieved from diverse system and storage media formats, and data sharing and disseminating can be made across various computing environments and data formats.

Mass storage system software for the existing mainframe and a robot tape storage device will be used to test a 1 terabyte mass storage system during 1991. A sample of representative data-base activities and scientific disciplines will be used to evaluate the system. This mass storage prototype will become the basis for developing the hardware, software, and basic data backup and archiving techniques necessary to handle the very large USGS data bases of the future.

Electronic Mail

By Paul Celluzzi

omputer electronic mail (Email) is fast becoming a vital part of everyday business life. Email, which has evolved from simple message sending to complex information exchange, provides an efficient method of transferring information in electronic form. This information may be simple text messages, complex documents, graphics, facsimile, or binary programs. In the future, Email will even include voice and video annotation.

Email is an effective method for moving documents, messages, and data and has improved information flow and strengthened communication within the USGS. Throughout. the bureau, scientists are recognizing the advantages offered by Email technology for communicating with colleagues and for conducting joint research from geographically dispersed locations.

The USGS is investigating technology that will allow information of all types to be

moved across different computer systems, application environments, and organizational boundaries. The diverse computer hardware and communication systems in use, however, complicate achieving a bureauwide system. Because each system has its own proprietary Email, implementing a single Email system is impractical. Instead, an alternative is to integrate the many separate USGS systems and connect to non-USGS systems. This approach gives each user the freedom to choose the Email system that best addresses local applications, provides for minimal disruption to existing user applications, preserves the investment in existing software, and avoids expensive retraining of users, who are able to continue to use their familiar local Email.

A central electronic Email switch will provide the necessary protocol conversion and mail routing when a user is sending information from one Email system to another. Also, a common central directory of user names and addresses will give USGS employees the ability to transfer electronic information to any user, regardless of computer system or Email software, simply by knowing the destination address. The sender will not need to know where the destination user is physically located, what computer system is being used, or on what Email system the destination user is registered.

Email is an end-user technology that, to be truly successful, must conform to design criteria that address user needs. The system must be universally accessible to users, allow transmission of different data types, be easy and inexpensive to use, be executable from within the user's local computing environment, and support interfaces to Email systems external to the USGS.

In the future, all USGS Email systems will support a standard convention for mail interchange that is part of a broader communication protocol suite known as GOSIP (Government Open Systems Interconnection Profile, discussed in "Open Systems Communication Standards," p. 88).

product, ADI, that allows access to Arctic digital data and information. The ADI hypermedia prototype is based on a printed journal format and, therefore, includes a series of papers listed in a table of contents (fig. 1). The reader can browse through the articles within the electronic journal by selecting an item from the table of contents or by choosing a link icon from within a section. By choosing the link icon, the reader then moves to related material. For example, the reader can begin searching through a data directory that is provided within the ADI (fig. 2).

While browsing the data directory, the reader can choose to link to the actual data set that may contain tabular numeric data (fig. 3A) or spatial imagery (fig. 3B). Upon reviewing the data set, the reader can then link to the full text of a journal article that provides analysis of the data. In addition to providing spatial data, such as side looking airborne radar data, the ADI includes image processing software that allows the reader to analyze data, such as the ice thickness of an Alaskan glacier.

ADI will be distributed to Arctic researchers and policymakers who use CD-ROM

Microcomputer

Graphic Presentations

By Brian Schachte

A

s a leader in the earth sciences and cartography, the USGS continually presents research projects and findings to the scientific community and to public audiences. Research itself has been greatly improved by new computing technologies, which enable earth scientists to analyze, model, and display complex data sets more quickly and efficiently than ever before. Presenting these results, however, is hampered by these same technological advances because, until recently, user-specific hardware and software could not be linked to communications and presentation materials or equipment.

Microcomputer animation and graphics software now make it possible to capture, manipulate, and enhance computer-based data. Connecting the microcomputer to various output devices produces illustrations

that can be carried easily for use at scientific meetings and at other presentations.

Postproduction printers, both color and gray scale, produce publication-quality illustrations. High-resolution color-image copiers produce high-quality paper and overhead transparency hard copies. Microcomputerbased film recorders produce 35-millimeter slides and 4- by 5-inch or 8- by 10-inch color negatives. Large-format posters can be made from these negatives. A series of images, output to a video cassette recorder (VCR), when combined with voice-over, music, and animated titling on a video tape, provides researchers and scientists with a presentation product that requires only a television and a VCR.

Microcomputer graphics at the USGS have produced several video cassette recordings, including one about national streamflow conditions and another about sedimentary basin analysis as it pertains to petroleum geology. These two videos contain a series of complex images that represent spatially oriented research projects.

Whether the final presentation product is paper or overhead transparency, 35millimeter slides or video tape, the use of microcomputer graphics makes it possible to demonstrate current research and results more easily and effectively.

Open Systems Communication Standards

By Paul Celluzzi

T

The diverse nature of the USGS mission requires a computing and communication environment to match this diversity. Over the years, different computer systems have proliferated and evolved independently of the issues of compatibility with other computer systems or an overall bureauwide computing strategy. There is now, however, an increased need to exchange information and to integrate applications.

The USGS situation is not unique; major compatibility issues exist throughout government and industry. In response, mechanisms are being developed to connect incompatible computer systems. A major initiative, sponsored by the International Standards Organization (ISO) and supported by contributions from national and international standards agencies, has resulted in an internationally accepted model for data communication and

standards that define the protocols for communication among different computer systems. These standards are referred to as open system standards because they are published in the public domain and have unrestricted availability. Open standards provide universal connectivity among computer systems.

The U.S. Government, under the direction of the National Institute for Standards and Technology, has adopted a set of specifications from the ISO standards that is collectively known as GOSIP (Government Open Systems Interconnection Profile). GOSIP defines a common method for government computer systems to communicate. Because GOSIP is published as a government FIPS (Federal Information Processing Standard), implementation of GOSIP is mandatory for all Federal agencies. A series of releases by the Federal Government will put GOSIP into effect.

The first release of GOSIP defines lower level communication protocols and file transfer and electronic mail applications. Later releases will include specifications for terminal login, Email directory services, compound document content architecture, and additional network options. Eventually, all government computer and communication systems will be required to conform to GOSIP. Government systems will then be connected and able to share resources.

The advantages of compatibility among computer systems are obvious; the USGS is now emphasizing the use of open systems protocols. A transition plan, in preparation, will convert the current USGS mix of proprietary and restricted protocols to GOSIP compliance. This approach initially provides for the use of existing protocols, then parallel and mixed protocols, and finally full GOSIP protocol to achieve a gradual transition having minimal impact on existing applications.

Eventually all USGS computer systems will be able to communicate and share resources regardless of the type of system. The GOSIP protocols will create a single standard for networking. Benefits to the USGS include a single standard that greatly simplifies the design, implementation, operation, and management of complex networks and reduces the costs of procuring and implementing the networks. Also, the investment in networks is protected from future changes because of the adoption of industry standard protocols. A growth path is established for changing and expanding networks. Finally, the resulting integration of diverse systems into a single manageable network provides a pathway to distributed processing and integrated applications across computer systems.