CAMPUS-WIDE INFORMATION SYSTEMS Judy Hallman1 May 19, 1992 University of North Carolina at Chapel Hill INTRODUCTION Campus mainframe computers are the repository for a wealth of information, but faculty, staff, and students have not traditionally had easy access to this information. Most people have not had easy access to the computers; they did not know the commands needed to find information, nor did they have the security clearance to explore the data in the campus computers. While the increase in microcomputers on campuses and significant improvements in computer networking have facilitated access to mainframe computers, mainframe operating systems and the need for security for data have not changed significantly. What has evolved is a new set of tools for providing access to campus information, collectively called campus-wide information systems (CWISs). A CWIS provides campus information of general interest, online, accessible from virtually every workstation on campus that has communications capability. The systems are menu driven, but provide short cuts for experienced users. A typical system contains the campus directory, event schedules, guides, policies, and other documents. Several issues are emerging, including the following ones: -- A variety of software is in use. Their screen displays and menu presentations vary as do their means for navigation. What are the similarities and differences between them? -- How can the displays and access mechanisms be improved to help people find the information they need? -- How do people know that these services exist and what information they have access to? -- The definition of "public information" is being questioned; which data can be freely displayed, which should have restricted access, and which can and should be censored? -- Text formatted for the printed page usually won't display exactly the same on a computer screen. What guidelines should be followed to preserve the design of the authors and editors? -- Should data be changed after it enters the CWIS to keep it current or should it faithfully represent the source document? -- Who should be responsible for maintaining data in the CWIS--the authors and editors or the people running the CWIS software? And who should run the systems--computer specialists or librarians? -- How can CWISs be linked together so that people on one campus can easily access information on another campus? These are issues that require the combined skills of librarians and computer specialists. CWISs make a significant contribution to the availability of information. This paper first examines the emergence of CWISs as viable systems of information distribution, tracing the contributions of key individuals. Next, practical problems facing those starting a CWIS are addressed, followed by a discussion of possible future directions for CWISs. Appendices describe an electronic discussion group, called CWIS-L, which has been invaluable to those who provide CWISs, and how to try out some of the CWISs available on the Internet. CURRENT SYSTEMS What Is a CWIS? A CWIS provides campus information of general interest, online, accessible from virtually every workstation on campus that has communications capability. This definition excludes Student Information Systems, which provide students with personal information concerning their records. Similarly, it excludes visitor information systems that are not available on the campus computing network. Most CWISs require a userid, and some also require a password, but they vary in access control. Some CWISs only display information, some allow anonymous user input, some provide access to special data (such as medical records) to special users, and some provide access to all campus computing facilities and services. Most campuses provide a userid and password for visitors to look at information in the CWIS, but access to services, such as electronic mail, are limited to registered users. What Information Is Provided by a CWIS? The types of information provided by most CWISs include the following: -- event calendars, including sports, music/film/theater, lectures/seminars/workshops, academic calendars -- promotional/expository information, including books, pamphlets, newsletters, directories, primarily describing campus functions and services -- lists, such as job openings, available housing, class schedules, buy/sell bulletin boards, rides needed/offered -- grant and funding opportunities, current research activities, faculty publications -- access to the online library catalog -- e-mail/forums/discussion groups/digests. Further, some systems also include community information, an inclusion that has been subject to discussion. While some campuses feel their time is best used expanding the campus database, others see local events and activities as important to the campus community. Some campuses include bus/train schedules, weather forecasts, fortunes, community guides (including movie schedules and restaurant information), agricultural extension information, and news feeds. One of the most interesting services provided by CWISs is psychological counseling. Since 1986, Cornellians have been asking Uncle Ezra, an online counselor, for help with their problems. Questions are asked anonymously through electronic mail. Professional counselors then post both the questions and their responses in CUINFO where anyone can look at them. This service has been valuable to people with problems, and fun and interesting to people who use CUINFO. Cornell has even published a book called The Best of Uncle Ezra. Other campuses, such as Appalachian State University, have adopted this service.2 What Are the Benefits of CWISs? CWISs provide a variety of benefits, including the following: -- All the information is in one place. CWISs provide one-stop shopping for quick answers to simple questions, like how many students there are on campus or what time the basketball game starts. If you are already online, it might be faster to look up a phone number in the database than to find your printed copy of the phone book. -- Information is available around the clock. -- Information may be more current than printed copies. Campus directories are a good example. Online versions are likely to be updated frequently while the paper version may only be printed once a year. -- Information can be accessed from afar. While CWISs service the local campuses, they are especially valuable for remote users. Campus directories and course catalogs are likely to be used by people off-campus. When you get home from a conference and discover you want to discuss something with someone you met at the conference, you can look up his or her phone number and electronic mail address in that campus' phone directory. Similarly, you might want to see if a university has a course in a particular subject and who teaches it. -- Information can be accessed by more people. A CWIS expands the audience for newsletters, course catalogs, and even postings of events. For example, the University of North Carolina at Chapel Hill (UNC-CH) posts the publications of the Institute for Academic Technology and a list of library-related events of interest to librarians worldwide. -- CWISs have educational benefits. They are a lot like cookbooks; you may start off looking up one thing and end up reading a lot of unrelated information. It is easy to browse a variety of publications on the local campus and on other campuses on the Internet. -- CWISs can provide online archives. At UNC-CH, old issues of newsletters are kept in the system. -- CWISs have the potential for saving paper. It is reasonable to think that in the future, campus directories, class schedules, and course evaluations, for example, might only be available in electronic form. -- Having facts in a CWIS can reduce phone calls and save personnel time. How Did CWISs Get Started? It is difficult to determine exactly when CWISs got started or who had the first one. For example, mainframe logon messages, sometimes called the "message of the day," date back to the early days of mainframe computers and provide a crude announcement service. The University of Illinois at Champaign-Urbana provided information about their campus in the early 70s using PLATO, a system developed on that campus to provide self-paced instruction in a variety of subjects using touch-screen terminals.3 Stanford had a similar system in the 70s using SPIRES, software noted for its excellent full-text search capabilities and marketed by Stanford. In the 70s and 80s several campuses experimented with providing documents to users of mainframe computing services. However, Rita Saltz of Princeton questions whether anything can really count as a real CWIS prior to the micro revolution that put electronic access on the desktops of a large part of the campus.4 Certainly Cornell had one of the first CWISs as we know them today, championed by Steven L. Worona. According to Worona,5 CUINFO at Cornell began to take shape in Spring 1982. Two commands were made available on their IBM computer system: ROSTER to display the University's course roster and EXAM to display the final examination schedule. Worona notes that these schedules "were invariably obsolete even before they reached students' hands. Since both of these documents were prepared for publication on the central computer system and regularly updated (although not reprinted), the possibility for electronic access to current information was tantalizing." The CUINFO command (installed in October 1982) was created as a generalization of EXAM and ROSTER. As Worona points out,6 at the time these CWISs began, campus computers were used for number crunching and data analysis and the users were technically oriented. CWISs expanded the audience to all members of the university community, including many who never had reason to use a computer before. Several CWISs came into being during 1986- 87. What Software Is Currently in Use? CWISs run on a wide range of computers (often hardware already on campus) and the software is still evolving. Like the evolution of word processing software, the process will probably result in some outstanding systems that will dominate, establishing consistency in menus and ways of moving around. Right now there is no clear standard. Current systems present similar menus, but vary in navigational methods, resources they access, and the difficulty in installing data. The basic concepts of these systems are essentially the same. The user is presented with a menu and chooses an item by entering its number or part of its name. Some choices lead to other menus. Eventually the user reaches text that is displayed one screen (or page) at a time. Users can move forward or back, one page at a time, they can return to the last menu displayed, or they can return to the initial (main) menu. Short cuts are provided for moving directly to information of interest, and indexes or maps tell users the keywords to use in the short cut method. How menu choices are made, the commands and key sequences for paging through text, and the short cut methods vary. The following describes systems currently in use with the design goals of some of their developers. Three samples of opening menus are included to show different menu presentations. See APPENDIX B for information on how to obtain a file that tells how to connect to CWISs on the Internet, what the systems contain, who to contact for more information, and for an easy way to browse through some of the services. CUINFO: According to Steve Worona of Cornell, CUINFO (written in IBM system 370 assembler language) was originally developed for use under the VM operating system that runs on IBM mainframes. While CUINFO could not be easily transported to other operating systems, "a complete re-implementation would not be a formidable task."7 Yale also runs CUINFO.8 Here is Cornell's opening screen (February 1992): C U I N F O - Cornell University's Electronic Information Source MAIN MENU Select... For items such as... NEWS Weather, Announce, Grad. Bulletin, Safety Reports, Updates EVENTS Calendars, Athletics, Colloquia, Theatre, Music, Movies DIALOGS Drug IQ Network, Mr. Chips, Uncle Ezra, Auntie Em, Suggest SERVICES Computing, Support, Housing, Transport, Food, Careers, Jobs, OEO, Directories, Volunteer, Library Or select a more general category like... ACADEM Information about Cornell instruction and research ADMIN Items from and about Cornell administrative offices ITHACA General information for and about the Ithaca area Or type the name of any specific CUINFO entry such as WEATHER, DIRECT, or EZRA. (Type INDEX for a list of available items, or DETAILS for descriptions.) Please select a topic or type HELP for hints. (Blank to exit.) --> Worona describes the CUINFO user interface as follows: The CUINFO interface is based on activities with which everyone is familiar -- selecting from a list and flipping through pages -- and we have yet to encounter a person who doesn't immediately understand the system. What we find more and more surprising is the wide variety of information that can be distributed effectively by this mechanism. The key is taking the trouble to organize the data before putting it online. The approach of many information retrieval systems is to provide the user with powerful searching tools, utilizing keywords, free text, boolean logic, and anything else the designer can throw in. These tools are then applied by the user to a data base whose internal organization is unknown or invisible. In this context, powerful tools are almost a necessity. But the rules for using powerful tools tend to be difficult to learn and easy to forget. The standard solution for this problem -- online "HELP" -- is not so much a solution as proof that the problem exists. CUINFO takes the opposite approach. The searching tools are extremely simple, but therefore usable by anyone. The burden is on the information providers and system administrators to structure their data in such a way that it can be readily processed by these simple tools.9 VTX: VTX is a product of Digital Equipment Corporation (DEC). It runs on DEC VAX computers and is available in the Education Software Package. Campuses using this software include Appalachian State University, Clemson University, New Mexico State University, North Carolina State University, Pima County Community College, University of New Hampshire, University of North Carolina at Chapel Hill, University of North Carolina at Greensboro, University of North Carolina at Wilmington, and Western Carolina University.10 One special feature of VTX is the ability to link services together using DECNet . The University of North Carolina uses DECNet to link together the services at Appalachian State University, North Carolina State University, the University of North Carolina at Chapel Hill, the University of North Carolina at Greensboro, the University of North Carolina at Wilmington, and Western Carolina University.11 VTX also supports voice output using DECTalk.12 Navigation in VTX relies on use of function keys and some key sequences. New Mexico State University modified the user interface allowing users to type commands (like Q for quit). Several of the campuses that use VTX have applied the modifications from New Mexico, but these modifications are not supported by DEC. Further, they only work for VT100 terminal emulation, not for users emulating IBM 3270 terminals. DEC is addressing these problems in a new version of VTX, version 5.0. PNN: Howard Strauss is responsible for the development of Princeton News Network (PNN). Princeton runs it under VM/CMS, UNIX,13 and Hypercard (on Macintoshes). All versions are available to universities at no cost. Several campuses use this software, including: Arizona State University's PEGASUS and ASEDD (under VM/CMS), Northwestern University, Notre Dame (under VM), San Diego State University (on a Sun SPACstation 1 running SunOS), and University of Colorado at Boulder (running on both VMS and UNIX).14 Strauss says that the major design goal for PNN was to provide an intuitive, user-friendly interface for a variety of platforms. For example, to scroll forward to the next page, F8 works (for the CMS users), PgDn works (for the PC users), abbreviations of FORWARD work, tabbing to the word FORWARD displayed on the command line and then pressing return works, and entering the number 22 to scroll ahead 22 lines works. Macintosh users can use scroll bars and they can use the mouse to point to commands.15 Indiana's AIE: Unlike CUINFO, VTX, and PNN, Indiana University's Academic Information Environment (AIE) provides access to computing resources.16 Indiana wrote their own software (in Pascal) to run under VMS. They paid particular attention to the user interface; the screens were designed by a standards team consisting of a graphic artist, an editor, a programmer, and a CAI consultant.17 Here is the AIE menu (March 1992): Keyword:TOP Wed, Mar 4, 1992 IU Academic Information Environment IU Scholars' Network [1] News & General Information [9] Library Services [2] How to Use the AIE [10] Administrative Services [3] Computing Info & Help [11] Academic Support Services [4] Where to Compute [12] University Life [5] VAX Files & Utilities [13] Instructional Computing [6] VAX Applications Software [14] Network Information & Services [7] Microcomputing Services [15] Miscellaneous Services [8] Mail & Message Services [M] Mail [H] Help [C] Comments [$] VMS Command [A] Addresses [K] Keywords [B] Back [V] Exit to VMS [I] IU Events [O] AIE On/Off [N] Next [Q] Quit/Logout >> News has been updated. To read it, type AIENEWS. << To enable/disable the AIE at login, select the [O] On/Off option. Type 1-15, option, or keyword > MUSIC/SP: Some campuses, including McGill, Lafayette, and Lehigh, are using software called MUSIC (Multi-User System for Interactive Computing/System Product) for their CWISs. MUSIC was developed at McGill University in Montreal, Canada, and is distributed by IBM.18 According to Roy Miller, one of its developers, MUSIC "includes a full-function CWIS system that supports menus, full-text keyword searching, direct updating of sections by the data providers, electronic feedback from viewers to data providers, and access to external systems through TCP/IP."19 TechInfo: MIT developed TechInfo which runs on a MacPlus with one megabyte or more of memory, System 6.0.3 or better, and licensed MacTCP drivers. Source code is available free. Campuses running this software include Mississippi State University and the University of Pennsylvania.20 Tim McGovern says that one of the design goals of TechInfo was to distribute the responsibility for content to the information providers. Thus, MIT designed tools that information providers could use over the network to enter new information and to maintain information.21 Other systems: There are several new systems under development that utilize client/server architecture to present information so that users can navigate through it using the tools they have on their workstations. Each database is presented by a server, and the user's interface to the databases is controlled by client software run on the user's workstation. Thus the user interface is different for each type of client and is natural to the user. Linking Information Systems There are at least three implementations of software that facilitates logging onto remote services: LIBTEL (a UNIX script), LIBS (an implementation that runs under VMS and UNIX), and Hytelnet (a hypertext system for UNIX and VMS). These products provide menu access to libraries, CWISs, and other information systems worldwide. You simply select the service you would like to connect to and the software makes the connection for you. You do not need to know the Internet address for the service or the appropriate terminal emulation (VT100 or TN3270). With current implementations, once you connect to a service, the way you navigate is determined by the service you are using. Perhaps before long, there will be a single user interface to similar services, so that online catalog searches, for example, are done using one command language, regardless of the software used by libraries being searched. APPENDIX B describes how to use the LIBTEL software at UNC-CH. The Internet Gopher developed at the University of Minnesota uses client/server architecture. Gopher is described by Mark McCahill, of the University of Minnesota, as a distributed document delivery service that allows access to documents residing on multiple hosts. Users can look at documents from menus or prepare queries and receive a list of documents that match the search criteria. You can access the gopher system as a VT100 client via Telnet.22 Another approach to linking campus systems is the Campus-Wide Information Systems Protocol (CWISP), described by Strauss as a protocol (based on Z39.50 standards) that will allow users of a CWIS to access data of other CWISs while preserving the user interface of the local CWIS.23 However, Strauss says that deciding to use the Z39.50 protocol added complexity to CWISP and it was taking a long time to implement . Meanwhile Gopher appeared on the scene and has become the defacto CWISP. He says that work on CWISP is on hold.24 WAIS (Wide-Area Information Server) is a protocol, evangelized by Brewster Kahle of Thinking Machines Corp., that allows users to perform searches of multiple databases at remote sites worldwide. Queries are simple English sentences or phrases. Matches are assigned usefulness values (based on criteria such as the number of matches in the selection) and listed at your workstation by value. Once a good document is found, you can use relevance feedback to find more documents like it, and you can then easily retrieve items of interest on your workstation. Currently, text and graphics can be retrieved; inclusion of sound and full-motion video is anticipated in the near future.25 WAIS also utilizes client/server software. Unlike current implementations of Gopher, WAIS uses the Z39.50 communications protocol. Perhaps one solution to the CWISP problem would be to implement Gopher in WAIS. Another approach to presenting information is the wide- area hypermedia initiative called World Wide Web (WWW or W3). Initiated at CERN, the European Particle Physics Laboratory in Geneva, Switzerland, W3 displays vary depending upon the type of workstation you are using. While for a VT100 it displays text with imbedded numbers and if you enter one of the numbers you get more information, for more sophisticated workstations it displays multifont text; linked text is underlined or boxed as a button and you can point-and-click to follow the link. Another function of W3 is an index search. According to Tim Berners-Lee of CERN, "Some documents are tagged as 'index cover pages'. When you get one of those, you can do a search by typing in keywords (like WAIS)."26 Further according to Berners-Lee, other systems, like WAIS and Gopher, "can be represented as a combination of hypertext links and searchable indexes."27 Here is the WWW opening screen (April 1992) as displayed for a VT100 terminal:28 The World Wide Web project (23/50) WORLD WIDE WEB The WorldWideWeb (W3) is a wide-area hypermedia[1] information retrieval initiative aiming to give universal access to a large universe of documents. General Project Information See also: an executive summary[2] of the project, Mailing lists[3] you can join, Policy[4] , latest W3 news[5] . Project Status[6] A list of project components and their current state. (e.g. Line Mode[7] ,X11 Viola[8] , X11 Erwise[9] , NeXTStep[10] , Daemon[11] ) People[12] A list of some people involved in the project. History[13] A summary of the history of the project. How can I help[14]? If you would like to support the web.. Technical details 1-26, Back, Up, for more, Quit, or Help: Regardless of the particular software used, we can expect to see more information accessed directly on the computer on which the data is maintained; there will be less copying of information to change its location or format. OPERATIONAL ASPECTS Who Initiates the CWIS? Early systems were defined by imaginative programmers (like Steve Worona of Cornell and Howard Strauss of Princeton) who observed the difficulty of getting information to the people who need it, when they need it. They were encouraged and supported by the people responsible for the information. Those who came later liked what they saw and wanted to provide similar services on their own campuses. On some campuses (North Carolina State University, UNC- CH, Northwestern University, for example), implementation of the campus information system was initiated by a high-level university administrator. On many campuses, implementation of the service was driven by information providers wishing to get specific information, such as calendars of events and class rosters, to students, faculty, and staff in a timely manner. Getting Support for the CWIS In an interesting discussion on CWIS-L about the driving force behind the implementation of a CWIS, Strauss said that "developing a system is often the easy part. Getting anyone to use it or support it can be orders of magnitude more difficult." He started a list of criteria for a CWIS,29 which was enhanced by contributions from Worona30 and Terry Mathias of Southern Illinois University at Carbondale.31 They made the following points: -- Try out a few CWISs before developing a new one. -- Start by defining the services your users want and then finding the tools to provide those services. -- Make sure that the CWIS contains the information people want to see and that it is presented clearly. -- Make the system entertaining and have things that people want to know so that they will use the system. If people use the system, it can also inform and educate them. A daily police blotter, for example, is compelling, informative, and educational. -- Don't overlook useful information outside the university. CWIS users have lives that extend outside their offices and classrooms. -- The CWIS must be easy to access, for example, from public terminals in a variety of locations around the campus. -- A CWIS needs an assertive, resourceful person who can tap a variety of campus information structures. Who Maintains the CWIS? Most CWISs are operated by the campus computer centers. Northwestern University is an exception; there the library is responsible for the CWIS. According to Brian Nielsen, Assistant University Librarian, "The Library more or less allied with administrative computing, arguing that current surplus capacity on an IBM mainframe should be utilized to build a more centralized utility; the Library argued that some 'editorial responsibility' was required to make a CWIS that was easy to use by any student/faculty member/staff, and agreed to take on such responsibility itself."32 The work associated with maintaining the database is distributed between the information providers and the organization responsible for the CWIS itself. On some campuses the information providers install the data themselves; on others, the organization responsible for the CWIS installs the data. Advisory Committees Many campuses have established committees to assist in establishing policies. At UNC-CH, the original steering committee consisted of a student, a computer science faculty member involved in text research, a journalism faculty member with experience with videotex systems, three librarian/programmers (one of whom had experience with a CWIS), and the three people from the computing center responsible for developing the system. The steering committee met several times during the planning stages of the service, helping to determine policies and define the menu structure. Some of the outcomes were: -- The committee reviewed menu proposals, tested the prototype, and recommended significant changes. -- The committee recommended a topic orientation instead of an object orientation; subjects from books were used as menu items rather than having the books themselves as menu items. -- The committee helped establish the CWIS management position as publishers, not writers or editors. Policies Timothy J. Foley (Lehigh) has described the need for a clear computing and information policy that addresses possible legal liabilities, censorship, resource management, information ownership, and user responsibilities.33 Policies need to be established for determining what campus information is considered to be public information. Some campuses are having difficulty obtaining permission to install in the CWIS some information that appears in print on campus: campus directories, for example. Advisory committees can be extremely helpful in this area. Policies also need to be established for how the information will be organized and how the work of installing information is divided between the information providers and the organization responsible for the system.34 Some campuses have documents describing policies and procedures for entering data into the CWIS.35 Cost Computer hardware and software costs vary widely. Many campuses run their services on hardware that was already on campus; some have acquired new, relatively inexpensive workstations. Some campuses use free software, others use proprietary software, and some develop their own software. The majority of the cost of a CWIS is personnel time to monitor the database; add, delete, and change information in the database; train information providers; acquire and add new sections to the database; and to keep university faculty, staff, and students informed about what is in the database and how to use it. Putting Information into the System The following aspects of entering text into the database are addressed by the information providers and/or the service providers: -- Converting the information from its current form to a file on the CWIS machine The majority of the information in a CWIS comes from a computer file, but the difficulty in moving it from one system to another will vary depending on the software and hardware used on each end. Computer programs may be needed to convert the formatting codes used by the originating word processor to something useful on the CWIS machine. Note that this step could be eliminated in the future by CWIS software capable of accessing information directly on the machine where it is maintained. Some information, such as announcements for special events, may come to the service providers as phone calls or printed announcements and need to be keyed into the system. -- Preparing the file for display on a computer screen Files prepared for the printed page usually need reformatting for a computer screen. The first steps are to change the font and character size so the characters are monospaced (all the same width) and the size they will be on a computer screen, for example, Courier font at 12 points, and to change the line length to something like 65 (screen displays are usually a maximum of 80 characters). The next step is to create an ASCII file; that is, one with no special characters in it. There usually is an option in the word processing software to do this, but some do a better job than others. (You may have seen e- mail messages that have a capital R instead of an open quotation mark and a capital S instead of the close quotation mark.) Regardless of how it is done, bullets, em dashes, en dashes, true (curled) quotation marks, and accent marks should be changed to standard keyboard characters. Tabs can present a problem, depending upon the word processing software used. When saving an ASCII file, some software inserts blanks to preserve spacing, but some leave the tab character. When displayed on the user's terminal, tabs at best will not provide the spacing the editor intended. Tables may be too long for the computer screen and need to be split. Pie charts and bar charts need to be retyped as tables. Sometimes tables or forms are omitted from the CWIS and readers are referred to printed copy. Footnotes often need to be relocated within the text. -- Changing menus Menu structures need to be evaluated and changes made, as appropriate. -- Adding highlighting Some CWIS software provides highlighting features, such as bold, flashing bold, and inverse video. While it can be useful to highlight headings and footers, highlighting has its hazards; how it looks on a particular screen depends on the terminal and the terminal emulation software. For example, at UNC-CH, inverse video is used to display the title and date of the source document in the bottom line of the screen. However, many terminals on campus do not have inverse video capabilities, and the source identification in the footer is often difficult to distinguish from the text itself. -- Installing the data and checking it Procedures are followed to install the data in the database and check each screen. At UNC-CH, a lot of time is spent converting files from their original form to the format for the database, adding highlighting and the commands to actually install the data, and checking it after it is entered. There was a time when the time-consuming nature of this activity at UNC-CH was questioned. Then in an e-mail message to CWIS-L, David Millman (Columbia) expressed curiosity as to how Howard Strauss and Rita Saltz handled the day-to-day maintenance of their CWIS,36 and Rita Saltz (Princeton) responded with her famous "groan" message, describing how her staff takes information in whatever form they can get it in and installs it in their database.37 In an interesting discussion of ways to put weather reports in CWISs, several campuses admitted to calling weather stations, taking notes, and typing in the reports. But Gary Anderson (Virginia Institute of Marine Science, College of William and Mary) described automatic procedures for obtaining weather and updating their CWIS several times a day. He noted that "The 4am posting has been especially helpful to those about to put out to sea for a day of fieldwork!"38 In fact, most campuses have developed some programs and scripts to aid in installing data. Strauss strongly encouraged CWIS managers to work on the maintenance problems, saying that the problems of maintaining a CWIS can be solved, and that "Arizona State University is running a version of PNN and has highly automated the update procedures."39 CWIS Managers as Publishers Most who provide CWISs think that their task is to gather existing, edited and reviewed publications and present the information in a useful fashion. One question that arises regards how much fixing should be done before information is put in the CWIS. Should spelling and punctuation errors be corrected? Should printed publications be changed to keep them current? When a publication is put in the CWIS at UNC-CH, the name and date of the publication is provided in a footer that appears on each screen. At the beginning of a document, a note is usually included containing at least the department and a phone number of the department responsible for the document. As a matter of policy, the documents in the database contain exactly the same information as the printed documents. If a document is changed, permission is sought from the editor of the document and/or a note is included explaining what was changed and why. UNC-CH has a rather unique problem of multiple sources for information because publications are divided and the information is organized by subject. For example, there is overlap between The Undergraduate Bulletin and the student handbook, called The Source. The information providers for both documents agreed that information in The Undergraduate Bulletin takes precedence over that from The Source. Thus, if you look at the section on "Grades" under "Academics," you see the section from The Undergraduate Bulletin. Taking Information Out of the System Most campuses have addressed the need to remove information when it is out of date; old information looks bad and takes disk space. In many cases, the software allows specification of an expiration date that is carried with the text but hidden from the viewers, and the text is automatically deleted after the specified date. But there are some interesting issues associated with removing information, particularly events, from the system. It is not always clear how much text should be associated with the expiration date; for example, should a sports schedule (such as a football schedule) be treated as one text that is removed at the end of the season or should each event be removed the day after it happens? If an event spans several days, do you take it out after the first day or the last? Exhibits that span several days should be kept in the system through the last day, but information about workshops is not valuable through the last day because people who missed the first day probably cannot participate. For seasonal events, should the entry remain on the menu after the last event of the season, before the next schedule is available, or should the menu be changed to remove the selection? If the menu selection is left in the system, what message should the user see when there are no more events in the season? At UNC-CH, events are removed the day after they happen, menu choices are left for schedules even if no events are listed, and the message "There are no events to post at this time." is inserted by the software that removes events when it takes out the last in the schedule. Designing the Menus Like many campuses, UNC-CH began designing its system by planning the menus. Existing systems were examined, items that might go into the database were identified, and a prototype containing just the menus (no data) was built . The advisory committee tested the prototype. The main menu contained too much text; there were fourteen items on it, each with a long description of what was in that area. It did not invite people to use the system. UNC-CH initiated its service with a menu that had jobs and grant/funding information under "News" and academics, student life, and services and facilities under "Ask INFO about..." Before long, categories much like those in the prototype were moved back to the main menu, because they most clearly represented the information in the database; however the items were shortened and presented in two columns. The main menu for a CWIS presents political, practical, and design challenges, as many people want their information presented on the main menu. It is very handy to have an advisory committee to share in the decision-making process when questions arise about design, particularly the main menu. Issues of breadth versus depth also challenge designers of CWIS menus. In the early days of its service, UNC-CH worked toward a set of menus designed to include all information on campus that might become available, but items were added to the menus only when the information became available in the database. Thus, users would sometimes have to move through menus of only one or two items each to get to the information they wanted. Even after a couple of years of development, there wasn't enough information available in the database to warrant the number of levels of menus, and users didn't seem to know what information was available to them in the database. After about three years of operation and continued development, UNC-CH redesigned its menus building them for the data already in the system or clearly on its way. The depth of menus was reduced, often by putting a large number of items in one menu. When developing a new service, some campuses present full menus before all of the information is available. If a user selects an item that is not in the system, they get an appropriate message. Some campuses highlight information that is available. For example, both Appalachian State University and Western Carolina University use inverse video to highlight the item numbers of items in the system. Another consideration associated with designing menus for a CWIS is how to construct menus for a large publication. Is it better to show the full Table of Contents, with the subheadings in a long menu that may exceed one screen, providing menu choices at logical points (more like hypertext), or should the menus be hierarchical, one menu of the major headings, and then several menus of subheadings? If menus are subject-oriented (as they are at UNC-CH), how do you combine those from several publications into a meaningful menu? One last note on menus: If newsletter articles are presented as a menu, the newsletter editors should re- evaluate titles of regular monthly articles, such as "Recent and Relevant" and "Feature of the Month," to provide more informative titles. How Much User Documentation Does a CWIS Require? A CWIS should require minimal documentation. Users need to know how to connect to it, how to move around, and what information is available in the database. Indiana University has paid particular attention to documentation needs. According to Marsha Snyder Waren,40 Indiana's initial Academic Information Environment (AIE) came with easy-to-use menus, online help, and 115 pages of paper documentation (describing, for example, how to search the online card catalog and how to use Dow Jones News/Retrieval). AIE's documentation was then reduced to a handful of how-to sheets. Waren continues to evaluate CWIS documentation needs. In October 1991, she asked CWIS-L, "Do you encourage your users to survive on electrons alone? Or do you provide something on paper to get them started more comfortably?"41 Responses indicated that several campuses provide indexes or maps as part of the service itself, and some have an area in the CWIS describing new additions to the service. Some campuses also provide starter and overview information on paper. Feedback: How Useful Is It? Most systems provide mechanisms for recording frequency of access to specific pages. Some CWISs (for example, Lehigh) provide usage statistics as part of the database. But how are usage statistics used? If segments of the database are not used, does that mean the information is not useful or that people don't know the information is in the database or that people can't find it? In these years when CWISs are still new, should CWIS managers take out information just because it isn't being used? There are still many people on campuses with CWISs that don't know the CWIS exists. As these services become more common and more useful, more people will learn about them. When a person looks for information in a campus database, they should find it, even if it is information that is not accessed very often. On the other hand, statistics do indicate what information is not being accessed so its placement in menus can be checked and advertising provided for it, if appropriate. One way CWIS managers find out what the users want is from phone calls and e-mail messages. At UNC-CH, if jobs are not posted on time, users call and complain. People call when something is wrong, and they often send e-mail describing what they would like added or changed. FUTURE DIRECTIONS The future for CWISs promises to be exciting. Sophisticated ways are being developed to link information systems; services are being expanded off campus to the community; and the idea of the library as the focal point for computer-based information systems is gaining force. Linking Information Systems While text-searching capabilities across databases worldwide, as provided by WAIS, present attractive alternatives to menu-driven information systems, they do not exclude the desirability of menu-driven systems. Some people will want to explore menu-driven systems, while people who know what they are looking for are more likely to want to use text-searching facilities. Hopefully, the information in CWISs worldwide can be linked together for text-searching inquiries as well as being available via traditional menus. For example, high school students should be able to search university catalogs electronically. They should be encouraged to go to their public libraries or computer work rooms in their schools to access campus information systems and look at eligibility requirements, course offerings, student activities, housing, and tuition costs as they begin to select the colleges they might want to attend. Similarly, it would be nice to be able to move easily from menu mode to text-searching mode and back again when looking at information. Consider the following scenario: A user works through some menu items and finds a section of interest on a local information system. The user then issues a command that says, "Show me more of these." If, for example, continuing education courses in public health are being displayed, the system might respond: "Would you like to see more continuing education courses in public health?" If the user responds "Yes," then text-searching software would search directories and present the user with a list of other selections. As more information systems are linked together, it will be increasingly difficult to tell where the information you need is physically located, but where it is will become less important, as long as you can get to the information electronically. Unless information providers are careful in how they identify information when they put it in databases, users may find a piece of text appropriate to their needs, but have difficulty identifying the source document. As Daniel Updegrove (University of Pennsylvania) pointed out in a message to CWIS-L, if everyone is committed to designing systems as Internet resources, there is a lot of information that could be shared. However, he also noted that content might need to be adjusted to provide more complete information, for example, adding areas codes to phone numbers.42 Further, information services will become more closely linked on the campus level. Some campuses already have one main menu for accessing mainframe computing services, the campus library, and the CWIS. Hopefully in the near future users will be able to make queries that would cross boundaries of information systems, such as the library holdings and the campus information system. This won't happen unless it is planned for and services using standard protocols like Z39.50 are developed. Expanding Toward Community-Wide Information Systems Campus-wide information services are expanding into the general community in two ways: campuses are encouraging the use of their services by alumni and prospective students, and campus personnel are helping to establish community information systems. Parents of currently-enrolled students, alumni, prospective faculty and students, and journalists would benefit from access to campus information. Electronic mail services would be particularly valuable for parents of currently-enrolled students. Electronic mail, bulletin boards, and news groups can be particularly valuable to alumni. Some communities have established local public information services. These are free services offering electronic mail, discussion groups, and ask-an-expert services, as well as accessibility to event schedules, directories, brochures, and the like. These systems are run by the communities themselves, with volunteers doing most of the work, but most are affiliated with a local university. For example, Cleveland Free-Net is sponsored by Case Western University, Cincinnati's TriState Online service is affiliated with the University of Cincinnati, Peoria's Heartland Free-Net is affiliated with Bradley University, and Youngstown Free-Net is affiliated with Youngstown State University. These universities provide campus information to the community system. Several more communities are in the process of establishing similar systems, and some campuses are moving toward putting their campus information systems into the community systems. There are several different models of community information systems. The Public Electronic Network (PEN) offered to the residents of Santa Monica, California, has received much attention. PEN is based on the UNIX operating system and software called Caucus. It was introduced to the public (local residents only) in February 1989.43 Several communities are using the Cleveland Free-Net model, and Free-Net users have formed National Public Telecomputing Network (NPTN), similar to National Public Radio. NPTN will assist in the development of community information systems and disseminate information features in areas such as health, education, government, and law to its affiliates. According to literature provided by NPTN, Cleveland's system began "in the fall of 1984 when Dr. Tom Grunder, then of CWRU's Department of Family Medicine, set up a single phone line, computerized, 'bulletin board' system called 'St. Silicon's Hospital and Information Dispensary' to test the effectiveness of telecomputing as a means of delivering general health information to the public."44 The Free-Net software is written in the C programming language and uses the UNIX operating system.45 Most of the community services are accessible from the Internet. APPENDIX B describes how you can explore them. Roles for Libraries Librarians should have central roles in expanding CWISs. While on most campuses the CWIS was implemented as a computer service, it is becoming increasingly clear that the task of overseeing CWISs belongs in the campus libraries. Librarians have traditionally gathered information, organized it, provided tools for locating it, and helped people find the information they need. Many of the issues facing providers of campus information have already been addressed by librarians; censorship, freedom of information, prohibition of keeping records of access by individuals, for example. It makes sense for computer centers to do the research and development work, to evaluate hardware and software, to develop tools for entering text, and to improve computer networking interfaces, but computer specialists are not well prepared for establishing policies regarding access to information. While campus libraries and computing centers have traditionally worked well together, there appears to be a need for closer interaction between the two. Libraries are also in a better position to provide free computing for CWISs than most computing centers are. During the last few years many libraries have implemented their own computing systems and services and now have the resources to support the CWISs. Meanwhile computing centers typically charge for services and find it increasingly difficult to justify providing information services free when at the same time charging for other uses. Libraries traditionally provide free services. Libraries are logical places to provide terminals for accessing campus and community information systems; people traditionally take their questions to the libraries. To help patrons use these services, librarians will need to continue to expand their knowledge of electronic information services. Librarians are already providing information to their users through campus and community information systems. These systems provide the mechanism for innovative ways to provide library reference services and information about special collections.45 Further, community information systems will need people to run them. Most of the community information systems currently in place have at least one full-time staff position to coordinate the effort. These positions seem well suited to the skills of librarians. The librarian's Strategic Vision Steering Committee could have been thinking about CWISs when they drafted the following vision statement: Strategic Vision for Professional Librarians Establish the basis for librarianship in the 21st century in: SERVICE By selecting and delivering information that users need at the point and moment of need By creating and maintaining systems which provide accurate and reliable information By promoting the design of information systems that require little or no learning time for effective use By correctly analyzing users' questions and providing them with the information they need (which may not be reflected accurately in their questions) By educating users to manage information By initiating contact with potential information seekers to ensure a widespread understanding of professional services available to them, including assistance for those who do not wish to use the library independently By furthering the development of the "virtual library," a concept of information housed electronically and deliverable without regard to its location or to time LEADERSHIP By taking responsibility for information policy development, information technology application, environmental awareness, information research, and risk-taking in making strategic choices in the information arena By accepting accountability for the information services we provide By identifying and collaborating with strategic partners and allies in information delivery . . . 47 These aspirations certainly reflect the requirements for making CWISs more valuable services in the future. APPENDIX A CWIS-L LISTSERVER An electronic discussion group, called CWIS-L, has been invaluable to those of us who provide CWISs. When we have a question or a problem or are thinking about trying something new, we post a question to the list. During the next few days we receive advice from our colleagues worldwide and can adjust our plans based on the experience of others. The questions and responses can uncover aspects and legal considerations we hadn't thought about. Participants have been free with advice and take the time to compose messages to help others. CWIS-L has provided much of the information for this paper. I subscribed to the list in January 1990 and since then have logged many of the messages I found particularly interesting. It has been fun to reread them while preparing this paper. Excerpts from several messages are included here. Further, members of the list have provided extensive help in preparing this paper. Steve Middlebrook of Washington University set up the electronic discussion group, or "listserver," called CWIS-L in January 1990. CWIS-L has been maintained by Timothy Bergeron (c09615tb@wuvmd.bitnet) of Washington University since October 1990. The stated purpose of CWIS-L is as follows: This list is for discussing the creation and implementation of campus-wide information systems. The term CWIS includes systems which make information and services publicly available on campus via kiosks, interactive computing systems and/or campus networks. Services routinely include directory information, calendars, bulletin boards, databases and library information. 48 Entering its third year of operation in January 1992, there were over 1,000 subscribers. The discussions on this listserver are also available in the Usenet news group bit.listserv.cwis-l, so there are more uncounted participants. To subscribe to this list, send to LISTSERV@WUVMD.BITNET or LISTSERV@WUVMD.WUSTL.EDU, the following message: SUBSCRIBE CWIS-L yourfirstname yourlastname For example, SUBSCRIBE CWIS-L Judy Hallman. Messages sent to CWIS-L@WUVMD.BITNET or CWIS- L@WUVMD.WUSTL.EDU are sent to everyone signed up to the list. If you reply to a message from CWIS-L, be aware that your reply will be directed to the entire list, not just to the sender of the message. Messages are archived; see your local computer people for information on how to use archives. APPENDIX B TRYING OUT SERVICES Besides providing the conduit for electronic mail, the Internet allows us to connect to CWISs worldwide. Campuses that have systems up and running provide information on how others can access their services via the Internet and try them out. For people who are designing new services or contemplating major changes, it is really helpful to be able to explore someone else's design. For people planning new services, it is also valuable to be able to try out the underlying software using someone else's implementation and see what the capabilities of the software are from the user's point of view. During the summer of 1990, I started gathering a list of services and occasionally posted it to the list. Art St. George (University of New Mexico), co-founder and maintainer of the Internet Library Guide, asked me to maintain this section for his publication. With the help of others on CWIS- L, I have done so. The first list was published in the Internet Library Guide in Fall 1990. Entries tell how to connect to the service, the hardware and software used, a contact person, and special items of interest in the system. You can pick up this list by anonymous ftp to ftp.oit.unc.edu; it is in pub/docs/cwis-l. See your local computer people for information on how to use ftp. LIBTEL software makes it easy to explore CWISs; it displays menus of services and then will connect you to the service you chose. You can try LIBTEL at UNC-CH using the Extended Bulletin Board Service (EBB). Telnet to ebb.oit.unc.edu, logon as ebb. (Note: You cannot use EBB if your terminal is emulating an IBM 3270-type terminal.) You will receive some informational messages and be asked for your name. From the EBB menu, select item #9 (Libraries and information systems). If you then select USA, you will be asked to choose the state in which the service you want resides. After you select a state, you will get a menu of services offered in that state. For example, if you select NY, you can select CUINFO at Cornell. The EBB software will then make the connection for you, if you so choose; you do not need to know the Internet address of the service you want to use. Further, EBB provides the proper terminal emulation for you (telnet or tn3270).49 NOTES 1. (hallman@unc.bitnet). Note that e-mail addresses can change. Those listed in this paper were valid in April 1992. 2. For an extensive list of information provided by CWISs, see: Rita Saltz (rita@pucc.bitnet). "Report of CWIS-L discussion session at Snowmass," Campus-Wide Information Systems (cwis-l@wuvmd.bitnet 7 Aug, 1991). The message contains the working notes from the August 5 workshop session, "Operating and Maintaining a Campus-wide Information System," at the General Directors' Seminar/22nd Annual Seminar on Academic Computing, held at Snowmass Village, Colorado. 3. David Lassner (david@hula.oit.hawaii.edu). "RE: Was CUINFO the first," Campus-Wide Information Systems (cwis- l@wuvmd.bitnet, 14 Jan, 1992). 4. Rita Saltz (rita@pucc.bitnet). "Was CUINFO the first," Campus-Wide Information Systems (cwis-l@wuvmd.bitnet, 13 Jan, 1992). 5. Steven L.Worona (slw@cornella.bitnet). "An Informal Overview of CUINFO (Cornell's Computer-Based Bulletin Board System)," (no date). 6. Ibid. 7. Ibid. 8. The contact person for CUINFO is Steve Worona (slw@cornella.bitnet). 9. Worona. 10. There is a listserver for VTX, VTX-L@NCSUVM.BITNET. 11. Users on one UNC campus simply select another campus from a menu, as shown in the following menu from the UNC-CH system: To connect directly (via DECNet) to the information service on another University of North Carolina campus, select one of the following services: 1. Appalachian State University, VideoText System 2. North Carolina State University, Happenings! 3. University of North Carolina at Greensboro, Minerva 4. University of North Carolina at Wilmington, SeaBoard 5. Western Carolina University Information System For information on how to connect to information systems worldwide, select the following item: 6. How to connect to information services via the UNC-CH campus extended bulletin board Commands: Help Quit Main Backup Find Enter menu choice number or command: There is no noticeable degradation in response time when looking at data on another service regardless of the distance between campuses (for example, Appalachian State University and UNC-Wilmington are more than 400 miles apart). Eventually we hope to have all sixteen of the University of North Carolina campuses on this link, making it easy for us to share information. Besides access between UNC campus services at the main menu level, DECNet allows connection directly from any menu choice or by special "global" keywords. For example, a person on the Chapel Hill campus who selects "Publications" from the main menu and then selects the electronic publication "Postmodern Culture," will view the publication installed in the North Carolina State University database in Raleigh. 12. Appalachian State University is currently using this feature which, according to Ernest Jones, "provides voice responses to an Alumni job opportunities file." Ernest Jones (jonesel@appstate.bitnet). "RE: CWIS and Voice Service," Campus-Wide Information Systems (cwis-l@wuvmd.bitnet, 19 Dec, 1990). 13. UNIX is a Trademark of Bell Laboratories. 14. The contact person for PNN is Rita Saltz (rita@pucc.bitnet). Direct system and development questions to Howard Strauss (howard@pucc.bitnet). 15. Howard Strauss (D8897@pucc.bitnet). "User-friendlyness of CWIS," Campus-Wide Information Systems (cwis-l@wuvmd.bitnet, 18 Oct, 1990). 16. Indiana's AIE is generally only available to registered users, but it is possible to obtain special permission to look at AIE by contacting Pete Percival (percival@iubacs.bitnet). 17. Caroline Beeb. "Friendly but Streamlined: Implementing an Integrated University Computing Environment," Proceedings, ACM SIGUCCS User Services Conference XVIII, 1990, pp. 31-34. 18. Direct inquires to Roy Miller (ccrmmus@mcgillm.bitnet). 19. Roy Miller. "Comments on your CWIS paper," E-mail message to hallman@unc.bitnet, 13 Mar, 1992 20. The contact person is Tim McGovern (tjm@mit.edu). 21. Tim McGovern (tjm@eagle.mit.edu). "Re: Manpower estimates and CWIS data," Campus-Wide Information Systems (cwis- l@wuvmd.bitnet, 4 Jan, 1991). 22. Mark P. McCahill (mpm@boombox.micro.umn.edu). "new release of PC gopher client software available," Campus-Wide Information Systems (cwis-l@wuvmd.bitnet, 19 Sep, 1991). 23. Howard Strauss (D8897@pucc.bitnet). "CWISP," Campus-Wide Information Systems (cwis-l@wuvmd.bitnet, 19 Oct ,1990). 24. Howard Strauss (D8897@pucc.bitnet). "Status of CWISP???," e-mail message to hallman@unc.bitnet, 8 Apr, 1992. 25. A WAIS bibliography, prepared by Barbara Lincoln (barbara@think.com), is available for anonymous ftp from /pub/wais/wais-discussion/bibliography.txt@quake.think.com. The bibliography covers WAIS documents, WAIS-related articles/publications, and electronic services. Hard copies are available from Barbara Lincoln (barbara@think.com), Thinking Machines Corp., 1010 E1 Camino Real, Suite 310, Menlo Park, Ca, 94025, phone: 415-329-9300, fax: 415-329- 9329. 26. Tim Berners-Lee (timbl@info.cern.ch). "RE: Please check info for a paper on CWISs," e-mail message to hallman@unc.bitnet, 6 May, 1992. 27. Ibid. 28. The contact person for World Wide Web is Tim Berners-Lee (timbl@info.cern.ch). He says, "There's more information in the web. You can pick up line mode, NeXT or beta test X browsers by anonymous FTP from info.cern.ch -- also you could pick up a paper from the same node: /pub/www/doc/ENRAP_9202.ps for example, or a version of the W3 "book" from /pub/www/doc/the_www_book.ps.Z ( or .tex for LaTeX version)." 29. Howard Strauss (D8897@pucc.bitnet). "Re: Drive Force Behind Implementing a CWIS," Campus-Wide Information Systems (cwis-l@wuvmd.bitnet, 28 Sep, 1990). 30. Steven L.Worona (slw@cornella.bitnet). "Re: Drive Force Behind Implementing a CWIS," Campus-Wide Information Systems (cwis-l@wuvmd.bitnet, 2 Oct, 1990. 31. Terry Mathias (ge0515@siucvmb.bitnet). "driving forces in CWIS creation," Campus-Wide Information Systems (cwis- l@wuvmd.bitnet, 2 Oct, 1990). 32. Brian Nielsen (bnielsen@nuacvm.bitnet). "Re: Drive Force Behind Implementing a CWIS," Campus-Wide Information Systems (cwis-l@wuvmd.bitnet, 27 Sep, 1990). 33. See: Timothy J.Foley (tjf0@ns1.cc.lehigh.edu). "Developing a Campus Computing and Information Policy: Issue and Concerns," CAUSE/EFFECT Vol. 14, no. 4, Winter 1991, pp. 25-29 and 33. See also "Developing a Computing & Information Policy," Proceedings, ACM SIGUCCS User Services Conference XVIII, 1990, pp 127-130. 34. As part of her work toward a Masters in Communication and Information Studies at Rutgers, Hannah Kaufman, of Princeton, has prepared a report "Campus-Wide Information Systems: Factors Affecting Willingness to Contribute Information." A summary is available from her via e-mail to hannahk@pucc.bitnet. She notes, for example, that people seem to contribute "to provide information to a different (not larger) group" and because of "a belief in online information." 35. Kansas State University, for example, has a document describing policies and procedures to be used for including information in their system, called UNICORN (UNIversity Central Online Resource Network). The document includes instructions for users to prepare and install their own information. For more information, contact Betsy Edwards (betsy@ksuvm.bitnet). UNC-Chapel Hill has a similar document; contact Judy Hallman (hallman@unc.bitnet). 36. David Millman (dsm@cunixf.cc.columbia.edu). "Forced e- mail," Campus-Wide Information Systems (cwis-l@wuvmd.bitnet, 2 Oct, 1990). 37. Rita Saltz (rita@pucc.bitnet). "Day-to-day maintenance of PNN (groan)," Campus-Wide Information Systems (cwis- l@wuvmd.bitnet, 3 Oct, 1990). 38. Gary Anderson (gary@ches.cs.vims.edu). "Re: Weather," Campus-Wide Information Systems (cwis-l@wuvmd.bitnet, 19 Jul, 1991). 39. Howard Strauss (D8897@pucc.bitnet). "Re: Non-University Info," Campus-Wide Information Systems (cwis-l@wuvmd.bitnet, 11 Jul, 1991). 40. Marsha Snyder Waren. "How much Paper Do You Need to Support an Electronic Information System,?" Proceedings, ACM SIGUCCS User Services Conference XVIII, 1990, pp. 355-360. 41. Marsha Snyder Waren.(waren@unixpop.ucs.indiana.edu). "Paper maps of CWISs.," Campus-Wide Information Systems (cwis-l@wuvmd.bitnet, 24 Oct, 1991). 42. Daniel Updegrove (updegrove@dccs.upenn.edu). "CWIS as Internet resources?," Campus-Wide Information Systems (cwis- l@wuvmd.bitnet, 13 Apr, 1990). 43. Articles discussing PEN include: Gayle Hanson. "Making Waves Via Computer," Insight, 27 Jan, 1992 and Pamela Varley. "Electronic Democracy," Technology Review, Edited at the Massachusetts Institute of Technology, November/December 1991. 44. From literature from NPTN. For more information about NPTN and Free-Net software, contact T. M. Grunder, Ed.D. - President, National Public Telecomputing Network, Box 1987, Cleveland, Ohio 44106. Voice: (216) 368-2733. FAX: (216) 368-5436. Internet: aa001@cleveland.freenet.edu. 45. Ibid. 46. Anne K. Abate and Rosemary Young. "Community Access Bulletin Boards: Cincinnati Librarians Become Involved." special libraries (spring 1992): 113-117. 47. Sue Martin (skmartin@guvax.bitnet). "Draft Vision statement," Public-Access Computer Systems Forum (pacs- l@uhupvm1.bitnet, 17 Jan 1992). 48. Timothy Bergeron (c09615tb@wuvmd.bitnet). "Re: Info on CWIS," Campus-Wide Information Systems (cwis-l@wuvmd.bitnet, 7 Nov, 1990). 49. The Extended Bulletin Board is a development project of the Office of Information Technology, University of North Carolina at Chapel Hill, and is experimental. Services are often expanded, modified, or dropped as appropriate, and no service-level guarantees are made to the users.