n o oL r\ U D P I E ^ if Museum and University Data, Program and Information Exchange NATIONAL SCIENCE COMPUTER NETWORK The National Science Foundation has announced (July, 1972) an expanded research program relative to the development of a national science computer network, which is described in announce ment number NSF-72-16. The following paragraphs are taken from that announcement, since they are particularly pertinent to any future exchange of data, programs, and information among museums and universities, and should therefore be of interest to MUDPIE readers. "BACKGROUND.--For several years the National Science Founda¬ tion has been sponsoring research generally applicable to the development ot a National Science Computer Network. Such a net¬ work would provide its users with access to computing facilities, science information systems, and other computer-based resources on a regional and national basis. A National Science Computer Network would have profound implications for resource sharing in research and education. Facilities and resources of higher quality than might otherwise be the case could be available, and the duplication of some facilities would be eliminated. In particular, a national network would offer the possibility of integrating computing and science information services in ways to strengthen the Nation's research and education programs.. "...A comprehensive and interrelated set of project activi¬ ties will be undertaken to permit exploration and evaluation of the many dimensions and facets of a nationwide network. Spec¬ ial attention will be given to collaborative efforts, independent of institutional affiliation, for development of innovative and effective approaches to resource sharing....The expanded program will provide for a trial National Science Computer Network uti¬ lizing existing computer communications technology and computer resources to support these network activities.... "...(The) network is envisioned as permitting from every sector of our society to participate at and without Federal subvention... A series of discu preliminary proposals is suggested which will perm tion to gauge the community's sense of direction, and requirements. "USER FUNCTIONS:... (include)... design and use of large-scale information and data systems,...collaborative efforts on computer based research methodology, including, for example, advanced computational techniques and hierarchical computing and informa- institutions minimum cost ssions and it the Founda- interests, - 2 - tion services,...developing, computer-based instructional approaches to effective user tional computing activities. testing and systems and integration w evaluating specialized materials,...innovative of research and educa- It is apparent from this series of abstracted statements that NSF is working along the same lines as long since visualized as profitable by MUDPIE participants. I have, therefore, prepared a statement of what MUDPIE is all about, and have submitted it along with other pertinent materials to the Foundation, as a preliminary statement of interest. If the Foundation indicates a wish to explore further with us, we should be prepared to do so. Some of you have been associated with MUDPIE since its in¬ ception, and I know your interests, activities, and areas within which you would be willing to participate. Others among you have done little more than to ask to be put on the list for the newsletter, and I cannot say very much about your interests. Since we are now embarking on what might be the first steps to¬ ward implementation of a functioning network, I need to know what each of you might be willing to do in the way of participation. If you are already using time-shared computing, this might be a simple matter of indicating a willingness to switch to a mutual contractor so the network could immediately become a reality. If you are currently doing nothing, because your institution cannot afford a teletype connection for you, how much would you need to take this first step, if some help were forthcoming? If you are already well involved in time-sharing, are you inter¬ ested in some kind of supervisory role in a planned network, to aid in development and growth (a MUDPIE Council?). All of this should in no way interfere with the direct, individual participation of anyone in the proposed NSF program. I would like to know about other projects that will be suggested to NSF and submitted for support, and I will be glad to have information for future MUDPIES concerning such projects. Keep in mind that MUDPIE now has about 300 readers, among whom may be several who might be of value in any project.—JAP MUSEUM AND UNIVERSITY DATA, PROGRAM AND INFORMATION EXCHANGE The Museum and University Data, Program change had its origin in 1967, as a result of at several institutions in the use of time-sh museum research and in the potential for deve museum network for sharing information. The attention has been on work in systematic and but much of the work done to date has had wid throughout biological research, and interest (hereafter referred to as "MUDPIE") has been nal emphasis has been widened to include asso investigation such as ecology. and Information Ex- the growing interest ared computers in lopment of an inter¬ primary focus of taxonomic biology, espread application in the exchange such that the origi- ciated areas of . Basically, MUDPIE represents a network, at present held to¬ gether by its newsletter. The plan has been, since the inception of MUDPIE, to take advantage of commercial time-sharing computer facilities for direct exchange of ideas and uses of the computer as designed and implemented by the participants. There are at the present time many users of commercial facilities among museums, with fewer, perhaps, in the universities. The latter is due to the fact that universities have moved more strongly in the direction of in-house availability of computer laborato¬ ries for use by the professional staff in their research. Museum scientists, however, found commercial time-sharing a blessing, since it provided them with access to computers in a fashion usually not previously available to them. In many cases, museums have been exchanging programs and data files which are stored in local facilities, creating a considerable duplication and added expense that could easily be eliminated through the use of common availability to all museum users of a centralized memory storage capability, as would be provided by a commercial contractor with nationwide access to his central computer. This is, of course, presently possible, since many contractors offer whole or partial service to the nation, and on several occasions participants in MUDPIE have already taken advantage of this potential. The American Museum of Natural History and the United States National Museum had mutual access to each other's stored data and programs for a considerable period while both were contracting with the company CEIR, located in Washington, D.C., and the arrangement was discontinued only when the company re¬ vised its corporate structure, and both museums moved to new contractors. The National Museum and Michigan State University have had mutual storage access through the GE time-sharing sys¬ tem, which is national in scope, but rather expensive for most museum users. Demonstrations of the capacity and potential of time-sharing in museum work have been made in all parts of the country, using either the connection of the Smithsonian with Dialcom, a D.C. company, or with GE, through Michigan State University. It has been a basic assumption for MUDPIE that the only material stored for full-time, mutual access in the network would be that for which use is highly repetitive. There are many collec tions of facts or organized sets of data that are constantly used by biologists in their day-to-day work or research, and most of these are amenable to computerization and on-line storage, given sufficient storage space. Among these are programs which permit standard calculations of biometric parameters and statistical correlations; taxonomic data matrices and programs for processing them; standard tables of biological data; answers to repetitive queries from the public; and many others. Progress in several of these areas has already been made by workers participating in MUDPIE, and functional programs, matrices, and answering services have been tested and are working. To take them up in turn, and in detail ♦ * 1. FUNCTIONAL PROGRAMS. Approximately 60 programs written for use by biologists in general and systematists in particular have been listed in various issues of MUDPIE. All such programs are available from the programmers, but at the present time must be sent through the mails as paper or magnetic tape or as a print-out. In addition, the Smithsonian has published a set of programs correlated with Simpson, Roe and Lewontin's book on quantitative biology. The set of programs (copy attached) has been distributed to all MUDPIE participants, and has also been added to the library facilities of several time-share com¬ panies, including one in Australia (used by the National Museum in Sydney). The Smithsonian has also published a program to permit analysis of biogeographical data in such a way as to distinguish biotically distinct areas, which, although original¬ ly designed to permit systematists to work with collections, has proven to be of use in pollution work, and has been adopted by environmental investigators. These are the kinds of programs, highly repetitive in their use, which should be centrally stored for use by biologists in all parts of the country, rather than individually stored on literally dozens of independent discs. During the time of mutual sharing by the American Museum and the National Museum, costs to each were reduced through the mu¬ tual access to programs prepared by the other, and a national network would correspondingly reduce costs for all those participating. 2. TAXONOMIC DATA MATRICES. The concept of storage of highly organized and coded taxonomic data has been fully devel¬ oped by MUDPIE participants. The matrix, as currently designed, will permit storage of any and all kinds of data pertaining to a taxonomic group, regardless of size or variation of the group. It is organized in such a way that programs written either in BASIC or FORTRAN can access the data. It compresses an enormous amount of data into a comparatively small space, and permits rapid retrieval in many different forms. For example, one of the most repetitive activities of a systematist is the identification of unknown specimens both in the course of his own work and as a service to nonsystematic users of taxonomic data. The taxonomic matrix can be used as a random entry key to process information concerning the characters of unknowns, taking advantage of the computer's speed in surveying the entire range of knowledge concerning all taxa concerned. It is a swift, efficient, and highly accurate method (within the limits of the accuracy of the data as originally stored) for eliminating one of the most time consuming tasks of systematics. In addi¬ tion, the same matrices are already being used for preparation of dichotomous keys which are printed by the computer; the print¬ ing of diagnoses or descriptions of taxa; preparation of contrasts between or comparison of two taxa, giving differentiating and similar characters; the preparation of "peek-a-boo" random entry field keys on punched cards; quizzing of students on their under¬ standing of taxonomic groups; organization of groups showing • ^ ;1 5L t d (