The New York Public Library has developed and implemented facilities for input and display of data in nonroman vernacular scripts in its Automated Bibliographic Control System. Hebrew has been the first nonroman script implemented. The input conventions were derived from an American National Standards Institute standard for Computer Compatible Transliteration of Hebrew. Practices regarding the use of nonroman text and character sets were derived from bibliographic practices employed by the Library of Congress. This paper attempts to describe how problems of input and display within the context of a computer-based system were solved. In addition it presents an innovative approach to filing nonroman entries in a catalog containing mixed scripts, which offers a solution to the problems posed by romanization.
Two innovative uses of MAchine Readable Cataloging (MARC) data whose realization is imminent are the reporting of library holdings to the National Union Catalog (NUC reporting) and the multilateral interchange of bibliographic information (MARC redistribution). These two uses are found to place very similar demands on the data required for input. Some of these demands are novel (or at least newly recognized) and imply a necessity to rework certain aspects of the definition of the MARC format. A brief outline and discussion of some of the necessary modifications to the definition are presented.
The methodology of cost-effectiveness analysis described in this report was developed in order to evaluate the automation of Colorado State University's library circulation system. The analysis was based on data from a time study of the circulation department operations. Unit costs for the old manual system and the new semi-automated system were found to be $0.365 and $0.474 respectively, but the new system was more effective in some respects.