lib-s-mocs-kmc364-20141005045627 257 Technical Communications ANNOUNCEMENTS ISAD Institute on Bibliographic Networking Information Science and Automation Division (IS AD) of the American Library Association will hold an Institute in New Orleans on February 28-March 1, 1974 at the Monteleone Hotel in the French Quarter. The subject of the institute will be "Alternatives in Bibliographic Net- working, or How to Use Automation Without Doing It Yourself." The seminar will review the options available in co- operative cataloging and library networks, provide a framework for identifying prob- lems and selecting alternative cataloging systems on a functional basis, and suggest evaluation strategies and decision models to aid in making choices among alterna- tive bibliographic networking systems. The institute is designed to assist the participant in solving problems and in se- lecting the best system for a library. Methods of cost analysis and evaluation of alternative systems will be presented and special attention will be given to comparing on-line systems with micro- fiche-based systems. The speakers and panelists are recog- nized authorities in bibliographic network- ing and automated cataloging systems and will include: James Rizzolo, New York Public Library; Maryann Duggan, SLICE; Jean L. Connor, New York State Library; Maurice Freedman, Hennepin County Li- brary, Minneapolis; Brett Butler, Informa- tion Design, Inc.; and Michael Malinconi- co, New York Public Library. The cost will be $60 for ALA members and $75 for nonmembers. For hotel reser- v~tion information and a registration blank, write to Donald P. Hammer· ISAD; American Library Association; 50 E. Huron St.; Chicago, IL 60611. P.S. Mardi Gras is February 26! ISAD Forms Committee on Technical StandariU for Library Automation (TESLA) The Information Science and Automa- tion Division of the American Library As- sociation now has a Committee on Tech- nical Standards for Library Automation (TESLA). TESLA, recently formed with the ap- proval of the ISAD Board of Directors will act primarily as: a clearinghouse fo; technical standards relating to library au- tomation; a focal point for information re- lating to automation standards; and a co- ordinator of standards proposals with ap- propriate organizations, e.g., the American National Standards Institute the Electron- ic Industries Association, N~tional Associ- ation of State Information Systems. The committee's initial work will be to formulate areas and priorities in which standards are required, to document exist- ing standards sources, and to develop a "library" of applicable standards to be drawn upon by the membership of ALA. According to the new committee's chairman, John Kountz, California State Universities and Colleges, "It is auspicious that this time be selected for the imple- mentation of a standards committee for library automation. With the current in- troduction en masse of production library automation systems and the fading of re- search and development activities, such standards will come into good use as they may be developed for library automation. In addition, the close linkage with new developments such as the Information In- dustries Association and the availability of standardized data bases, hardware, and communication standards are becom- ing requirements. The standards which shall be emphasized in the committee ac- tivities are those relating to areas of inter- est- for administrators and automators 258 Journal of Library Automation Vol. 6/ 4 December 1973 alike. These standards are intended to fill the void for future library automation op- erations." The committee efforts should be mea- sured in terms of facilitating the automa- tion of library functions as required on an individual library basis. Information relat- ing to the standards committee activities and its scope, or general information re- lating to library information technical standards, should be addressed to: ALA/ ISAD Committee on Technical Standards for Library Automation, John Kountz, Chairman, 5670 Wilshire Blvd., Suite 900, Los Angeles, CA 90036. Formation of an Ad Hoc Disetl$sion Group on Serials Data Bases As a result of an informal meeting held during the ALA Conference in Las Vegas to discuss the problems associated with the establishment and maintenance of union lists of serials, an Ad Hoc Discus- sion Group on Serials Data Bases was formed, with Richard Anable acting as in- terim coordinator. The Council on Library Resources agreed to fund a meeting of the group's steering committee on September 21, 1973 at York University in Toronto, Canada. Many of the major union list ac- tivities on this continent will be represent- ed as well as the National Libraries and ISDS National Centers from both Canada and the United States. A list of the subgroups that have been formed gives a good idea of the individu- al problem areas which the group is tack- ling: a. Record format comparison b. Minimum record data element re- quirements c. Cooperative conversion arrangements d. Organizational relationships and grant support e. Holding statement notation f. Bibliographic standards g. Authority files h. Software evaluation and exchange A detailed description of the history and activities of the Discussion Group can be found on page 207 of this issue. For further information contact: Rich- ard Anable, York University, Downsview, Ontario, Canada, M3J 2R2, ( 416) 667- 3789. TECHNICAL EXCHANGES File Conversion Using Optical Scanning: A Comparison of the Systems Employed by the University of Minnesota and the University of California, Berkeley By this time most large libraries in the U.S. have converted into machine-read~ able form at least some of their IDes. Most of them, however, have used relatively in- efficient techniques (such as key-punch- ing) or relatively expensive ones (such as on-line data entry). It was with pleasure, then, that I read Ms. Grosch's recent ar- ticle ("Computer-based Subject Authority Files at the University of Minnesota Li- braries," Journal of Library Automation, Dec. 1972) describing a conversion tech- nique that she, like the library at the University of California at Berkeley, has found to be extremely cost effective, namely optical character recognition using a CDC 915 scanner. Berkeley has used (and still is using) this technique in its efforts to create what will soon be among the largest machine- readable serials files of any university in the world. That me currently contains records for over 50,000 serials (in the MARC structure). It is expected to con- tain records for about 90,000 unique ti- tles (approximately 30 million characters) before the end of the current fiscal year. Based on our eJ~:perience in this undertak- ing, I would like to offer the following comments on the use of the CDC 915 scanner as it is used in Minneapolis and in Berkeley. Costs-It should be crystal clear that the main reason for using the scanner is cost of the keyboarding device. That is, the keyboarding device for the CDC 911S scanner is an ordinary ten pitch Selectric typewriter which can be purchased for under $500.00 or rented for from $II.OO to $30.00 per month. When not used as a computer input device the machine functions as a normal office typewriter. A device like an MT/ST that rents for about $110.00 a month costs about $.60 an hour for every hour it is used, or ten times as much. Keyboard operators for a typewriter are easily obtained since there is no need to train an operator in the idiosyncrasies of keypunch cards, CRT terminals, magnetic or paper tape devices, etc. Keyboarding is fast and easy, especially when com- pared to a key punch. Mistakes are easily corrected by, for example, merely crossing out the character ( s) in error. Keyboarding on a Selectric for a scan- ner and keyboarding on a device like the MT/ST both require a "converter" (the scanner itself or the MT/ST-to-computer- tape converter) . These "converters" are equally available and the decision to use one keyboarding device over another should not hinge on the "availability" of such "converters," as is usually the case. In addition to selecting a cost-effective keyboarding device, Minnesota has also operated a system that delivers the data to the keyboarding device in an efficient manner: the typing is done from the source document itself, rather than from a copy of that document that has been transcribed onto a "coding sheet" or a photocopy of that document. Ms. Grosch points out that photocopying the source document would have raised the project costs by about 50 percent. In addition, keyboarding from photocopied documents would probably have been much slower and less accurate. The Berkeley typists al- so keyboard from the original document, even when that document is a public cata- log card that must be temporarily marked up in order to resolve ambiguities for the typists. Supplies-It is true that the ordinary Selectric typewriter (without the pinfeed platen) performs satisfactorily. Thus, one does not need continuous forms for the typewriter. Indeed, it is not necessary even to use a "stock form"; plain 20 pound white long grain paper will do. We use Zellerbach's Hammermill Bond 820, which costs $2 a ream. At Minnesota, using this paper instead of the "stock form" would probably have reduced the supplies cost from $400 to less than $25. Technical Communicat!ons 259 Had a keypunch been used, the operation would probably have required about $150 worth of IBM cards. Scanner Throughput-Careful design of the format of the data on the typed sheet can substantially improve through- put on the CDC 915 scanner. With dou- ble spaced typing (three lines per inch), the CDC scanner is capable of reading data at the rate of over a half million characters an hour, or about twice as fast as was actually achieved at Minnesota. Thus, with altered design of the input format, about half of the cost of the "con- verter" -the scanner-could· have been saved, representing an additional savings of $500. The principle applied to maximize throughput on a scanner such as the CDC 915 is to enter as much data as possible on a line and as many lines as possible on a page without crowding the data so much as to cause the machine to misread. (The machine enforces stricter tolerances as its capabilities are pushed to their lim- its.) One wants to get as much as possible on a line for the same reason that one wants to get as much as possible onto a punched card: there is a fair amount of machine overhead involved in advancing to the next line and/or page. The Berkeley system uses a sheet of pa- per that is 8}~ x 14 inches in size, and the typists type each line a full 63~ inches long. Typing is double-spaced (even though the machine is capable of handling single-spaced typing) because this in- creases the vertical skew tolerance from ~ of a character height to a full character height. Figure 1 is an example of a page typed at Berkeley. At Berkeley, more than one field may be placed on a line, each field being sep- arated by the "fork" character (Y) . Like Minnesota, typists identify each field by a one-character code at the beginning of the field (A for Author, T for title, H for holdings, C for call number, B for branch library location, etc.). Typists are instruct- ed to type until the margin locks. The be- ginning of each logical record is identified by the ''chair" character (r\) plus the typ- ist's initials at the beginning of the line. Thus the entire line is utilized, and the 260 Journal of Library Automation Vol. 6/4 December 1973 nSSYA=FAOYTPROCEEDINGS OF A SYMPOSIUM ON MAN MADE FORESTS AND TH EIR INDUSTRIAL IMPORTANCE, CANBERRA, 19b7YH1-3, 19b7//YCSD118.FS nSSYA=FAOYTREPORT ON A SURVEY OF THE AWASH RIVER BASINYH1-S, 1965// nSSYA=FAOYTETUDES PEDROHYDROLOGIQUES, TOGOYH1-3, 19b7//YCS599.C7Fb Fig.l. Berkeley Optical Scanner Input. machine is not required to read a large number of blank spaces at the beginning of the line (which, as Ms. Grosch points out, it has trouble doing since it cannot readily tell whether six blanks may, in fact, be really five or seven blanks) . We generally do not proofread the sheets after they are typed. We have found that when proofreading is necessary (usually during training), it is not difficult to proofread data typed in the format that we use. Data Element Identification-At Berke- ley, as at Minnesota, the typist identifies the data element (e.g., the author or the title) rather than relying on a computer algorithm of the kind used by the Library of Congress or the Institute of Library Re- search (automatic format recognition). This approach was selected because it was felt (a) that the typist could perform this task better than the computer could, and (b) that the routine nature of the typing job necessitated the insertion of more meaningful tasks for the typists. The data presented to the typists for interpretation can be in a wide variety of languages and may be transcribed on the source docu- ment according to any one of the conven- tions used by the library during the past several decades. Typing Throughput-The Berkeley conversion system includes the use of cer- tain "super abbreviations" that typists may use in place of commonly occurring words or phrases. All such abbreviations are two or three characters in length and are pre- ceded by an equal sign. For example, "= F AO" is translated into "Food and Ag- ricultural Organization of the United Na- tions," by the computer software. Al- though this substantially improves key- boarding throughput, its chief advantage is the insurance that the long phrase is en- tered into the :file correctly and consistent- ly. I personally find the requirement that the typist at Minnesota type the "format recognition line'; at the top of each sheet in .order to avoid the necessity of a "com- plete rerunning of the job" to be not only wasteful, but playing brinkmanship with systems design. Expanding the Character Set-Al- though the CDC 915 scanner is capable of reading only the OCR A font (an all upper case font), it is relatively simple to produce upper-and-lower case output from data input via the CDC 915. Two alternatives are: 1. Have the typist key a special char- acter that means "next character is to be capitalized" before each upper case character (the technique used by typists throughout the Western world, in the form of the shift key) . If, for the CDC scanner, the dollar sign were chosen to be that special character, the "$JOHN" would rep- resent "John" and "JOHN" would represent "john." This technique can be used to expand the keyboard to include diacritical marks. A Berke~ ley typist keys "ESP AN%EOL" to produce "espafiol," since the com- puter translates %E into a tilde over the preceding character. 2. Do all capitalization by logic con- tained within the software. A primi- tive computer algorithm might sim- ply say "capitalize the first word of every sentence plus the following proper nouns .... " The Berkeley li- brary currently uses such a tech- nique for the capitalization of words in serial entries. This has been done in order to print out the serial en- tries following standard rules of Technical Communications 261 CALL NUMBER University of California - Berkeley - General "Library Serials Key Word Index Page 372 HOLDINGS QA76.AlC545 Z699.AlH3 QA76.Al!4.R4 QA74.Ali65 QA16.All5549 QA16.A1!555 COMPUTERS QA76.A1A36 ENGI Advances in COMPUTERS , QA76.AlA36 HATH Advances in COMPUTERS , , , •••• QA76.AlC56 ENG! COMPUTERS and Automation •• , •• , RA409.5.AlC65 BIOL COMPUTERS and Biomedical Research • QC.145.2.C6 ENG! COMPUTERS and Fluids. , .•• , ••. , , , , •• 1,1960- 1, 1960- 2, 1953- 1, 1967- 1, 1913- 1968 TA64l.C65 LIBR COMPUTERS and Libraries; an Australian Directory. 1, 1911-ENG! COMPUTERS and Structures. , • • • • • . • • • , • QA76.5.A1C65 MAIN COMPUTERS and the Humanities ••••••••••• 1, Sep. 196* 5, 1910/11- 0n Order UNDE COMPUTERS and the Humanities ••••••••••• On Order HATH COMPUTERS, Control and Information Theory •••• LB1028.5.A1C* ~~g~ g~~~~~~~e igna~~~6tEi~di~e~~;in~der9r~d~~te • c~r~i~uia: P~o~e~dln9s: f~~ 8~~~ AlH65 * i:gi ~~~i~~~ng~n~~r:~~;t~~n~ir~~~i~:E~~Ste;~n!~~egg:Pb~~~~~' PrOc~edi~gS 1, 1970- 1, 1970- 1,1958- 1,1969- 1:18, Dec.l* 1, Mar. 196* 1, 1971- TI<6540.I55 ENG! IEEE Transactions on COMPUTERS , •••••• , • , • Z699.AlP76 LIBR Program; News of COMPUTERS in Libraries •••• , , , , ENGI Quarterly Bibliography COMPUTERS and Data Processing ••• 8~l~:~~2Q3 :i~ ~!~f~rH:r ag~i3f~~~~y. o: ?~P:rr~R~ ~n~ ~a~a. P:o:e~s~n? : QA4 7. T7 MATH Tracts for COMPUTERS. • , , • . • , , , , • • • , , • • , 1, 1971- 1, 1919; 2* 11-161; 1S- Fig. 2. Berkeley's Serial Key Word Index: Sample Page. style, rather than the traditional rules of librarianship, namely every significant word in the title is cap- italized. (Did the library practice arise because early typewriters had shift keys that were hard to use?) Our computer algorithm says essen- tially "capitalize all words in the en- try except the following insignificant ones .... " This technique has creat- ed an upper-lower case file without having typists use the shift key, or its equivalent, .at least a half million times. Figure 2, a page from Berke- ley's Serials Key Word Index, illus- trates the results of this system. The Real Problem-! do not mean to imply that everything is rosy in file con- version land. A file conversion is a messy, difficult and essentially unproductive task, no matter how well done, because it merely transforms existing data into an- other form and in so doing exposes, for all to see, the "many ancient errors" which we do not want to see. It also exposes the "ambiguities" that were perhaps better left ambiguous, not to mention the incon- sistencies that have cropped up as library practices varied. I would suggest that any file conversion that works from files that have been built up over some time period requires more in the way of resources for the "cleansing" than for the conversion. That is, in the case of the subject authority files at Min- nesota, I would guess that far more than $5,296.21 (the total amount spent on typ- ists, keyboards, computers, supplies, etc.) was spent resolving ambiguities (before the drawer was handed to the typist) and "cleansing" the data in the one year be- tween the time when the data had been converted and the time that they were put to use. This has been our experience at Berkeley. Stephen Silberstein University of California, Berkeley 262 Journal of Library Automation Vol. 6/4 December 1973 REPORTS-LIBRARY PROJECTS AND ACTIVITIES Bucknell University Plans Entire Bibliographic File to Go On-Line Bucknell University's already strong computer-usage program is expected to be strengthened in 1973/74 to permit stu~ dents and faculty to conduct fast, accurate searches of the university library from any of thirty-five campus terminals. A $28,000 grant to the Bucknell University library from the Council on Library Resources is supporting this program. Seventy-five percent of Bucknell's stu- dents already use the campus computer in course work. And Bucknell's on-line li- brary data base includes records of ap-- proximately 25,000 of the library's 200,000 books. The council grant will en- able additional computer storage to be rented to permit the entire bibliographic .file at Bucknell to go on~line. The com- plete flle is already in machine-readable form. While Bucknell's current system enables a search of the on-line files by author- title, title alone, and Library of Congress (LC) number, its enlarged plan calls for subject search capability as well. Using LC classification numbers, a user will be able to ask the computer to locate and dis- play the authors and titles associated with the subject of interest, examine the near neighbors of his original hit in the file, or he may pick an author's name from the response and enter the system again on the author's name to see what else the au- thor may have written. Stanford University Data File Directory The Stanford University Data File Di~ rectory, compiled by Douglas Ferguson, is available as an example of a library- produced access publication for computer~ ized data files on a university campus. The directory lists and describes colle~ tions of social, economic, political, and scientific research data on punched cards, computer tape, and disk, located on the Stanford campus. Each file description di- rects the user to documentation and pub- lished research in the university library collection or elsewhere. Access to each data file is controlled by the owner and is listed in each file description. The di~ rectory is available, for prepayment of $4, from the Financial Office, Stanford University Libraries, Stanford, CA 94304. STANDARDS Editor Note: The recent flurry of activi- ty concerning standards which affect l~ brary automation, dilta bases, etc., is pointed up in the several actions reported in the last issues of TC. Perhaps the fu- tility of keeping up with standards and the need for a clearinghouse type of op- eration is best recognized by noting a sample of some recently adopted stan.. dilrds which now have or will potentially have ramifications in library automation. The following list does not represent a complete accounting of all pertinent stan- dards due to lack of a comprehensive source. Selected ANSI Standards Many ANSI Standards published in the ANSI categories of "Information Process- ing Systems" and "Information Systems" may be of interest to ISAD members. Selected items are listed below. The new American National Standards Insti- tute (ANSI) catalog is available free of charge from the Institute's Sales Depart- ment at 1430 Broadway, New York, NY 10018. The catalog lists " ISO Standards" and "ISO Recommendations" as well. X3.14 RECORDED MAGNETIC TAPE FOR INFORMATION INTERCHANGE (200 CPI, NRZI) (REVISION OF ANSI X3.14-I969)--Provides the standard technique for recording American Nation~ al Standard Code for Information Inter- change (ASCII), X3.4-l968, on magnetic tape at 200 characters per inch ( CPI) using nonretum-to-zero-change on ones (NRZI) recording techniques. Approval date: December 12, 1972. X3.38 COMPUTER CODE FOR STATES-X3.38-I912 provides two-digit numeric codes and two-character alpha- ------~------------------- betic abbreviations for both the states and the District of Columbia. The numeric codes will allow the states and the Dis- trict of Columbia to be sorted into alpha- betic sequence. ANSI X3.38-1972 may be obtained from the American National Standards Institute at $1.25 per copy. It was devel- oped under the secretariat of the Busi- ness Equipment Manufacturers Associa- tion. X3.31 STRUCTURE FOR THE IDEN- TIFICATION OF THE COUNTIES OF THE UNITED STATES FOR INFOR- MATION INTERCHANGE (NEW STANDARD )-Identifies a three-digit numeric code structure for the counties of the states of the United States, includ- ing the District of Columbia. Supersedes the listing which appeared in the March 26, 1971 issue of Standards Action. Ap- proval date: March 14, 1973. X3.39 RECORDED MAGNETIC TAPE FOR INFORMATION INTERCHANGE (1600 CPI, PHASE ENCODED) (NEW STANDARD)-Presents the standard technique for recording the coded charac- ter set provided in American National Standard Code for Information Inter- change, X3.4-1968 (ASCII) on magnetic tape at 1600 characters per inch (CPI) using phase recording techniques. Ap- proval date: March 7, 1973. X3.40 UNRECORDED MAGNETIC TAPE FOR INFORMATION INTER- CHANGE (9-TRACK 200 AND 800 CPI, NRZI, AND 1600 CPI, PE) (NEW STANDARD)-Presents the min- imum requirements for the physical and magnetic interchangeability requirements of ~-inch wide magnetic tape and reels between information processing systems, communication systems, and associated equipment using American National Stan- dard Code for Information Interchange, X3.4-1968 (ASCII). Approval date: March 5, 1973. BSR X3.41 CODE EXTENSION TECH- NIQUES FOR USE WITH THE 7-BIT CODED CHARACTER SET FOR ASCII (ANSI X3.4-1968) (NEW ,PROPOSED STANDARD)-Provides means for aug- menting the standard repertory of 128 Technical Communications 263 characters of American National Standard Code for Information Interchange, X3.4- 1968 (ASCII), with additional graphics or control functions, by extending the 7-bit code while remaining in a 7-bit en- vironment, or increasing to an 8-bit en- vironment in which ASCII is a subset. Order from: Business Equipment Man- ufacturers Association; 1828 L St., NW; Washington, DC 20036. Single copy price: Free. BSR X3.47 IDENTIFICATION OF NAMED POPULATED PLACES AND RELATED ENTITIES OF THE STATES OF THE UNITED STATES, STRUC- TURE FOR THE (NEW PROPOSED STANDARD)-Provides the structure for an unambiguous, five digit code for named populated cities, towns, villages, and similar communities and for several categories of named entities similar to these in one or more important respects. Order from: Business Equipment Man- ufacturers Association; 1828 L St., NW, Washington, DC 20036. Single copy price: Free. BSR X11 .6 OPERATIONAL DATA PROCESSING APPLICATIONS CON- TAINING CONSTITUTIONALLY PRO- TECTED DATA, DOCUMENTATION REQUIREMENTS FOR (NEW PRO- POSED STANDARD)-Provides all those involved with operating electronic data processing applications, involving consti- tutionally protected data, with a list of minimum documentary requirements which apply to such applications. Order from: Society of C ertified Data Processors, 38 Main St., Hudson, MA 01749. Single copy price: $2.00. BSR Xll.l CATEGORIES OF ERROR- CREATING CHARACTERISTICS OF VARIOUS DATA STORAGE SYSTEMS USED WITH ELECTRONIC DATA PROCESSING APPLICATIONS (NEW PROPOSED ST ANDARD)- Provides the consumers of electronic data processing applications and the suppliers and imple- mentors of such applications with a tech- nique for defining the error-generating capabilities that exist in the data st orage system used to hold the consumer data. It is one of a series of data storage stan- 264 Journal of Library Automation Vol. 6/4 December 1973 dards being prepared by the Society of Certified Data Processors Technical Stan- dards Committee, to provide a method whereby the application implementor ·and the application consumer may communi- cate easily, allowing the application con- sumer to take the responsibility for the ac- curacy of the maintenance of the data base by electronic data processing sys- tems. . Order from: Society of Certilled Data Processors, ATTN: Chairman, Technical Standards Committee, 38 Main St., Hud- son. MA 01749. Single copy price: $2.00. BSR X11.2 DATA ITEMS STORED IN GENERAL DATA BASES, CLASSIFI- CATION . OF (NEW PROPOSED STANDARD)-Provides the suppliers of data to a general data base \vith a means of communication with the operation of the base regarding the characteristics of the data items being supplied. Order from: Society of Certified Data P~o~ssors, ATTN: Chainnan, Technical Standards Committee, 38 Main St., Hud- son,.MA01749. Single copy price: $2.00. BSR Xl1.3 DATA BASE PROCESS- ING ACTIVITIES BASED ON DATA ITEMS USED, CATEGORIES OF (NEW PROPOSED STANDARD)-Provides the application designers of data base appli- cations and the operators of several data bases with a means of describing the characteristics of the data items stored in the data base. Order from: Society of Certified Data Processors, ATTN: Chairman, Technical Standards Committee, 38 Main St., Hud- son, MA 01749. Single copy price: $2.00. BSR X2.3.4-1959 CHARTING PAPER- WORK PROCEDURES , METHO D OF -This standard was one of the original input docume nts considered in the devel- opm.ent of American National Standard Fi.owchart Symbols and Their Usage in Information Processing, X3.5-1970 (orig- inally ANSI X3.5-1966). However, ANSI X2.3.4-1959 was not considered sufficient- ly useful to serve the needs of the commu- nity which now uses ANSI X3.5, nor at that time did X3 have responsibility for ANSI X2.3.4 or feel that it should initiate action to modify the older standard. The subject standard was subsequently as~ signed to American National Standards Committee X3 for review and revision, re- affirmation or withdrawal. Current review finds no interest in this standard, either in the form of users of the standard or of an organization desiring to assume its maintenance. Order from: American National Stan- dards Institute, Dept. BSR, 1430 Broad- way, New York, NY 10018. Single copy price: $1.00. SC/ 20 STANDARD SERIAL COD- ING-The American National Standard Identi.Scation Number· for Serial Publica- tions, .Z39.9-1971 is available from ANSI at $2.25 per copy. In June 1970, ISO/ TC 46/WG 1· accepted the system as out- lined in Z39.9-1971 -as the basis for the international standard numbering system. A final ISSN standard was presented to the Plenary Session on TG 46 in October 1972 at the Hague. The International Center (IC) of the International Serials Data System (ISDS) is responsible for the administration of the ISSN as a central authority. The IC-ISDS was established with headquarters in the Bibliotheque Nationale with financial support being shared by the French Government and UNESCO. The National Serials Data Program (NSDP) has been selected to serve as the United States National Center and as such is the sole agency responsible for the con- trol and assignment of ISSN in the U.S. (Note-The ANSI !STAB (Informa- tion Systems Technical Advisory Board) rejected the proposed ANSI Z219.1-1971 , Use of CODEN For Periodical Title Ab- breviations. This proposal had been sub- mitted to ANSI by the American Society for Testing and Materials in 1971 for ap- proval as an American National Standard; Z39 members were asked to comment on it during the public review in July and August 1971. After considerable discus- sion the 1ST AB came to the conclusion that the proposed standard was in conflict with 239.9-1971, the. ANSI Identification Number for Serial Publications.) SC/ 2 MACHINE INPUT RECORDS -The members of SC/ 2 have agreed that this standard cannot be written at this time. The purpose of the proposed standard was for general information in- terchange at the interface between data processing terminal equipment {such as data processors, data media input/ output devices, office machines, etc.) and data communications equipment (such as data sets, modems, etc.) . The decision was based on the fact that the problem of de- signing a format is not being addressed here (that standard already exists, name- ly Z39.2-1971) but rather the problem of ne twork protocol. Therefore, the trans- mission of the bibliographic record itself, taken in this context, is only a small part of the total picture. Subcommittee 2 has concluded, how- ever, that in the light of future develop- ments in network protocol, bibliographic data should be transmitted in the Z39.2- 1971 interchange format standard. In or- der to further this recommendation, the present Z39.2-1971, the American Nation- al Standard for Bibliographic Information Interchange on Magnetic Tape, will be revised by SCI 2 to reflect a broader scope, i.e., information interchange in digital form, with appropriate sections in the document describing the existing stan- dards for different media (the first of these would be magnetic tape since this standard already exists). This should have the effect of using the standard format in future systems via telecommunications as well as via magnetic tape. The additional sections discussing various media will aid the user of the format regardless of the media involved. INPUT To the Editor: Say it isn't so. Tell me that, as editor of Technical Communications, you are not responsible for the item on page 65 of val. 6, no. 1. I refer to the squib headed "Tomorrow's Library: Spools of Tape." I am particularly offended to see this kind of outdated foolishne ss promoted after noting two pages earlier that the new di- rections for Technical Communications will involve pertinent information about technical developments. How could a Technical Communications 265 publication entitled College Management possibly contribute technically significant information about such a specialized and sophisticated area as library automation? In general, I think blue sky articles are inappropriate for TC. Carl M. Spaulding Council on Library Resources The new format and content of Techni- cal Communications is expected to evolve, and thf18 no step function change was an- ticipated. In the meantime, while operat- ing on an accelerated publication schedule I have attempted to find pertinent (if not completely appropriate) articles for TC. I would like to see more contributions of hardcore technical communications from the field, but until people accept the new design for TC and contribute to it, the se- lections wiU be scarce. Incidentally, I have received some com- ment to the contrary, that perhaps a " Bltte Sky?" category of news notes in TC would serve the useful purpose af providing an- other perspective, or putting "far out" items into context. Certainly, contributions of the type submitted by Stephen Silber- stein in this issue and Justine Roberts in the last issue of TC represent the direc- tions envisaged for TC's content. In most technical fields there's a place for the proposed TC type of forum, and r m confident library automation and technology have a similar need. I would appreciate more readers' comments, and more importantly, brief write-ups of the technical aspects of your accomplishments ami. findings which would be of interest to ISAD members.-DLB POTPOURRI UNISIST International Serials Data System The International Serials Data System { ISDS ) as establi~hed within the frame- work of the UNISIST program, is an in- ternational network of operational centers, jointly responsible for the creation and maintenance of computer-based data banks. 266 Journal of Library Automation Vol. 6/ 4 December 1973 The objectives of the ISDS system are: a. To develop and maintain an inter- national register of serial publica- tions containing all the necessary in- formation for the identification of the serials. b. To define and promote the use of a standard code (ISSN) for the unique identification of each serial. c. To facilitate retrieval of scientific and technical information in serials. d. To make this information currently available to all countries, organiza- tions, or individual users. e. To establish a network of communi- cations between libraries, secondary information services, publishers of serial literature, and international or- ganizations . f. To promote international standards for bibliographic description, com- munication formats, and information exchange in the area of serial pub- lications. The ISDS is designed as a two-tier sys- tem consisting of: an International Centre (IC) National and Regional Centres The ISDS-International Centre is estab- lished in Paris by agreement between Unesco and the French Government. It is temporarily located at the Bibliotheque Nationale. The ISDS-IC will establish an interna- tional file of serials from all countries. This file will be limited, initially, to sci- entific and technical publications, and will be gradually extended to include all dis- ciplines. Each serial will receive an Internation- al Standard Serial Number (ISSN), which has been developed by the International Organization for Standardization (ISO). Products which could be derived from the International Serials Data System are as follows: Titles Index; ISSN Index; ISDS Reg- ister of . Periodicals (Register); Clas- sified Titles Index ( CTI); New and Amended Titles Index (N & AT); Cumulated New Titles (CNT); Per- muted Index; Microform Reference File (MRF). A magnetic tape service will be pro- vided of the current master file, and of the new and amended titles. The responsibility for the establishment of National or Regional Centres belongs to Unesco member states, and associate members who wish to participate in the UNISIST program. Upon establishment each National Cen- tre will obtain a block of ISSNs from the International Centre and will gradually take over the responsibility for the regis- tration of serials published in its territory. A regular information exchange pro- gram will be established between the na- tional centers and the international cen- ter. The international register will thus be a regularly updated cumulation of the ini- tial file established by the IC and the Na- tional or Regional files. Serials published in countries with no National or Regional Centres will be regis- tered by the International Centre, which will endeavor to obtain the necessary in- formation. The relationship with users of ISDS is primarily through National or Regional Centres, but this general rule does not ex- clude direct contact with the Internation- al Centre. The building of a consistent internation- al file of serials implies close cooperation between all members of ISDS. The work in all countries will be based on a common set of rules concerning: bib- liographic description, communication for- mat, character sets, abbreviations, trans- literation, etc. Coordination between all members of the system is one of the main tasks of the International Centre. Close cooperation has also been estab- lished with various international organi- zations, the objectives of which are close- ly related to those of ISDS. In November 1972 the Director-Gener- al of Unesco informed member states of the creation of the International Centre and has invited them to cooperate in ISDS by establishing national or regional cen- ters. To assist in the creation of these na- tional or regional centers provisional guideUnes were made available. These guidelines are at present being finalized and will shortly be widely distributed in English, French, Spanish, and Russian. The response of member states was most encouraging and to date the follow- ing countries have set up or are in the process of setting up national or regional centers: Argentina, Australia, Austria, Canada, Colombia, Dahomey, France, Federal Republic of Germany, Guatemala, India, Italy, Malta, New Zealand, Nigeria, Philippines, Union of Soviet Socialist Re- publics, United Kingdom, and United States of America. For further information and ISSN as- signment contact IS OS-Intemational Cen- tre, Bibliotheque Nationale, 58 rue de Richelieu, Paris 2eme, France. ADL to CondtJCt Study of the Data Base Publishing Industry Arthur D. Little, Inc., the Cambridge, Massachusetts, consulting firm, is launch- ing a major study of the data base pub- lishing industry. The study, which will be available on a subscription basis, will cov- er present and future technology utiliza- tion, economics, markets, and business and competitive structure. More specifi- cally, the study will: • Characterize typical data base pub- lishing activities in terms of markets, products, sales strategies, methods of data base collection, distribution, etc.; • Identify the current and expected roles of private industry sectors, gov- ernment, and professional associa- tions; • Analyze existing and latent markets for data base publishing ventures and estimate market growth over the next five years; • Describe criteria for analyzing the economics of data base publishing services and pricing them; • Review hardware, software, and de- velopments likely to affect the indus- try in the next five years, including emergence of lower-cost switched data networks; • Describe the probable impacts of Technical Communications 267 public policy and regulatory devel- opments, including copyright legis- lation, patentability of software, and concern over protection of confiden- tiality of personal information, and • Characterize the reasons for past failures of certain data base publish- ing ventures and propose strategies for successful involvement. The study will be directed by Vincent Giuliano and Robert Kvaal. Dr. Giuliano has extensive experience working with major information dissemination systems, ranging from libraries to telecommunica- tions-based computer systems. He has led a variety of systems development, systems analysis, evaluation, and market research projects at ADL. Mr. Kvaal has focused his recent work on strategic planning issues facing com- puter services companies, and on assisting computer users in financial institutions, retail and distribution companies. This work has included operational and man- agement audits, planning and implemen- tation assistance, management informa- tion systems development, and the overall design of a nationwide teleprocessing sys- tem. According to Giuliano and Kvaal, data base publishing enterprises tend to evolve through well-defined stages of automation and business development: maintenance of manual data bases (reports, clippings, etc.) and the manual preparation of con- ventional printed products; partial com- puterization of the data base and some computer usage in preparation of conven- tional printed products; considerable au- tomation of the data base and output process; offering of information retrieval and specialized search services on an over- night or phone call basis; and offering di- rect access to the data base via remote computer terminals. "But," Giuliano and Kvaal note, "the growing tendency of data base enterprises to evolve along this scale is creating dislocations in many of them, while at the same time, offering new op- portunities for participants and suppliers. This uncertainty makes a study such as ADL's especially useful at this point in the industry's development." 268 Journal of Library Automation Vol. 6/ 4 December 1973 The results of ADL's study will be pre- sented to clients in published form and in group meetings held in appropriate loca- tions. The cost to each subscriber is $2,000. Additional information may be obtained from Philip A. Untersee (617- 864-5770). PERTINENT PUBLICATIONS New 1973 ACM Publication Catalog The new expanded thirty-four page Publication Catalog of the Association for Computing Machinery has been released. The catalog covers technical publications in over thirty major segments of the com- puting and automation field. Copies are available upon request by writing to: Pub- lication Services Department, Association for Computing Machinery, 1133 Avenue of the Americas, New York, NY 10036. Proceedings of 1973 National Computer Conference The Proceedings of the 1973 National Computer Conference & Exposition are now available from the American Federa- tion of Information Processing Societies, Inc. (AFIPS). The Conference Proceedings, Volume 42, contains more than 160 technical pa- pers and abstracts covering a wide range of topics in Computer Science & Technol- ogy and Methods & Applications featured at the recent '73 NCC, June 4-8 in New York. The price of the 920 page hard-cov- er volume is $40. A reduced rate of $20 is available for prepaid orders from mem- bers of the AFIPS' Constituent Societies stating their affiliation and membership number. Copies of the Proceedings may be ordered from AFIPS Press, 210 Sum- mit Ave., Montvale, NJ 07645. Computerized Serials Systems The LARC Association announces a new publication series entitled Computer- ized Serials Systems. Each volume in the series will consist of six issues published · at bimonthly intervals in both paperback and hardbound editions. Each issue will be authored and edited by a person di- rectly affiliated with the project reported, and each issue will be devoted to papers relating to an automated serials project undertaken by a specific library. The for- mat of the new series is designed to pro- mote understanding through clear narra- tive description and extensive illustrative materials. For details concerning the purchase of individual issues or a subscription to the complete volume contact LARC Press, 105-117 West Fourth Ave., Peoria, IL 61602. --