lib-MOCS-KMC364-20131012114231 on their regular CLSI equipment in the main library and station branch. On days when housekeeping chores are scheduled, the console operator's job includes turning on the Apples so we can begin serving the public when the doors open at 9:00 a.m. Unless downtime persists for more than a day, no other routines are done except checkouts. Under some circumstances, cer- tain materials might be checked in on the Apple, but it is not desirable to do this for newer materials on which holds may have been placed. When the LIBS 100 is online again, the checkout station is switched back to normal mode and the Apple takes over the informa- tion desk's port for dumping, rendering that terminal inoperative. Dumping continues around the clock until all transactions have been processed from both Apples. Normal activities proceed at all other terminals. Diskettes are dumped in chronological or- der. As the dumping process operates, a file of transactions eliciting error or exception messages from the LIBS 100 is created on the Apple diskette. This file is available for attention at a later time for manual entry into the database. The chief asset of the dumping process is the accuracy achieved by automatic input- ting. When we used paper and pencil, not only was the original writing time consum- ing, but manual data entry was difficult be- cause of illegible handwriting, inaccurate transcription of the numbers, inaccurate in- putting into the database, and lack of avail- able personnel for the job. The CTI system resolves all of these difficulties, but a price is paid in the loss of the dumping terminal's services. The public may be less disturbed if a terminal in a nonpublic area is used. But to the department involved, access to the database is a central part of their work and its loss severely limits their output. In fact, dependence on the automated circulation system by all departments in the library has been swift and universal even though we originally assumed the terminals outside the circulation department would be used spar- ingly. Plans are being made to store personnel records in machine-readable form on disk- ettes. Other developments are being put on a back burner until we have less frequent Communications 299 need for the Apples as backups. However, LEVELS, Great Neck Library's Youth De- partment, has several Apples of its own on which budding "computerniks" practice their art. For them there are few limits to possible applications-perhaps only the outermost boundaries of imagination. REFERENCE 1. Joseph Covino and Sheila Intner, "An Infor- mal Survey of the CTI Computer Backup Sys- tem,'' journal of Library Automation 14:108-10 Oune 1981). Computer-to-Computer Communication in the Acquisition Process Sandra K. PAUL: SKP Associates, New York City. In the 1970s, we entered the period of computer-to-computer communication; we now appear to have reached the second stage of development. Today more than seventy publishers are equipped to receive computer tape orders and input them di- rectly to their order fulfillment systems; twenty-six publishers can produce com- puter invoices and credits for their cus- tomers; six are capable of sending monthly updating information about titles, prices, publication dates, and books declared out of print. All of this, however, is based on a system through which computer tapes are sent from buyer to seller and back via the United States mail. The next step- computer-to-terminal or computer-to- computer communication-is just around the corner. HISTORICAL PERSPECTIVE How did this happen? It started in Sep- tember 1974 when DeWitt C. ("Bud") Baker, newly appointed president of the Baker & Taylor Company, envisioned the savings his company could find if their cus- tomers provided the International Stan- dard Book Number (ISBN) on their orders. He also believed that the volume of paper created by the computer was expensive and time-consuming for publishers to handle. 300 Journal of Library Automation Vol. 14/4 December 1981 Always a visionary, he believed that com- puters communicating directly with each other would not only save time and money, but would prevent human errors intro- duced by research clerks or keypunchers. He invited publishers, booksellers, librar- ians, wholesalers, representatives of school systems, and others to a full-day meeting at theW. R. Grace building. This diverse group of individuals dis- cussed the ISBN-what it was, what it might do. By the end of the day, the group defined two areas in which efforts might bear fruit. One was educational- publishers needed to be told the importance of printing valid ISBNs on their books, in their promotional materials, in their adver- tising, and on any other source of ordering information, and on the invoices and pack- ing lists they send to their customers. Wholesalers, librarians, and booksellers needed to be shown the efficiency their use of ISBNs on orders introduces to the fulfill- ment process at publishing houses and wholesaler offices. These functions were as- signed to an ISBN Publicity Committee, chaired by Franklyn ("Lee") Rodgers of Scribner Book Company. The second func- tion was the design of computer-to- computer formats for orders and invoices that would be keyed to the use of ISBN as title identifier and would be industry-wide in scope. This function was undertaken by an ISBN Data Transmission Committee, chaired by David Wolverton, then of Bro- dart. The ISBN Publicity Committee pro- duced a booklet and posters, distributed them at all major conventions, made press releases available, and prepared articles for inclusion in the newsletter of all of the ma- jor industry associations. The committee surveyed the use of ISBNs by publishers and published a list of in-house contacts for ISBNs. The committee's program was a suc- cess! FORMAT DEVELOPMENT The ISBN Data Transmission Committee had a more difficult task. The first question they faced was one of basic approach. Im- mediately they decided to proceed with a format for orders rather than invoices. Next, they reviewed the level to which the format would be directed. Believing it was more appropriate to "crawl before they walked," they decided to develop a format that could be generated on computer tape, which would be mailed from buyer to seller through the United States mail. Once there had been experience with that format, work would begin on direct computer-to- computer communication formats and pro- tocols. The final decision related to form. The majority of people volunteering their time to work on this committee came from the major book publishing houses. Ad- ditional members included two major bookstore chains, B. Dalton and Walden- books; the New York Public Library (NYPL) and three major wholesalers, Bro- dart, Baker & Taylor, and the Ingram Book Company. Representatives from NYPL, R. R. Bowker Company, and the library wholesaling organizations were familiar with American National Standard Z39.2- Bibliographic Information Interchange on Magnetic Tape. They felt that this stan- dard, which is the basis for the MARC tapes, should also become the basis for an order to be sent on magnetic tape. The ma- jority of the committee, however, was not only unfamiliar with Z39.2, but with the concept of programming for variable length records and/or fields. These data- processing managers argued that the for- mat basically would consist of sending a quantity and the ISBN for each title or- dered, not the sending of bibliographic re- cords as such. After review of a strong and well thought out letter from Michael Malin- conico supporting the use of Z39.2, the ma- jority held to their decision and the subcom- mittee, chaired by Tom Brady, then of Baker & Taylor, was assigned responsibility for developing the first computer-to- computer order format. It is a fixed length field and record format. Debate continued throughout its devel- opment. Each publisher hoped to have to do minimal programming in order to inter- face the new format and the input require- ments of his or her specific order fulfillment system. Provision was made for minimal bibliographic information if an ISBN was unknown. Polling of the members resulted in decisions to limit author and title to thirty characters, for instance. Shortly before the format was approved by the committee, Dick Fontaine, then sales manager of B. Dalton (now president) and Dick Lieberman, sales manager of Random House decided that they would begin send- ing tapes in the mail as of January 1975. Once the format had been approved, other publishers joined the group. Orders were sent from Baker & Taylor and Brodart to Random House, John Wiley, Prentice Hall, and Doubleday. B. Dalton continued to send Random House tape orders in a slightly different version of the format. By the end of 1976, it appeared that the task of the ISBN Publicity Committee had become publicizing the order format, rather than the ISBN as such. The two com- mittees decided to merge in March 1977, selecting the name Book Industry Systems Advisory Committee (BISAC), as much be- cause it was a pronounceable acronym as for any other reason. Development of an invoice format, originally considered of im- mediate importance by Waldenbooks, had been shelved after that company lost inter- est and the individual chairing the subcom- mittee working on the format left the field. The next step became reports of experi- ence with the order format. Fields left open for individual use came under review for standardized coding; procedures were de- veloped for the marking of information on the outside of the tapes and paperwork to accompany it; some publishers refused to accept tapes without ISBNs, while others pondered the procedure to separate titles without ISBN from those with and then merge the two for discounting purposes. With all of its inadequacies, the format was working. Publishers reported saving up to one week of time in processing tape orders. Random House analyzed returns for wrong title or wrong edition one year after they began receiving tape orders from Dalton, Baker & Taylor, and Brodart. They found an extraordinary 47 percent decrease in that type of misshipment to those three cus- tomers. With a few years experience under their belt, BISAC decided to revise the format to accommodate the inadequacies and prob- lems members had found with it. In addi- tion, the R. R. Bowker Company and OCLC, Inc., had just announced their in- Communications 301 tentions of developing acquisition systems which would replace them in the role of "order forwarders." They would prepare orders for other organizations and transmit those orders, on tape or directly online, to the vendor of the organization's choice. This forced BISAC to include a field for an "order placer" in addition to the traditional "bill-to" and "ship-to" customer name and address. The revision was approved in Feb- ruary 1977 and called "Format 112." (It is this version of the format which has been programmed by publishers and wholesalers noted in the introductory paragraph). BISAC members began pressuring B. Dalton to convert from their original "pre- format 1" format to Format 112. In analyz- ing the cost of doing so, Dalton also consid- ered the potential saving they would have if invoices from publishers were received on tape. Although they have never made those figures public, the potential was so great that Jim Nermyr, then their vice president of data processing, agreed to chair a sub- committee to develop an invoice format, paralleling as closely as possible, the order format. Once that was approved Jim made "selling trips" to New York and elsewhere, convincing over forty publishers to pro- gram for the invoice format in return for receiving orders on tape in Format 112 from Dalton. Finally, BISAC members began express- ing concern about misinformation on or- ders and invoices. Typically, when two or- ganizations agreed to communicate using the standardized formats, they would ex- change tapes of titles, descriptions, and the appropriate ISBN for each. They would produce error reports and the purchasers would bring their computer records in line with the publisher's. However, once price changes occurred, books were made out of print, or publication dates changed for not- yet-published titles, orders carried erro- neous information. To resolve this, a sub- committee, chaired by Andrew Uszak of R. R. Bowker Company, set about developing what has come to be known as the "Title Update Format." This format allows a pub- lisher to send a monthly tape of all titles on file indicating those fields that have changed since the prior month, or simply sending the ISBN and changed field infor- 302 Journal of Library Automation Vol. 14/4 December 1981 mation. Six publishers are now sending in- formation in this format to the Ingram Book Company on a monthly basis; others will be doing so shortly. (The format is also the basis for information college textbook publishers send to update the monthly AAP Microfiche Service.) MOST RECENT CHANGES At its May 1981 meeting, BISAC ap- proved minor modifications to its order and invoice formats. These changes included in- creasing the zip code field to nine digits and specifying a seven-digit field for the Stan- dard Address Number. The invoice format was modified to accommodate its use for sending credits as well as invoices. These new formats were released in August 1981 and are "titled" Order Format #3 and In- voice Format #2. We guess that it will take at least a year before the bulk of those orga- nizations now sending tapes in Order For- mat #2 and Invoice Format# 1 program for the revisions. During 1980, BISAC members began ex- pressing concern that the "crawl before you walk" philosophy had stopped in the crawl stage. Baker & Taylor and Brodart, in par- ticular, expressed concern that the genera- tion of tapes was expensive and using the mails introduced such delays that orders were filled more promptly when phoned into publishers than when sent in the BI- SAC format. In addition, OCLC, which had programmed the order format into their new acquisition system, agreed that sending this information on tape would be far less effective than transmitting it online to the major vendors. In 1980, BISAC es- tablished a subcommittee, chaired by Jim Long of OCLC, Inc., to develop an alterna- tive version of the order format. This ver- sion, with variable length fields and re- cords, is intended for use in a communication mode between the main frames of two computers. We expect there will be deep consideration and long debate on this proposed version at BISAC meetings in the next few months, with passage ex- pected in 1982. Finally, BISAC brought its formats to the attention of American National Standards Committee Z39 (ANSC Z39) when format #2 was completed. At its May 1981 meet- ing, the committee decided to ask Z39 to officially begin work on formalizing both the Order Format #3 and the new variable- length order-format alternative as Ameri- can National Standards. The Z39 program committee and executive council agreed; Ernest Muro of Baker & Taylor is chairing the Z39 subcommittee charged with this task. As BISAC activities became more widely known, this ad hoc committee strained the resources of its volunteer officers in answer- ing requests for information, for copies of the formats, and in preparing and dissemi- nating the minutes of their meetings to an ever-increasing number of interested indi- viduals and organizations. In 1980, BISAC approached the Book Industry Study Group, Inc. (BISG) with the suggestion that they become a permanent committee of that research organization, whose mem- bership also included publishers, librarians, booksellers, and wholesalers, as well as book manufacturers. The BISG agreed and today supports BISAC activities through its offices. At the end of this communication is the address which can be used to request any of the formats from the BISG office. THE FUTURE Automation is here and here to stay. Indi- viduallibraries that once considered it im- possible to imagine being able to afford a computer now have several-and soon will have more-acquisition systems available to them through independent vendors and through the national bibliographic utilities. Wholesalers are gaining computer sophisti- cation, as are publishing houses. As an in- dustry, we are lucky that those volunteer data-processing types who formed BISAC and kept it alive were each willing to make the compromises necessary to provide us with a standardized industry-wide format. Other industries have not been so lucky, with major vendors using their dollar- volume clout to demand that their cus- tomers accept orders in their own , unique formats. However, the need for standard- ization is known. In 1979 the American Na- tional Standards Institute approved the cre- ation of a new committee-ANSC X12-Business Data Interchange. This committee is charged with developing ana- tiona! standard format for transmission of orders, invoices, and other transactions re- lated to the sale of merchandise, and the payment for that sale through electronic funds transfer. BISAC and ANSC Z39 have been carefully reviewing the progress this new committee is making. It appears that the formats that result from their efforts will be variable length fields and suffi- ciently general in nature to fit the needs of librarians, booksellers, wholesalers, and publishers, along with those involved in the sale and purchase of all other types of com- modities. The traditions and laws of this country preclude any organization from "forcing" a library to make use of these standardized formats. However, the cost savings, the guarantee of accuracy of the record re- ceived, and the speed with which the order reaches the fulfillment center suggest that these formats will increase in use in the fu- ture. We also anticipate that more and more use of the formats will be in an online trans- mission mode, rather than in the form of computer tapes in the mail. As the volume Communications 303 of transmissions grows, we expect that some day messages from purchasers will be forced into queues to reach the more popu- lar suppliers. To the extent that the major wholesalers provide terminals to their cus- tomers and/or facilities to accommodate a large number of transmissions, their queues may be minimal. However, it will be inter- esting to discover how individual publishers will cope with this situation. Readers who are interested in receiving copies of Order Format #3, Invoice Format #2 or Title Update Format# 1, should write to: Book Industry Study Group, Inc., 160 Fifth Ave., New York, NY 10010. There is no charge for these formats. If all three are requested and first class mail is requested, postal costs are billed to the recipient. Those interested in active participation on BISAC should send a letter to the organization stat- ing that request. Finally, those interested in receiving copies of the minutes of the bi- monthly meetings held between September and May should send a request, accompa- nied by a check for thirty-five dollars to the BISG address. Hungry for answers in the life sciences? No matter how unusual your search question , the new 1981 810SIS Search Guide can help you find the right answers. In response to a request for references on " people eating insects as food," an information scientist searched 810SIS Previews online: " ... I wanted a file with a broad subject coverage and wide range of search options. The first step I took in my search strategy was to use the free-text faci lity for the word 'edible' . Then I used the keywords insect , insecta, and insects, which I identified from the 810SIS Search Guide. The Guide also led me to several biosystematic codes .. . " from a prize-winning 8/0S/S Search Tourna - ment essay. For your online searches of biological and medical research, you need the right words ... the right concepts ... the right codes ... You'll find them all in the: 1981 BIOSIS Search Guide The price? Just $75.00. For further information, or to place your order, contact: BIOSIS Customer Services, 2100 Arch Street , Philadelphia, PA 19103. 800-523-4806 or 215-568-4016.