ACRL News Issue (B) of College & Research Libraries 2 3 6 /C&RL News W hat’s hot and w hat’s not By Denise A. Troll A look at the automation vendor exhibits at Midwinter I spent six hours visiting the booths of library automation vendors in the exhibit hall at the ALA Midwinter Meeting in Philadelphia. In most cases, a vendor representative was available to talk so I introduced myself as a reporter for C&RL News, showed my yellow press ribbon, and asked a leading hypothetical question: If I were a customer, why should I buy your prod­ uct instead of the competition’s? After the con­ ference, I reviewed my notes and the handouts and press kits that I gathered from the vendors. This article reports on the adventure that en­ sued. It is presented not in chronological or­ der, but categorically, the events described in terms of what’s hot and what’s not in the com­ mercial world of library automation. W h a t ’s hot Six trends emerged as I navigated the busy aisles o f the exhibit hall and reviewed the marketing literature: 1. Library m an agem en t software. M vendors are automating library staff functions and technical services. Like cars, the products range from the inexpensive compact model to the expensive four-door sedan. Vendors like Nichols Advanced Technologies, Inc., and the Library Corporation address the needs of the small library by automating one or more func­ tions, typically circulation and cataloging. Brodart TIPS software automates collection analysis and development functions. Other ven­ dors, targeting larger budgets, are building so­ phisticated systems that integrate staff and public (OPAC) access to library databases and services. Integrated system software is modular in de­ an sign and sales. The requisite modules of a fully integrated system appear to be circulation, cata­ loging, acquisitions, serials control and the OPAC. Many vendors— including Ameritech, CARL, DRA, Endeavor, Gaylord, Innovative In­ terfaces, Sirsi, and VTLS— have these modules now. Other vendors have one or more of the key modules in development (e.g., Mandarin and Tapestiy). The modules typically include transaction logging and reporting. Some ven­ dors have additional modules in production or under construction, including data conversion, authority control, and media booking. The trend is toward client-server architecture and relational database management. There also seems to be a trend to migrate to Unix although there are exceptions, and different strategies are being used to build systems— e.g., RDBMs versus tra­ ditional keyword/boolean engines like BRS. 2. T he W orld Wide Web (W W W ) and o th e r In tern et services. If anything stuck out as new in the automation exhibits at ALA, it was the number of vendors advertising URLs as access points to their real or sample data or to information about their products and ser­ vices. SilverPlatter, the Library of Congress, and yma ny o f the fully integrated system vendors noted above provided promotional handouts that described their use of the WWW. Faxon put its URL on the buttons it was giving away. Mosaic and Netscape were the prevalent cli­ ents. The most striking applications of this tech­ nology— Sirsi’s WebCat and OCLC’s WebZ— integrate WWW and Z39.50 services. WebCat dynamically creates “pages” from Sirsi’s pro­ prietary server or any compliant Z39.50 server database. In addition to WWW access, software ven­ dors are providing gateways to other Internet services, e.g., telnet, ftp, gopher, WAIS, and electronic mail. The presentation o f the gate­ Denise A. Troll is h e a d o f research a n d development, library autom ation a t Carnegie Mellon University Libraries; e-m ail: trolh@ andrew .cm u.edu mailto:trolh@andrew.cmu.edu A pril 1 9 9 5 / 2 3 7 way varies and with it the user’s access route to information. When the gateway is integrated with the OPAC, users simply follow links in retrieved information. When Internet services are packaged separately, users must select ei­ ther a service or a destination prior to search­ ing for inform ation. For exam p le, HSLC’s NetSelect software provides connection to the Internet, menu-driven access to Internet ser­ vices, and a feature that enables users to moni­ tor Internet discussion lists without subscribing to them. Sirsi’s VIZION software organizes In­ ternet services in a desktop portfolio o f over a thousand Internet destinations that can be searched, sorted, and filtered. For sites without local expertise, the Global Library Connection Program from C.L.A.S.S. will design the net­ work, configure the machines and popular In­ ternet service software, write the necessary scripts, and even customize the home page on the WWW. 3. E n h an ced OPAC features. In additi to integrating OPAC and Internet services, en­ hanced OPACs provide access to local and re­ mote databases online or on CD-ROM. Again, the architecture is client-server with software modules sold separately. For example, Sirsi and Ameritech sell integrated OPAC and library management system software separately from their system software for locally mounted ref­ erence databases and indexes. Gaylord provides gateways to IAC and RLG databases. VTLS sells a gateway module to CD-ROM databases. The need for interoperability with remote catalogs and databases foregrounds the use of standards. The popular retrieval standard is Z39.50. Many vendors in the exhibit hall dem­ onstrated Z39.50 clients interoperating with Z39.50 servers at different sites. DRA’s Z39.50 client (DRA Find) can search multiple databases at multiple sites simultaneously and retrieve local holdings information from the sites that support this service using Z39.50. Ameritech, CARL, and TRW recently announced efforts to provide Z39.50 interoperability with SilverPlatter databases. (IAC and UMI databases will soon be available from SilverPlatter.) Note that Z39.50 implementations do not necessarily talk to one another and libraries need to be aware of the “real” versus “theoretical” compatibility. Even when systems are compatible, databases on remote servers may not have all the indexes that are available in local databases, so retrieval may not meet the user’s expectations. An addi­ tional standard being applied for interoperability o among vendors is EDI, electronic data inter­ change. The newest enhancements to OPAC tech­ nology provide patron self-service features. Depending on whether the features are config­ ured to be mediated or not, patrons can either submit requests for or actually check out books, put them on hold or reserve, renew or recall them, and inquire about overdue books and fines. Many vendors use the 3M software for self check-out services. A new development is the provision o f these services over the phone. CARL advertises Phone Circ and Phone Notices; VTLS has Phone Works. Several integrated sys­ tems— e.g., Best-Seller and Sirsi— also provide e-mail services so that users can send ques­ tions, comments, or bug reports to library staff. Multilingual user interfaces are available but not widespread. For example, Nichols offers an OPAC in English and Spanish; Winnebago offers an OPAC in English, Spanish, and French. n 4 . More data, faster updates. Another trend in the exhibit hall was the proliferation of data and data formats available online, on CD-ROM, or on magnetic tape, and comments about timely updates and distribution procedures. The focus is on full-text news, business informa­ tion, government documents, scholarly and lit­ erary works, and hard-to-find publications. Target markets range from K– 12 through schol­ ars and professionals. A product called Broad­ cast News provides searchable full-text tran­ scripts from over 80 television news programs. Softline Information, Inc., is keying in the full text of ethnic materials that have limited distri­ bution and are therefore often inaccessible. The pinnacle o f current data delivery is OCLC’s PRISM PromptCat service that enables users to retrieve MARC records o f new materials prior to publication. 5. Clients and co lo r. Character-based cli­ ents are available, but the trend is conspicu­ ously toward graphical user interfaces. The most popular graphical client is for a PC running Microsoft Windows. The second most popular platform is the Macintosh. H.W. Wilson intro­ duced “the first Macintosh-based CD-ROM prod­ uct.” IAC offers InfoTrac databases for the Macintosh. Several integrated system vendors (e.g., DRA, Sirsi, and OCLC) will have Macintosh clients in 1995. Character-based clients on PCs and Macs typically have a bright blue screen with red and yellow highlights. The graphical user interfaces use the same primary colors or an entire palette o f bright colors. Motif (UNIX) 2 3 8 /C& RL News clients are rare and use more subtle colors or gray scale displays. 6. D ocum ent delivery and multimedia. More and more vendors are handling or plan­ ning to handle full-text documents in ASCII and image formats. Access to image documents is typically provided through links to database records or browsers that enable users to navi­ gate hierarchies of documents— e.g., selecting a journal title, then a volume, issue, and article. Some clients, like UMI’s ProQuest and RLG’s Ariel for Windows, can display and print the documents. Others enable users to request in­ terlibrary loan or fax or mail delivery (e.g., Ei’s Article Express International). Image and ASCII documents are being provided by publishers, information brokers, or the local site. OCLC’s FirstSearch enables users to specify the deliv­ ery method and document supplier. OCLC’s SiteSearch Image Extension module, RLG’s Ariel, and other products enable sites to build image collections (including electronic reserves) for desktop delivery. Nichols’ Athena software provides facilities for linking color photographs to patron records in their circulation systems. Audio and full-mo tion video are becoming more prevalent. GEAC and DRA demonstrated sounds linked to data­ base records. The newest applications o f sound technology are the telephone self-service mod­ ules (described above). W h a t ’ s n o t h ot In the course o f my discussions with the ven­ dors, I had the opportunity to ask follow-up questions. It was here that I learned what was not hot in library automation, an observation that was confirmed by my review o f the pro­ motional literature. 1. Aw areness o f th e com petition. The thing I learned that is not hot is competitive awareness: library automation vendors often don’t know what the competition is doing. It was alarming how many vendors admitted that they couldn’t tell me what features or function­ ality they offered that the competition did not. Only one vendor pulled up a chair, invited me to sit down, introduced herself as “David in a sea of Goliaths,” and talked to me at length about the philosophy of customer service as problem-solving that sets Best-Seller apart from the competition. Since most vendors couldn’t answer my question, they rattled off a grocery list o f buzz words: client-server, standards, In­ ternet, graphical user interface, parameteriza­ fi tion. I heard again and again that their product was “fully functional,” but no one I asked could specify a function that they had that their com­ petition did not. In one case, a vendor mistak­ enly cited a feature as new and unavailable from anyone else; the feature has been avail­ able from competitors for years. 2. Aw areness o f th e cu sto m er. The sec­ ond thing I learned was that many vendor rep­ resentatives know little if anything past the marketing hype. No one I asked could tell me what version o f Z39.50 they were running, though they assured me that it was the newest one. Sometimes the representatives I talked with openly admitted that they were unfamiliar with the software and could not adequately demon­ strate or describe its features or underlying ar­ chitecture. Other times they tried to muddle through, not necessarily to the advantage of the product. I suspected that many of the ven­ dors underestimate the knowledge and sophis­ tication o f their customers. During my two breaks from chatting with vendors, I asked people who had been in the exhibit hall their impression of the vendors’ discourse, the way they were pitching their products and answer­ ing questions. The people I asked had the same impression I did and a feeling of disappoint­ ment. They’re looking for professional prod­ ucts, product literature, and vendor represen­ tatives who know and can demonstrate or describe how the products work from the per­ spective of the user, the system administrator, and the software architect. Budgets are tight and libraries must shop wisely. Many of us are leery of canned demos, weary of media hype, and would prefer an intelligent conversation. 3. Awareness o f hum an factors research . Vendors also apparently know little if anything rsta bout human factors and interface design. A simple review o f the published research would have prevented some of the atrocious inter­ faces demonstrated in the exhibit hall and pic­ tured in the promotional materials. The ex ­ amples I’ll cite here pertain to the use of color, but I could just as easily have picked examples that pertain to readability research. Many ven­ dors dem onstrated PC windows interfaces where the window frame (menu bar, scroll bars, and rows o f icons) danced in various vibrant colors. One vendor demonstrated an interface in which each field in a database record ap­ peared in a different color. Another vendor demonstrated lime green database records with red letters. I asked this vendor if he knew what April 1 9 9 5 /2 3 9 these records would look like to the color blind and w hat p ercentage o f the population was color blind with red and green. He responded that the colors w ere custom izable, and pro­ ceed ed to dem onstrate unsuccessfully how to change the color o f the many com ponents. During the attempted dem o, my mind w an­ d ered to Edward T u fte’s w ork on graphical design. Tufte describes the b est user interface as conveying the maximum am ount o f infor­ mation in the smallest, sim plest (least busy or distracting) sp ace. A good design will fo re­ ground the information contained within the w indow , not the w indow fram e; the two will not com pete for the user’s attention. Tufte cites the cartographer’s w ork as the best exam ple o f color interface design. T he implication is that a user friendly graphical user interface will look m ore like a good surface map than a rainbow. T he u se o f bright primary and second ary col­ ors in dancing rows o f m enus, icons, and text indicates that vendors are not applying relevant human factors research. SIRSI is an exam ple o f a vend or undertaking testing with users based on HCI. This is the b est approach to client de­ sign and developm ent. 4. Client development. I also learned t the client business itself is not hot. Tw o com ­ ha ments w ere quite telling. O ne vendor told me that a few years from now the client will be w hatever Mosaic evolves into. A nother vendor told m e that the client will b e w hatever Bill G ates wants it to be. What is striking about these com m ents is that the vendors s ee them ­ selves as being out o f the client business in the near future— w hich may explain why so many o f them are exploiting Mosaic instead o f c o n ­ ducting o r applying hum an factors research. 5. Technology for access not ownersh T he final observation I m ade that falls into the “not hot” category is that m any o f the CD-ROM vendors are reluctant to change media. I cat­ egorize this as “not hot” b ecau se it indicates that vendors have not yet adopted the philoso­ phy that has b e e n the library mantra for years: access not ownership. Selling stand-alone sys­ tem s w hose perform ances severely limit simul­ taneous use under a bann er that proclaim s to take the library into the 21st century som ehow strikes a dissonant chord in the con text o f the libraries’ harm onious vision o f future interop­ erability and possibility. As the cost o f mag­ netic disk storage d ecreases and the num ber o f successfully interoperable system s increases, t o n e can only h o p e that vendors will follow suit. ip.