The Books In Print (BIP) case study illustrates how one company, R. R. Bowker, has taken an existing print product and created a separate valueadded electronic product.

When you're looking for a book one indispensable item is Books in Print. Produced by R. R. Bowker, it is present in every bookstore and library. Finding information in it, however, can be a problem if you don't know enough relevant items like the complete name of the author, book, or publisher. Obviously an on-line version is immensely useful. Reed Reference Electronic Publishing publishes a CDROM called "Books In Print PLUS." Thousands of independent book stores, small and large chains, including Barnes & Nobel and Borders use this CD-ROM, to locate and order books. Most of the information for this case study was provided by Skip Slawson of Reed Technology and Information Services Inc. (RTIS), and Marty Brooks of Reed Reference Publishing.

The BIP database contains 1.6 million entries of active titles, with another 700,000 inactive titles. There is also another 70,000 entries of publishers and a linked database of subject headings.

The database itself exists in a proprietary format called Bowker Power. The information is stored in a standard relational database. Information gets entered into the database via online keying, off-line keying and EDI transmissions from the publishers.

The paper book is created by processing the database via a proprietary composition program. After printing the pages are proofed.

The paper and CD-ROM versions of Books In Print are released monthly. Following is the description of the CD-ROM production process:

The CD-ROM database is processed via a proprietary premastering and indexing program at RTIS-Online Computer Systems. Data is output at Bowker in "Bowker Power" format and sent to RTIS via overnight delivery. Records are built for the bibliographic data, the publisher data, and the full-text review data. Subject files are incorporated. Nineteen indexes are created. In addition, inventory data from Ingram is added. The result is mounted on a VAX computer and is checked by the staff. Each index is checked and statistics are compiled for the number of entries in each index and the total number of hits in each index and compared to the previous month. Then, Bowker staff compiles the final root directory software and sends it to RTIS via the Internet. They create a WORM, ship it back to us at the same time they send a final tape to our disc pressing facility. We approve the WORM and release the tape for manufacture.

Quality Control (QC) is an important part of any data processing operation, such as the production of this CD-ROM. Every stage has the potential for error. QC procedures are numerous and occur at all stages of data input and CD-ROM creation. As one example, when a new record is entered the ISBN (a unique book identification number) is checked against another database. If the ISBN does exists it is checked against the author/title information to see whether it is really the same record as an existing one or a new intellectual work with an incorrect ISBN number.

In addition there is a large staff of database editors who correct and update information. Reed receives tapes and EDI transmissions from publishers, usually on a monthly basis. Weekly data feeds also occur occasionally.

An experimental Windows version of the Books In Print CD-ROM product integrates the BIP software with the rest of the PC.

There are several variations of the product. Books In Print PLUS costs $1095 per year and is updated monthly. Books In Print with Book Reviews PLUS costs $1595 per year. The Book Review version of the CD-ROM contains brief reviews of many books. INGRAM-Books In Print PLUS has weekly updates. INGRAMs is a major supplier of books to

bookstores. The CD-ROM contains information about inventory and warehouse locations. Ordering software is integrated with the CD-ROM in both versions.

Taking advantage of the value of their database Reed Publishing is starting to move their information to the Web. (Their Web site is at http://www.reedref.com.) Currently they have available for free the Books Out of Print database. At some time in 1996 the Books In Print database will be available, for a fee, on the Web. In addition the Web site will be available in certain situations on booksellers own Web sites, although with restricted information.

The CD-ROM product is intended for bookstores, not the mass market. In addition to a searchable version of Books In Print, an ordering system is built in. Book stores can, order the books they find, from a wide variety of suppliers. Of course, they must have accounts with these suppliers, but that is a normal part of the book business.

A company called MUZE has recently licensed the Books In Print database and will have versions available in kiosks for retail stores. MUZE currently has kiosks in music and video stores all over the world. Their Web site can be found at: http://www.muze.com.

Reed Reference Publishing is starting to exploit the value of their database in several ways. What was formerly a print only product has turned into a series of CD-ROM products, a Web site, and a licensable resource. This represents data reuse at its best.

This case study shows how a medium sized game company is using the Web to promote it's products.

Bethesda Softworks is the fourthlargest private entertainment PC publisher. It has produced a wide variety of PC titles including The Terminator: Future Shock, The Elder Scrolls: Arena, The Tenth Planet, and PBA Bowling. Chris Weaver, President, provided most of the information presented here.

The site, began operation in January 1995, receives approximately 30,000 hits per week. They receive approximately 250 email messages per day, with over 2000 a day during product launches. Sample games available on the site are downloaded at the rate of about 20-30 per hour.

Much of Bethesda Software's success is due to its highly detailed roleplayinggames (RPG) such The Elder Scrolls: Arena and its sequel Daggerfall. Following this line of thought, the Web site itself is structured as a selfguided tour through a medieval land, complete with detailed descriptions of the surrounding areas. The illustrations of the sites are highly detailed and follow medieval tones. For example, location names include: Archives, Castellan, Cauldron, Cistern, Courtyard, Tavern, and Tourney Field.

These locations often lead to "real" information about products, demos, press releases, and

employment opportunities. Clearly the goal is to provide an entertaining Web site that also points to "serious" content.

The site contains information about each of the companies game products as well as teaser images of soon-to-be-release games and demo versions. The Web site functions as an inexpensive recruiting mechanism with a "Billboard" of current job opportunities.

Through a registration process, you can register with the site. Registration has obvious benefits to the company, such as an ever growing list of names interested in the products. The benefits to the user, however, are questionable; the user must be enticed to register.

Bethesda accomplishes this by telling people, "Registering will make your stay here more comfortable, by allowing you to carry coin and by remembering which site you last visited. It will also reveal to you may secrets of Bethesda that would otherwise remain hidden to your eye." The coins registered users can collect can be used to get discounts on real Bethesda products.

One immediately noticeable aspect of the site, is that the URLs for each of the pages don't seem quite normal. This is because all of the pages are dynamically generated by cgiscripts. This allows Bethesda to serve up "text-only," "lite-graphics," "full-graphics," and pages for "registered-users," all from one common set of information. As you travel around the text is dynamically altered to reflect places visited and "coins" you have acquired. These "coins" can be used as discounts for Bethesda products.

The initial site design, developed by a group of people, took approximately 2 weeks. Maintenance takes about 3 to 4 hours per day, primarily due to the volume of email.

As the "winner" of a Point 5% top Web site award, and being selected as a Yahoo! cool site, clearly this site has accomplished many of it's design objectives. It's fun, interesting to experience, and provides useful content.

With the volume of email and downloads the site is clearly a powerful mechanism to distribute information. The site does, in a cost effective manner, distribute patches and help support existing customers as well as advertise and entice new ones. Future plans

This case study illustrates how an art supply company is able to advertise and sell its products via the Web. It is also shows how a small Web service provider has combined automated data base management with Web page design and electronic commerce.

This Maryland retail operation is entering the Web and electronic commerce in a big way. The concept is to create an on-line database both to manage products and to be used as the direct source of data for Web page generation. This system is called the "AccessAbility's Integrated Catalog." The Web designers and implementors are AccessAbility Internet Services Inc. Most of the information presented here comes from Scott Bodarky and Don Krapf of AccessAbility.

Store employees enter data via a convenient form-based interface. The structure of the data is flexible enough to accommodate virtually any product identification system already used by a merchant. An automated suite of programs automatically generates the actual Web pages. This is critical as data in the catalog changes frequently.

The concept is to give Web site developers access to an integrated database and Web site development suite. A custom designed database works in concert with Web page generation software to automatically produce the Web pages, once a database has been created. The process is iterative, allowing refinement of the site design and database management. The following sections discuss the process from the point of view of the three principal types of people using the site.

The Webmaster, the individual responsible for implementing the catalog, uses a form-based tool to create "prototype" files. These files, a combination of a style sheet and database definition are created by selecting the "Catalog Profile" section of the form.

Often the Webmaster and Merchant will be the same person, depending on the scale of the operation, of course.

The merchant is responsible for data entry and catalog updates. The catalog software is set up to handle catalog pages, detail pages, and images. Detail pages describe, in as much detail as desired, the product. One or several images can also be associated with the product. These images are named according to a merchant selected naming scheme for automated referencing.

The integrated catalog system can be viewed as a series of interacting components. All the products in a catalog are stored in the product database. The products are displayed on the web via the catalog display system. Information which applies to the catalog, rather than the products, such as how to calculate shipping and handling, is part of the catalog configuration. The integrated catalog system also provides a series of enhanced HTML resources which enable the site to perform a number of nifty functions, such as detect which browser someone is using.

The catalog is managed from the catalog control page, which provides access to all the components of the catalog. Orders are managed from the ordering administration page.

The merchant is responsible for creating and maintaining the product database (via the catalog control panel), writing and installing the prototype pages (again via the catalog control panel), and filling orders. All the shopping basket and ordering functionality is handled by the integrated catalog.

In typical Web site fashion, the customer can browse around the store adding or removing items from a shopping basket. An additional custom feature called "Smart Links," implemented as cgi scripts working with the server, enables the customer to go back to places previously visited. Customers can even leave the session and come back on another day to continue the shopping. The "Smart Links" feature is also a facility for the Web site designer as it eases navigation issues. The designer can simply place a "back" link in the

page and the user will go back to wherever they were, even if using frames. More unusual, the customer can create a record of purchases and locations in the virtual store, which can be used on subsequent visits to speed their way through. Finally, the customer can actually purchase the products via a secure Web server.

One final capability provided by AccessAbility is a technology they call "Enhanced HTML." Simply put, they have designed additional control commands which can be embedded within normal HTML. An enhanced HTML file is first interpreted by a program which works with the Web server, an Apache server, and the HTML is sent back to the client. An example application of this is to embed "if-then-else" type of control within the Web document. At run time the server, knowing the state of a user's travel's can decide, according to the enhanced HTML commands, to present the user with a particular Web page. This flexibility is only possible at run time and cannot be predesigned.

Let's look at some examples of the IF command. Suppose you are operating an online catalog, and wish to display special instructions for international customers. Your order form might have two checkboxes, one for customers who live in the US and one for customers who don't. Let's say that the name of the selector is "LOCATION", and that selecting "US" produces a value of 0, while selecting "non-US" produces a value of 1. We could test for this and provide conditional instructions as follow:

<P>For international shipping, the carrier we use varies depending on the destination. The shipping will be calculated and automatically added to your bill.</P>

While new, the Towson Art Web site, is an interesting experiment in integrated database and Web site management. The success of the site is as yet, untested, however the concepts have proven workable. In addition, Self service (Please see Section 1. 3 Web Maintenance in Chapter 1 World Wide Web for more discussion of self service Web services) clients of AccessAbility can take advantage of this powerful facility by designing their own catalogs using the many tools provided.

Here we see how one company leverages good oldfashioned email to get its message to customers.

Most of the information presented here is from email correspondence with Dianna Husum, Editorial Director of Tabor Griffin Communications, who was kind enough to answer my pesky questions.

How long have you been in business? Tabor Griffin has been publishing on the Internet since 1990. We started with the electronic version of the now defunct hardcopy Supercomputing Review. Eventually this evolved into HPCwire. In September 1994, we launched WEBster--The Cyberspace Surfer.

How long have you been publishing weekly? HPCwire has been weekly since its inception. WEBster is currently published every other week with plans to expand to weekly coverage.

How do you get your articles originally. (i.e. are they from wire services, or originally authored for WEBster?) We use a combination of wire services and freelance reporters.

Do you have any company standards for the process of getting the information into the central distribution point? We are set up like a traditional "newspaper" operation, in which certain editors are responsible for certain sections of each publication. We maintain a central database for notes, back issues and other research items.

How do you coordinate the work of all the authors? Again, it's pretty much a standard newsroom setup.

The managing editor is responsible for coordinating magazine content and freelance assignments. We operate on a query basis with freelance writers. Each week our staff "meets" via phone, e-mail, or conferencing to discuss issues, coverage, and deadlines for the next edition.

Are the "News Briefs" section articles primarily from press releases, and if so do you have any automation or semi-automation systems to help in the overall selection process? Articles in the news brief section do originate from press releases and newswire services. We use Compuserve as our news feed and set up filtered folders that capture files that contain specific key words. For WEBster we are viewing about 200 articles a day. For HPCwire it is only about 80. Then our editors view the capture files, pulling out those items that we plan to include in the issue. So automation is minimal, and a lot of handson copy editing and enhancement takes place prior to publication.

What kind ofcomputing equipment to you use? Our editorial/production department operates on 486s; back office, advertising, marketing, and accounting are on a combination of PCs and Macs and systems; and our Internet server is run off an SGI Indy and DEC station. So you could say we have a very heterogeneous network going.

What kind of mail system do you have as the server, or is it custom or a combination? We use a modified list serve.

Subject: WEBster Table of Contents: February 6 Content-Length: 9877 Gentle Reader,

with a number series starting with 6, 7 or 9. Plus you may retrieve the following WEBnews article:

Database integration is par for the course with MDG Computer Services release of Web Server 4D, a new server for the Macintosh.

TWO WAYS TO READ WEBster Send e-mail to i-want-it@webster.tgc.com In subject line, type the number immediately adjacent to the desired article or section. No message text is required. Example appears at end of this file.

Or check out our Web page at http://www.tgc.com/webster.html *****************************************************************************

***************************************************************************** SPOTlight (10155 this number retrieves entire section)

O.K. you've done all the research. Now its time to put all the information gleaned from those surf sessions to work. Here's

Three years in the making, Science for the Millennium is now officially open to the public. Through a video-rich, prototype online science exposition, visitors can explore advances in computing, networking and virtual reality.

Report on Rights Management Technologies: authoritative information dbHOTlinx: client-server information at your fingertips

Access WEBster's Back

**************************************************************************** RETRIEVAL EXAMPLE: To: i-want-it@webster.tgc.com

95006 )

HELPHELPHELPHELPHELPHELPHELPHELPHELPHELPHELPHELPHELPHELPHELPHELPHELPHELP ! HELP PLEH

Help me! I've ordered and I can't retrieve For help with retrieval problems send an e-mail message to sos@webster.tgc.com or human@webster.tgc.com for human intervention

What type of business model do you follow? We follow the model of a paid subscription magazine, essentially a traditional publishing model.

How many subscribers do you have? We essentially have two levels of subscription: paid/full access and unpaid/limited access. The paid group has full access to all articles ever published by that specific magazine. The unpaid group receives all of our mailings, but can only retrieve specific articles.

WEBster has about 38,000 addresses on its mailing list and about 6,000 paid subscribers. HPCwire has about 17,700 addresses on its mailing list and about 5,000 paid subscribers.

What other businesses are TGC involved in? We define ourselves primarily as a publishing house, but we also offer Internet and WWW consulting as well as publishing services for other businesses; (i.e., we have produced live publications for international events such as the G7 Ministerial Conference held last year in Brussels, and Internet publishing services for non-competing publications and internal newsletters for large corporations.)

Are you considering repackaging some of your archived material into other forms for sale? We have at times discussed this issue, but have not yet made any firm decisions.

What is your pricing structure? We have individual, group, and site license rates. HPCwire is priced at $97 per year for individuals or $49 per year for each member of a group of 3 or more. WEBster is $29 per year. These are flat rates and offer readers unlimited access in terms of numbers of articles ordered.

Can institutions get site-licenses? Yes, we have site license programs for each publication. These are based on a flat rate for a specific number of subscribers. The more subscribers you have, the lower the individual rate.

Are you seriously considering using any type of MIME data types such as audio or video in the future? We are always open to new content development or ways of presenting our material. However, the actual file type depends on our readers. We recently committed, our resources to Web delivery for both publications, this emerged from an overwhelming request from our readership. The difficulty with MIME and other specific data types is accessibility. When evaluating these technologies, it is important to keep in mind the mission of the publication (entertainment vs. information) as well as the nature of the readers. Our readers are primarily busy executives who are more interested in missioncritical information delivered in a timely, easytoread format rather than all the bells and whistles. Also at issue is portability. There is a lot to be said for good ol' ASCII. Everyone can accept ASCII files. Not everyone has MIME compatibility. So the key is determining the lowest common denominator for your readership and designing around that specification.

What do you see as the relationship between the HTML and email version? Our use of the e-mail version is two-fold. First, it serves as a reminder that new information is available. Think of it in terms of having your newspaper delivered to your front door every morning. A web page by itself is a very passive animal. The e-mail dynamic makes it more proactive.

Second, a large portion of our readership prefers reading the publication via e-mail as opposed to on the Web. So it remains an important delivery channel.

What is the hardest part of putting the publication together? I would have to say that the most challenging aspect is making sure that all the components of the process are in place and functioning in tandem from customer service to editorial, advertising to systems administration, and keeping the lines of communication open between those areas.

On the production side, the most difficult aspect is converting from ASCII to HTML, double checking all the links and all the other nitty gritty unglamourous details that suck up huge amounts of time and push you past deadline. Each issue of WEBster contains between 100 and 150 live outside links and about 100 "tgc"

What is the most surprising part of putting the publication together? With WEBster it would have to be the joy of discovery...the Web is growing so quickly that you never really know what waits at the end of the next link. The other surprise from both publications is the amount of feedback from readers. Coming from the "hard copy" publishing world, I am always amazed at the amount and extent of reader mail we receive and the "ownership" many readers feel toward these publications.(9)

The people at Tabor Griffin have clearly gone the low-tech route. Email provides a simple information dissemination method. Rather than try to create lots of new whiz-bang, content-free material, Tabor Griffin decided to stick with the real content and cost-effective distribution technology. The weekly summaries provided by both e-zine's get to the point and provide easy methods for article retrieval. Clearly, this type of distribution mechanism can be an effective tool for the dissemination of timely content.

The Internal Revenue Service (IRS) provides an example of how a large information provider, (tax forms and publications) uses the Web in coordination with the Internet to give people easy access to a large quantity of necessary information.

The Internal Revenue Service (IRS) an often despised government agency, actually has a sense of humor! Starting in early January of 1996, the IRS has created a Web site that's downright friendly. It's chock full of useful information and a terrific example of the imaginative use of advanced Web features. Most of the material for this case study was from information generously provided by Linda Wallace part of the Web team at the IRS.

A team of four IRS technicians (one part-time) and two research assistants develops and provides IRS electronic information services to the public. These products and services increase the responsiveness of IRS services and reduce costs. All the IRS applications for the Internet were conceived and laid out in-house.

The site was developed by contracting with FedWorld, part of the Department of Commerce's National Technical Information Service (NTIS), to provide operations and maintenance. NTIS then contracted with Websys, a small northern Virginia company specializing in Internet services, to develop the software and overall design. The content, applications and overall newspaper style layout was specified by the IRS staff.

In addition to the friendly 1950's art deco style the page features a globe that spins and a marquee of information. The opening screen lets the user select between a graphics or textonly version of the Web. The site actually detects browser types and adjusts the type of HTML delivered to the capabilities of the browser.

Before the Web site design and implementation, the IRS team benchmarked with a number of leading companies in electronic text management services to find industry best practices. They relied heavily on evaluation factors used in the Baldrige National Quality Award as they researched these best practices. Over the past ten years the IRS has had a strong commitment to SGML. They feel that without this strong foundation it would not have been possible to build this 3000 page Web site so quickly and accurately. The IRS text and graphics repository was a key component to the effective management of their data. The core business products, such as tax forms and publications, were moved into a self-help electronic access facility, the Web site.

More important than the Web site is the information it leads to. The IRS has placed all of their forms on-line in four formats. You can obtain the forms in PDF (Acrobat Reader format), PCL (Hewlett Packard's printer format), PS (PostScript) and SGML (Standard Generalized Markup Language).

The files themselves are located on an FTP site maintained by FEDWORLD, the premier disseminator of U.S. government information in electronic form.

In addition to the forms themselves, there are two versions of regulations on-line. One is the full text of various incomprehensible regulations. The second is a plain English version of the tax regulations for simple folk, like me. The plain English version exists as nicely formatted HTML pages.

In addressing the issues of business process and infrastructure, Linda Wallace (of the IRS) says:

The IRS staff concentrates primarily on applications development, identifying core business products and matching them to electronic media, and building infrastructure which will move the IRS into more multimedia service delivery. According to the Electronic Information Services staff, the infrastructure issue is the most difficult. It

means changing the way people build business products. Another challenge was helping internal business functions look at their products and services to find those that lend themselves to self-help electronic format and that are the top customer priorities.

As a customer driven organization the IRS really tries hard to keep their Web site responsive to, and focused on, the customer. One technical method they use is to make several Web browser interfaces available to the customer. These browser interfaces help people with "new and not-so-new technology and for persons with disabilities." Access to aswideas possible a customer base is very important. This must be balanced with the desire to present the most attractive Web site possible, which usually requires newer technologies, such as Java for example. Every Web page exists as HTML 3.0, HTML 1.0 and text only versions. In addition a variety of interactive applications, navigational metaphors, and search techniques make the search at hand easy and interesting for the novice and expert user.

Needless to say, a lot of people need to deal with the IRS. During the last week of the filing season (April 1996) the Web site was getting over 1 million hits every day. On the last day April 15th 1996 almost 1.9 million hits were logged. During the early time period of site deployment they were averaging 25,000 file downloads every day., and on the final day of the 96 tax season, over 100,000 files were downloaded. These include tax forms and publications.

The IRS Web site has in a short three month of operation received a great deal of (well justified) praise. They have won over 30 industry awards and great feedback from the media. A major accomplishment give most people's usual view of the IRS.

The New York Times said "Innovative and useful," and CNN reports, "Colorful, easy to

According to the IRS there have been over 52 million hits in about 3 months and most amazingly "people are actually writing in to say 'thank you' to the IRS." Imagine that!

The success of the IRS Web sites illustrates how to move a bureaucracy toward the future. Wide use of the information on this site, saves money, in this case taxpayer dollars. It's much cheaper for the IRS to provide information electronically than via traditional means. Technology is rarely the problem when introducing new services, usually old business practices are the enemy. Any large organization, government or private, has people with vested interests and resistance to change. This site above all else shows that it is possible to make real changes in established organizations and to do so with style.

This case study is a little special. Mostly it represents an attempt by yours truly to take material from one media, print, and transform it into another, the Web. First of all, the process has only barely begun. Initially, I was going to place the whole book with figures on the CD-ROM and Web site for this book. Time pressure and technical issues conspired to change the original project into another (more interesting) one.

Over the course of several months I will be turning the book into an integral part of the Case Studies section of this book's Web site at Prentice-Hall. l will document the process and problems encountered so that we may all learn from the process. I would especially like to thank Dan Gutman for letting me play with his book in this way.

Banana Bats & Ding Dong Balls: A Century of Baseball Invention by Dan Gutman is a traditional paperback book. It also represents an experiment, in putting existing material intended for one medium into electronic form.

I avoided the first problem, legal rights, because Dan Gutman is a friend who owns the copyright and I knew he could be persuaded, to let me experiment on his book. Next, I had to get the electronic form of the book. It turned out to exist in MacWrite format. Another reason for selecting this book is that it is a bunch of stories, and not a novel or piece of fiction that must be read from start to finish. It seemed to have potential for lots of hypertext links and jumping around.

Please stop by and visit our Web site at: http://www.prenhall.com to check on the progress. Finally it is my intent to continue to produce case studies and place them on the Web. If you feel that you'd like to explain how you put together a Web site, CD-ROM or other electronic publication please write to me at sandy@ability.net.