Don't start out without a road map!

If you are interested in SGML and XML, who could blame you for wanting to jump right into this book. You have in your hands the most comprehensive collection of detailed descriptions of SGML and XML resources ever assembled -- and even more than that.

There is such a lot here that it's worth taking a minute to get the whole picture. In fact, we so much want you to read this introduction that we've moved the typical front matter distractions, like the preface and acknowledgments, to page 661.

We promise that you absolutely, positively, do not need to be an SGML expert to read this introduction, or 96% of the rest of the SGML Buyer's Guide™ But the experts who have read it agree that the Guide offers insights even for experts, and that the place to start reading -- for everyone -- is right here.

So go ahead, browse through the book, but then come back and read this introduction before you read anything else.

What this book can do for you

Collected right here, in one place, you will find:

• Authoritative descriptions of more than 150 SGML and XML tools and services.
• Offerings categorized according to function based on the famous Whirlwind Guide to SGML Tools.
• Expert advice on what to look for in each tool and service category and what to ask the vendors and service providers,
• A CD-ROM with extensive SGML and XML materials, live demos, trialware, and 45 great genuinely free-use no-time-limit software products, including some available nowhere else.
• Descriptions and contact information for the support infrastructure for SGML and XML: organizations, books, periodicals, and other resources.
• The Sponsor Showcase Supplement, a print and CD resource where SGML and XML vendors and service providers present their own information directly to you.

You could use the SGML Buyer's Guide like an ordinary directory or catalog, flipping through its pages, reading about the tools and services that are available, and looking for the ones that are right for you.

But how do you know which ones are the right ones for you?

Every organization has unique requirements, just as it has unique products, editorial processes, working methods and even cultures. Indeed, that's one of SGML's great strengths: rather than force-fitting your organization's information and work methods into the formats required by a product, you can fit the products to your requirements.

The rewards of using SGML are well documented. In fact, much of that documentation is in resources described in this book. For example: $GML: The Billion Dollar Secret, described on page 627, tells how visionary companies use SGML to turn document bucket-brigades into information pipelines.

However, the very desirable flexibility that SGML allows can complicate the purchasing process. Really knowing your requirements is key.

So the SGML Buyer's Guide also provides a method you can use to help determine your precise requirements: HARP™ analysis. Using HARP analysis you can better understand aspects of your document processing system that are typically overlooked, but vital to building a system that works for you. HARP analysis will enable you to:

• Assess your publishing operations in an intuitive, visual way.
• Inspect your systems from the unique perspective of "document representation" -- how the information is stored in the computer.
• Find out that your processing can be made more efficient just by changing the way you currently use representations.
• Indicate where really big payoffs can be obtained from workflow analysis and reengineering.
• More precisely determine your needs for SGML and XML products and services.

Bear with us for a bit and we'll show you why HARP analysis should be an important part of the process you are engaged in.

Does your company manage its information?

How is information managed in your company? The answer depends on what type of information we talk about.

If we talk about the type of information that can be stored in database fields -- financial data, customer information, inventory levels -- then the information is probably under tight control. You use databases with appropriate access and security restrictions to collect, verify and maintain the information at all times. The information is stored in a well-understood format so that it can be used and reused for a variety of applications effectively. Your technology specialists analyze the requirements for new systems carefully before they build them, and changes to existing databases -- such as adding a new field of information into the accounts receivable database -- are accomplished only through a very careful, very formal procedure.

However, if we talk about the type of information stored in documents -- everything from owner's manuals and regulatory filings to contracts, brochures, even email -- then the information is probably loosely controlled if it can be called "controlled" at all.

For one thing, your documents are likely to be physically out of control. Unless your organization uses document management and workflow systems, you lack the means to reliably store, version and manage documents in a way that gets the right data to the right desktop at the right time, in a controlled and audited manner.

But document managment is just part of the puzzle. Even with a system of controls in place, the ways in which your documents are represented inside the computer are still likely to be out of control. Decisions about crucial issues like document file formats ("notations" in our terminology), character sets, and storage organization, are almost always made by default. They are the by-product of individual decisions to purchase a program for a single end purpose -- to format the pages of your owner's manuals or write Web pages. The fact is, document representation has a definitive impact on what else can be done with the information in a document. But that is rarely appreciated -- until someone else later tries to reuse it. It is almost never taken into account as an explicit part of the purchasing decision.

In contrast, enterprises that have learned to manage documents as information have done so by mastering a simple principle: Know your document representation! Don't let the software you buy determine the representation. Let the representation you need determine the software you buy.

HARP™ analysis is the tool to help you do exactly that. HARP analysis is a new technique for visualizing your information systems. It lets you see the representations you need in order to produce the information products you want, and then turns that analysis into requirements for tools.

Using HARP analysis will help you make the management of document representations an integral part of the purchasing process. It will help you determine what tools and services you need to manage your documents as real information assets.

Why care about document representation?

But what is all this fuss about "document representation" anyway? Why should you be concerned about how some software program you buy stores its files inside the computer? Shouldn't that be the software's problem? After all, isn't creating the right storage format part of what you are paying these software companies to do?

The answer is both yes and no. It is "yes" for individual outputs. Most software companies produce products that create one type of publication product and produce it well: print, electronic publication, slide show, Web page, etc. If you want the output and you like the product, you'll need to get your information into their format.

But the answer is "no" for your organization as a whole, because the right storage format for one product is the wrong storage format for all the others. If you don't take these formats and the other traits that go to make up what we call "document representation" into account when you design your system, then these proliferating, incompatible file formats will eat up your productivity.

Because representation is a key to productivity

When what is going on inside your computers is left out of your system planning and design, the result is copies upon copies of documents floating around your organization. Each copy is in the representation of a different department's word processor, slide-show package, or desktop publishing program. There's little opportunity to reuse common information and almost no way to leverage your investment in "intellectual capital" into new information products. Even keeping supposedly "identical" documents in sync becomes a maintenance headache.

Some organizations try to solve the problem by mandating a corporate standard word processor (thus implicitly mandating one standard representation). But even this solution usually fails, because different groups end up using such different page layouts and rendition formats that the document files they produce -- which faithfully capture those renditions -- are useless for anything beyond printing and reading. Their information still can't be shared.

And because document representations are subject to no centralized control, different groups, even individual writers, can decide to make changes, minor or major, to the way their documents are rendered without consulting anyone else. That means they change the computer representations as well, minimizing or eliminating their usefulness for other purposes. The end result for your company is a lot of paper -- and very little genuine, properly managed "information."

This ad-hoc approach has a very real -- and very negative -- impact on productivity. That impact is usually hidden because it happens where information moves across the boundary from one department, one group to another. But you can see it:

• When your training department works with draft user manuals from corporate communications...
• When customer service depends on the new feature bulletins and technical memos coming out of engineering...
• When manufacturing needs the schedules, specs and plans developed by project management...

Then the lack of control over documents makes it difficult for anyone to develop reliable, smoothly-functioning procedures.

Because publishing itself is changing constantly

There's even more to the problem than that! The changing definition of "publishing" itself is making our ad-hoc approach to documents more and more costly and constrictive. Once upon a time, we only had to concern ourselves with producing paper. Back then, problems managing document representations had a relatively minor impact. But today, when we use the phrase "publishing system," we are talking about a growing variety of channels for capturing, manipulating, delivering, and selling information.

• We can produce large volumes of printed, bound books, or we can use "print-on-demand" technology to deliver small slices of selected content.
• We can create up-to-the-minute analysis and commentary on volatile fields like finance, law, health care, or world affairs and immediately push it out to customers over a broad array of newswires, faxes and inter- or intranets.
• We can publish huge volumes of information on CD-ROMs.
• We can offer content over online services, and we can deliver information over the World Wide Web.

And our content is becoming increasingly interactive. Where once our customers might have been happy to get electronic versions of forms, now they expect the forms to do the work for them -- filling in known data, calculating final results. They expect electronic documents to be part reference, part expert system, part computer game. They expect this because if we aren't doing it, our competitors are.

Because information delivery is increasingly costly

We want to be able to take advantage of these opportunities, but we don't want to go broke doing it. And delivering information to those who need it is becoming increasingly costly.

The problem is that every one of these delivery mechanisms requires our content to be in a different representation, and changing representations costs money.

And if it has to be done by hand -- as it does in most organizations -- then it costs a lot! Lots of money, certainly, but also the time of skilled people who are needed urgently elsewhere.

To rework your content over and over for every new use is both time-consuming and expensive. Yet, if you are like most, your ability to build systems that efficiently process large volumes of content and turn it into new products is frustrated by the very ad-hoc nature of your publishing systems.

SGML is the right choice to solve these problems because it deals with document representation the right way. It does not eliminate the need for the individual representations that produce each output. But SGML lets you build the foundation, the fundamental representation, from which to generate those other representations cost-effectively.

SGML is the tool that can streamline the movement of your content (that intellectual capital your knowledge workers get paid the big bucks to produce) into those output-oriented forms most efficiently. SGML vendors know this and their products reflect it. SGML service providers know this. Now it is time that you found out about it too.

HARP™ analysis -- a new insight into publishing systems

But we don't really need to convince you that today's requirements for publishing systems are far more challenging than they have been in the past. The fact that you are reading this book suggests that you are grappling with this problem right now. You may even be in the process of specifying a new system for your organization. If so, you are probably experiencing some of the frustrations of finding the right combination of products and services that will satisfy your requirements. You may even be finding it difficult to determine what your requirements are!

This is exactly where the SGML Buyer's Guide will help. The SGML Buyer's Guide gives you new ways to analyze publishing systems and determine your requirements. These new insights do not focus on the physical side of the problem (the management of processes and files). The physical aspects of system analysis are well covered by workflow analysis and structured design methodologies like SADT (Structured Analysis and Design Technique) and Structured Systems Development. Rather, the SGML Buyer's Guide complements these methodologies with a new technique that gives you the tools to analyze the representation side of the problem.

The new technique is HARP analysis. HARP analysis is a methodology for analyzing what happens -- or what needs to happen -- to your information as it moves through your system from initial creation to final delivery.

• It enables you to capture critical information about the document representations (file formats, storage organizations, etc.) you use today and how they impact your system's productivity.
• It shows you how to determine the representations you need in order to produce the information products you want, in the most expeditious way possible, without boxing your organization into one proprietary corner or another.
• Finally, it shows you how to turn these insights into requirements that can be used to identify products and services, and to transform your plans into functioning systems.

You can either use HARP analysis on its own, or as a complement to workflow and business process analysis. In Part I of the SGML Buyer's Guide we introduce you to the visual component of the methodology through a graduated series of examples. Our goal is to teach you the basics without making it feel like we're back in school.

But the visual aspect of HARP analysis is just a graphical user interface. The complete methodology is described in Part II of the SGML Buyer's Guide. If you are experienced in system analysis and design, and you want to get right to the full story, you can turn directly to Part II for a detailed exploration of HARP analysis.

The SGML Buyer's Guide™ -- a tool for buying tools and services

The SGML Buyer's Guide is intended to be a tool, one that will help you determine your precise requirements for an information delivery system and identify the products and services that can help you build it.

Do you suspect that your systems for producing publications suffer from some of the problems described above?

Do you see your organization struggling to get information out on time and up-to-date?

Can you relate to the insidious way that multiple data formats impede your organization's productivity?

Are you investigating SGML and XML as a possible solution?

If you've answered "yes" to any of these questions, then you are holding the right book in your hands.

• It will give you the tools to assess the representations (the file formats, the character sets, etc.) used in your publishing operations in an intuitive, natural way.
• It will let you inspect how representations are created and transformed from one to another.
• It will highlight where the representations are causing bottlenecks and other obstacles to productivity.
• It will point out areas where changing the representation alone can improve processing, even within the same business process workflow.
• It will signal opportunities to improve your workflow by reducing the number of changes in representation, and help you focus workflow analysis on the right conditions to allow those opportunities to be realized.
• And it will help you to develop a list of the categories of products, the capabilities and requirements, and the processes and outcomes, that you need to build that better publishing system.

The SGML Buyer's Guide consists of five Parts, plus a Supplement and a CD-ROM.

Parts I and II introduce HARP analysis, the technique for analyzing and diagraming the sequence of internal representations used in any system that creates, collects, manages and publishes information.

Parts III, IV, and V provide a guide to tools, services, and resources, organized by the types of task they can perform (and cross-referenced by provider, product category, and other useful criteria). These Parts report objective facts about the products and services, along with descriptions supplied by their developers. We have deliberately avoided product comparisons and ratings because, ultimately, only you can decide which capabilities and features are important for your application.

Using the material in Parts III, IV, and V, you can identify those products, services and vendors that can meet your requirements and help you turn your planned system into a real, working one. Of course, you can use these Parts for that purpose whether or not you use HARP analysis to help determine those requirements.

A special supplement to the SGML Buyer's Guide is the Sponsor Showcase, a section where our sponsors provide you with their information about their products and services. Their story continues on the accompanying CD-ROM, which also contains demos, trialware, and freeware, so you can try products before you buy them. (We urge you to browse the CD-ROM. It has some wonderful software that other books would trumpet on their covers, but which we can't mention because we don't want to favor one sponsor over another.)

Unlike the material in the book proper, the content in the Sponsor Showcase is unabashedly subjective. It ought to be; it comes to you directly from our sponsors, and it's the place where they get to blow their own horn! Turn to this section to get "up-close and personal" with providers who interest you.

You are in the market for new tools. As you read this book, you will find that the SGML Buyer's Guide is much more than just a catalog of products. It is itself a tool; the very first tool you need to get your documents, and the systems that produce them, performing to the demanding standards of your company and your customers.