Essentially, this volume is two books in one. It is a heavily annotated descriptive bibliography of all of the standards, technical reports, draft standards, and would-be standards relating to Standard Generalized Markup Language (SGML). It is also an introductory class in SGML, condensed into 151 pages. The tutorial material will be of great value to people who are considering using SGML or are trying to make their initial selection of SGML systems and services. Each of these intermingled parts would probably benefit from separation. The portions of this book I like the best describe the basic concepts of SGML and their implementations concisely and accurately. The portions that describe the standards development process and the standards and technical reports that support the standard, while easier than reading the documents they describe, are quite heavy. I fear that the first chapter and the first sections of most chapters will scare off readers who could learn a lot from the rest of the book.
The organization of the book is similar to the organization of most courses on SGML. The “Introduction” includes a brief description of the objectives and uses of SGML. It then describes the SGML standard, a host of related standards, and the organizations that develop and promulgate them. The second chapter, “The Standard Generalized Markup Language,” describes the theoretical foundations of SGML and markup, what SGML is, how it works, and the options (called features) of SGML. The explanation of features includes descriptions of each, how they are used, and the author’s opinions of when they should be used.
Chapter 3 defines the three parts of “An SGML Document.” The functions, format, and syntax of the SGML declaration, document type definition, and document instance are described. This chapter also includes a section on “reducing overheads,” by which the author means reducing file size, which she seems to feel is far more important than I do. “Some Application Considerations” is a chapter on uses of SGML and good SGML practice. It discusses using SGML to encode the content and structure of a document rather than its format; use of implicit numbering; and handling of lists and tables. The discussion of tables is a passionate (but not very logical) plea to use some of the SGML features described in “An SGML Document” to make the coded version of tagged tables short and easy to read.
“The Integrated Information Base” discusses the concepts of creating and storing information rather than documents. The corporate information base the author advocates can be used both as a management information system and as a publishing system from which a variety of publications in a variety of formats can be produced. Guidelines for designing and using such an information base are presented along with the advantages of publishing from an integrated information base.
One of the biggest issues in SGML application development is addressed in the chapter entitled “Presenting SGML Documents.” SGML is an encoding system and must be rendered, or presented, to be used. Rendering systems for SGML include typesetting, desktop publishing, speech synthesis, and electronic display in hypertext or full-text search systems. The chapter begins with a discussion of the draft Document Style Semantics and Specification Language standard, which will be a standard for specifying the presentation of SGML documents.
“SGML Support Facilities” includes descriptions of standards and technical reports relating to the interchange and registration of SGML documents, as well as describing several related technical reports that are being drafted. “Hypermedia and Other New Work” briefly describes the new HyTime standard (an ISO standard for hypertext and hypermedia), the Standard Music Description Language, and binary SGML, which is still in the concept-only stage. In “Some Applications of SGML,” short descriptions of commercial, defense, and civilian government uses of SGML give some indication of the current usage.
“A Comparison of SGML and ODA” is the best description I have seen of these two standards, which many people see as being in conflict. The goals, histories, structures, and status of the two standards are objectively compared. This chapter is particularly useful, since most mentions of ODA (Open Document Architecture) in SGML documents are irrationally negative.
“SGML Product Guidance” includes an excellent list of questions to ask a software vendor. These questions are also an excellent test of knowledge of SGML; if you understand the questions and the answers to them, you know enough to make rational system component selections. If you do not understand the questions or answers, either learn more about SGML before spending any money or get an expert to help you get started. The listing of products is, as the book states, incomplete, but it does include all of the major vendors and most of the popular software products available. While this chapter will quickly be out of date, the final chapter, “Further Information,” describes the major sources of information about SGML, including addresses and telephone numbers. The reader can use these sources to get more current information on SGML standards efforts and products.
I recommend the book to two different audiences. SGML novices should skip the first chapter and the sections about standards and technical reports. For them, this book will be an excellent introduction to SGML. SGML users and developers should study the first chapter and the sections about standards, and review the rest of the book to see how clearly some difficult concepts can be described.
