A Practical Introduction to

SGML, the Standard Generalized Markup Language, deals with the structural markup of electronic documents. It was made an international standard by ISO in October 1986. SGML soon became very popular thanks in particular to its enthusiastic acceptance in the editing world, by large multinational companies, governmental organizations, and, more recently, by the ubiquity of HTML, HyperText Markup Language, the source language of structured documents on WWW. This article discusses the basic ideas of SGML and looks at a few interesting tools. It should provide the reader with a better understanding of the latest developments in the eld of electronic documents in general, and of SGML/HTML in particular. 1 Why SGML? Since the late eighties we have witnessed an ever quickening transition from book publishing exclusively on paper to various forms of electronic media. This evolution is merely a reeection of the fact that the computer and electronics have made inroads into almost every facet of human activity. In a world in which one has to deal with an ever-increasing amount of data is support of the computer is a particularly welcome alternative, for the preparation of telephone directories, dictionaries, or law texts|to mention just a few examples. In such cases it is not only the volume of the data that is important, but also the need for it to be kept constantly up-to-date. Once data have been stored in electronic form one can derive multiple products from a single source document. For instance, an address list can be turned into a directory on paper, but it can also be put on cdrom, as a database allowing interactive or e-mail access on the Internet or to print a series of labels. Using a set of law texts or a series of articles on history marked up in SGML, one can rst publish a textbook containing complete law texts, or a historic encyclopedia, and then provide regular updates or extract a series of articles on a given subject; one can also ooer a consultation service on Internet, via gopher, www or develop a hypertext system on cdrom. All these applications suppose that the information is not saved in a format that is only suited for printing (for example, WYSIWYG), but that its logical structure be clearly marked. To recapitulate, the strong points of a generic markup (in SGML) are the following: the quality of the source document is improved; the document can …