HTML Basics
Stanford University Libraries & Academic Information Resources

Exceptions

HTML was designed to be simple, and unfortunately the side effects of that design goal are occasionally unfortunate. Structural features that would have helped make the structure of documents clear, and helped to clarify the distinction between content and look & feel were left out of HTML in order to make it easier for authors to write and browsers to process.

It would have been beneficial, had HTML provided a section/subsection structure, with a heading for each section/subsection, much as most other major SGML-based markup languages do. As structure like the following is effectively impossible to represent in HTML as it exists today:

    <section id="sec1">
    <heading>Chapter 1</heading>
        <!-- This is part of sec 1 -->
        <p>My life has been rich...

            <subsection id="subsec1">
            <heading>Early School Days<heading>
                <p>First I went to school...
            </section>

            <subsection id="subsec2">
            <heading>Later School Days<heading>
                <p>The I went to school some more
            </section>

        <!-- This is part of sec1 -->
        <p>Yes, indeed my life has been rich

    </section>

It would also have been helpful if HTML had provided elements to designate at least some of the elements that appear on title pages. For example, there is no tag to indicate a subtitle, so the long title/subtitle combinations we encounter in the academic environment are really rather difficult to handle. It is expected that the Title be marked as <h1>, but clearly a subtitle needs to be distinguished from the title, so our only plausible choice is <h2>, which is really quite misleading. Similar problems obtain for other title page elements.

Future versions of HTML may clear up at least some of these problems. The wider use of richer SGML Document Type Definitions, such as those of the Text Encoding Initiative, can solve these problems in several ways either by automated conversion of the richer markup to HTML (which is then seen mainly as a delivery format), or by coupling (or replacing) traditional Web browsers with SGML-aware browsers, so that richer SGML documents can be disseminated via the Web and viewed in their full complexity.