Ordinary text (called `inline text' in this document) may contain a elements for special formatting and cross-referencing. Inline text appears in chapter, appendix and section titles, in the copyright summary, inside paragraphs and in other similar places.
There are a number of elements for denoting special significance of certain pieces of text. For all of them the end of the special text must be marked up explicitly, by using an explicit end tag <element>, the abbreviated end tag for closing the innermost element </> or the slash / which finishes the most abbreviated form of element markup (see SGML markup and metacharacters, Chapter 1).
Typically this will be represented by italics if available, or emboldened or underlined text, or in plain text formats with no character highlighting available by surrounding the text with asterisks like *this*.
Typically this will be represented by bold if available or in plain text formats with no character highlighting available by surrounding the text with asterisks like *this*.
Typically this will be represented by italics, or in plain text formats by surrounding the text with angle brackets like <this>. If several metasyntactic variables appear one after the other they should each be given their own <var> element.
This will usually be rendered using a fixed-width font; in plain-text formats quotes may be used around the element.
In output formats where character highlighting and various font styles are available this is usually represented by using a fixed-width font. In plain text output formats these names are not specially marked at all, as they occur frequently and introducing quoting would disrupt the flow of the text.
This will usually be rendered using a fixed-width font; in plain-text formats quotes may be used around the element.
This will usually be rendered using a fixed-width font; in plain-text formats quotes may be used around the element.
In formats where cross-references can be made non-intrusively by making a region of text a hyperlink without introducing in-line text this element will cause its contained text to become a hyperlink to the target of the cross-reference. In other formats this element will not have any effect.
There are a number of elements for introducing cross-references either to other parts of the same document or to other documents.
The intra-document cross-references are based on a scheme of reference identifiers. Each chapter, appendix, section, subsection etc. may have an id attribute giving its reference id - for example <chapt id="spong"> specifies that the chapter or appendix being started has reference id spong. This reference id can then be referred to in other parts of the document using the special cross-referencing elements.
The reference identifier will also be used for generating filenames and reference tokens for formats such as HTML which produce several output files; if no reference ids are specified then the chapter, appendix and section numbers will be used. It is a good idea to give at least all your chapter and appendix reference ids so that the filenames will not change if you change the order of the chapters, appendices or sections in your document.
The <ref> element does not have any contents; the chapter, appendix or section number and title and its page number or whatever is appropriate for the output format will be inserted into the text at the point where the tag appears.
Syntactically the cross reference is a noun phrase, suitable for uses like (see <ref id="...">) or further info is in <ref id="...">..
In some formats this will generate a true cross-reference which might (for example) be used to send email to the address quoted. In others it will just mark the text specially, usually including angle brackets < > around it.
Typically both elements will be rendered in a fixed width font; if possible, the <ftppath> will be made into a functional hyperlink to the named file or directory on the most recent <ftpsite>.
<ftppath> must always have been preceded by a <ftpsite> in the same chapter or appendix, but once one site has been named several paths may be appear.
Typically both elements will be rendered in a fixed width font; if possible, the <httppath> will be made into a functional hyperlink to the named file or directory on the most recent <httpsite>.
<httppath> must always have been preceded by a <httpsite> in the same chapter or appendix, but once one site has been named several paths may be appear.
Typically this element will be rendered in a fixed width font; if possible, id will be made into a functional hyperlink using name as place holder.
Footnotes may appear in most inline text, and are indicated by <footnote>...</footnote>. The text of the footnote itself will be removed and placed elsewhere (where depends on the format), and replaced with a reference or hyperlink to the footnote.
The contents of the footnote should be one or more paragraphs; the start of the first paragraph need not be marked explicitly. Inline markup elements such as character style do not take effect on the contents of footnotes defined inside them - the footnote gets a `clean slate'.
Footnotes may be nested, but this is rarely a good idea.
ardo@debian.org
ijackson@gnu.ai.mit.edu