Re: Tables, paragraphs, and structure (was Re: colours for bullets...)

Abigail (abigail@fnx.com)
Thu, 17 Apr 1997 19:13:06 -0400 (EDT)


From: abigail@fnx.com (Abigail)
Message-Id: <199704172313.TAA15926@fnx.com>
Subject: Re: Tables, paragraphs, and structure (was Re: colours for bullets...)
To: davidp@earthlink.net (David Perrell)
Date: Thu, 17 Apr 1997 19:13:06 -0400 (EDT)
Cc: www-html@w3.org
In-Reply-To: <199704172009.NAA15757@denmark.it.earthlink.net> from "David Perrell" at Apr 17, 97 01:00:59 pm

You, David Perrell, wrote:
++ 
++ IMG. A table is typically an ordered arrangement of data that clarifies
++ some relationship, and it often relates to a specific phrase in a
++ paragraph. There is no conceptual difference between tables and images
++ in this regard, and an image could easily be (and sometimes is) a
++ bitmap representation of a table. Would it be reasonable to render the
++ table inline as can be done with images? No. But it is reasonable to
++ tie the position of the table to a particular place in the paragraph.
++ Since there is no way of knowing before rendering time where that
++ position will be relative to the top of the paragraph, this can only be
++ done by embedding the table element within the paragraph.
++ 
++ Re OL and UL: A list rarely if ever stands alone as a concept. A list
++ is usually an appendage to a paragraph, not a separate entity. To give
++ a list equal space above and below is not a reasonable presentation.

One can use DIV around the paragraph and the list to indicate
a relationship.

++ Re the 'structure' of HTML in general: Is it a tree? Does each
++ subsequent lower heading level grow a new branch? Are paragraphs
++ children of the preceding heading? The existence of DIV implies there
++ is no such structure. To call six heading levels, paragraphs, arbitrary
++ divisions, and a bunch of other content models 'document structure'
++ seems a bit of an overstatement to me.

It is a tree, though not as you mention it. HTML has elements, and
elements are containers. If you make a directed graph G, where each
vertex V in the graph represents an element el(V) in your document, and
there is a link from vertex V to vertex W, iff el (W) appears directly
in el (V). Obviously, there cannot be a cycle: no element has 2
parents, and no element can contain one of its ancestors. Hence G is a
tree.


Abigail