- From: Jukka K. Korpela <jukka.k.korpela@kolumbus.fi>
- Date: Mon, 06 May 2013 16:56:44 +0300
- To: public-html-comments@w3.org
2013-05-06 16:33, Ian Yang wrote: > But when > you use <p> element purely as container, is that really a semantic > approach? No, and there is very little semantics in HTML in general. The word "semantic" as usually used as a buzzword really refers to structure, not semantics (meaning). In reality, both in HTML tradition and in HTML5 drafts, the <p> element means a block of text that may contain text-level markup, but not other markup. It largely corresponds to the word processor concept of paragraph, except that the latter is even wider: anything formatted as a block is a paragraph. When some text and text-level markup needs to be marked up as a block, there are usually two basic options: <p> and <div>. In special cases, other markup may apply (e.g., <blockquote> or <li> or <td>), but this depends on context. The real differences between <p> and <div> are: 1) <p> implies empty lines before and after, in default rendering, and may cause (short) pauses before and after in speech rendering 2) <div> content is not limited to text-level markup but can contain e.g. a list or a table. It's pointless to try to describe the difference in more "semantic" terms, or even in more structural terms. Such descriptions just result in confusion, endless questions, and debate. Of course you *can* ignore difference 1 if you think that non-CSS rendering does not matter, on the grounds that you can set vertical margins in CSS as desired. Then it would be natural to use <p> for all basic blocks of text (i.e., blocks that contain only text and text-level markup). And "block" can be seen as a physical concept, which makes things easier, or as a logical concept, in which case your would need to refer to things like dealing with a specific topic as a logical unit, such as one passage (paragraph) in a novel, a collection of interrelated data, or a unit of a form consisting of a label and associated field. Things would probably be easier if the word "paragraph" were not used at all (except perhaps when referring to paragraphs of prose) and <p> were defined as denoting a block of text. -- Yucca, http://www.cs.tut.fi/~jkorpela/
Received on Monday, 6 May 2013 13:57:08 UTC