- From: Ian Hickson <ian@hixie.ch>
- Date: Wed, 13 Apr 2005 11:20:27 +0000 (UTC)
On Fri, 7 Jan 2005, dolphinling wrote: > > "green" is just as meaningful as "subscript"--they're both purely > presentational, and we as people have attached meanings to certain > presentations. The semantics of "subscript" are completely different > from the semantics of "there are two of the (chemical) element that > immediately preceded this", but we have attached one to the other. There are some differences: * The semantic of <sub>/<sup> can often be deduced from context (especially if we define some specific contexts, like "sub inside var means it's the variable's subscript, as in the part of its name that identifies the specific variable in a family of variables"). * There are rarely, if ever, any semantics associated with a colour that can be understood without an explanation in the same document. Sometimes some content is coloured so that it can be differentiated, as in making test results red or green for different results, but that is always explained in the same document (if it is to be understood, anyway) and is rarely the only aspect of the document that indicates the difference (e.g. typically in such an example it would be the words "pass" and "fail" that were coloured). I would love to be able to unambiguously have the semantic for the "2" in "H<sub>2</sub>O". I don't see how to do it (at least not without introducing a ridiculous number of elements to HTML). I consider <sub> to be an acceptable (semantic) compromise. It does affect media other than visual media, which (as previously mentioned) is the criteria I usually use for working out if something is semantic or presentational. Another criteria is "could the presentation be changed without losing its meaning?". For example, with <em> clearly you can change the presentation without losing the fact that it is emphasis: whether it is bigger or italics doesn't make much difference. But with <sub> I don't think you can. If you change the presentation of <sub> you _do_ change the perceived meaning of the rendered content. -- Ian Hickson U+1047E )\._.,--....,'``. fL http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,. Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'
Received on Wednesday, 13 April 2005 04:20:27 UTC