- From: Jukka K. Korpela <jkorpela@cs.tut.fi>
- Date: Tue, 15 Apr 2008 16:24:45 +0300
Tina Holmboe wrote:
> If we wish to reproduce, as you mention, a work in which we can't
> really decide what the best /structural/ element sould be, then CSS
> comes to the rescue.
CSS is for optional presentational suggestions. When the issue is
presentation itself, it should no be a matter of suggestion. We should
be able to say, for example, "here's a word in italics, I don't know
why, but it is - or might be - essential that it is in italics". The
markup <i>foo</i> says this perfectly. The fact that in 999 times out of
1000 (would you believe 950 out of 1000?) italics is either purely
decorative or meant to express something definitely structural does not
change this.
One might conclude that such situations are so _rare_ that they don't
deserve markup elements in a language like HTML. But they _are_ in HTML,
so why take them away? (In an ideal markup language, we would perhaps
have just <font> with various attributes, instead of several font-level
elements.)
> Such reproduction is only important in a
> visual - ie. graphical - environment, since italics cannot be
> represented in either speech or "plain" text.
It cannot be presented as such, but it can be described, e.g. using
words or special expressions, like /foo/ or <italics> foo <end of
italics> - or even <i>foo</i>!
> Well. Unless we want the speech browser to actually read "This
> phrase was originally written in italics", but that is something it
> could derive from the stylesheet.
Why would it use such a stylesheet and not an aural stylesheet?
But it could read <i>foo</i> as "italics foo". Or maybe "| start italics
| foo | end italics |" where "|" indicates a short pause; this would
work better for <i> elements with many words inside them.
Basically, the difference between adequate and inadequate use of <i>
would be that for the adequate use, you would happily, even eagerly
accept such secondary ways of expressing italics, whereas for the
inadequate use, that would be foolish.
Jukka K. Korpela ("Yucca")
http://www.cs.tut.fi/~jkorpela/
Received on Tuesday, 15 April 2008 06:24:45 UTC