- From: Chris Lilley <chris@w3.org>
- Date: Sat, 22 Feb 2003 18:19:33 +0100
- To: www-style@w3.org, Etan Wexler <ewexler@stickdog.com>
- CC: Ian Hickson <ian@hixie.ch>
On Monday, November 18, 2002, 9:06:48 AM, Etan wrote: EW> Ian Hickson wrote to <www-style@w3.org> on 16 November 2002 in "Re: EW> word-spacing property" EW> (<mid:Pine.LNX.4.21.0211161216070.12577-100000@dhalsim.dreamhost.com>): >> The spec is already _very_ clear -- CSS cannot affect the DOM (the >> underlying text) in any way. EW> When I wrote "underlying text, I did not mean the Document Object Model. EW> Although the DOM, too, is underlying text, I meant the text exposed to EW> operations such as copy and search. Copy yes; search, that is not clear. EW> Depending on the user agent, CSS might affect this intermediary layer of EW> content. The 'text-transform' property, for instance, changes characters EW> to uppercase or lowercase equivalents. While this transformation may not EW> touch the DOM, It does not touch the DOM, as you say (unless the Views module gets revived). EW> it will likely change what the end user gets as content. It changes the presentation, yes. In CSS, the rendering tree is 'almost' like the content tree; or is a copy of that tree with changes (extra elements for first line and first letter, :before and :after, implied table and table tow, and so forth). word-spacing and text-transform similarly affect the copy of the text that is in the rendering tree (as opposed to using a small-caps font, which has no such effect on the text in the rendering tree). -- Chris mailto:chris@w3.org
Received on Saturday, 22 February 2003 14:22:24 UTC