- From: Ian Hickson <ian@hixie.ch>
- Date: Mon, 24 Oct 2005 04:12:56 +0000 (UTC)
- To: Boris Zbarsky <bzbarsky@mit.edu>
- Cc: www-style Mailing List <www-style@w3.org>
On Sun, 23 Oct 2005, Boris Zbarsky wrote: > > > > The language you mention doesn't affect this; an inline-table on the > > first line is still on the first line > > I had a table, not an inline-table. Oh right. That would be like having display:block inherit, then. > > What UAs generally seem to have done IIRC is made properties that > > normally inherit inherit from the ::first-line (and thus have two > > computed values, one for each part -- I'm not sure which they return > > for the DOM), and made properties that don't normally inherit inherit > > from the parent element. > > What Gecko does is to inherit everything from the first-line, but make > its decisions on what layout objects to create _before_ any of that > happens, so that the first-line can't affect that. Which of course > means that the style data and rendering model don't match. ;) Oh, and > all this is during initial page layout. On incremental changes the > behavior is highly inconsistent, depending on the ordering of changes > and exact types of changes... > > Which is what made me look into this to start with. :( The effective result is what I described, though; maybe one way of doing that would be to change to the model I described. If you do try that, please let us know how it works, so we can see if that might be want we want the spec to say. Did you imply that getComputedStyle() always ignored ::first-line? That would be one way of defining it which would work. -- Ian Hickson U+1047E )\._.,--....,'``. fL http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,. Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'
Received on Monday, 24 October 2005 04:13:02 UTC