- From: David Hyatt <hyatt@apple.com>
- Date: Thu, 26 Feb 2009 13:09:13 -0600
- To: "L. David Baron" <dbaron@dbaron.org>
- Cc: www-style@w3.org
On Feb 26, 2009, at 12:22 PM, L. David Baron wrote: > On Thursday 2009-02-26 17:30 +0100, Michael Jansson wrote: >> I don't understand how the height of the single line is being >> computed >> for the middle block element in Firefox and Opera. I would expect >> it to >> be 100px, since none of the inline elements are taller than that. >> However, I get a taller block in Firefox and Opera. I can also see >> that >> there is actually two line boxes if I select the text in the >> browser. I >> only expect one. > > Neither is taller than 100px, but they're not at the same position. > The half-leading goes evenly on each side of the font, so the line > is supposed to end up taller than 100px. > >> Second mystery is that the text is not being text-bottom aligned in >> Firefox and Opera for the middle case. Intuitively, I wouldn't expect >> setting the line-height to affect the vertical alignment. I would >> expect >> case 2 and 3 to give the same result. > > This is because 'text-bottom' aligns the bottom of the box (which > includes the half-leading, which is different sizes for different > pieces of text) to the bottom of the parent's text. (In this case, > there isn't any text in the parent, but it still has metrics.) Where is it specified that the half-leading should be included when doing text-bottom/text-top alignment? I couldn't find this in the spec. If a font can be thought of as being broken into half-leading + ascent + descent + half-leading, then WebKit is currently aligning to the top of the ascent when text-top is specified and the bottom of the descent when text-bottom is specified. Is that wrong? The result in WebKit makes more sense to me than what I'm seeing in Firefox and Opera. dave (hyatt@apple.com)
Received on Thursday, 26 February 2009 19:09:57 UTC