- From: Boris Zbarsky <bzbarsky@MIT.EDU>
- Date: Mon, 14 Mar 2011 11:35:11 -0400
- To: Peter Moulder <peter.moulder@monash.edu>, www-style@w3.org
On 3/13/11 4:16 PM, Peter Moulder wrote: > On Sun, Mar 13, 2011 at 02:47:43PM -0400, Boris Zbarsky wrote: >> For what it's worth, I believe this last case doesn't matter in the >> sense that the rendering is the same either way... > > If "the inline box [is] broken around the block-level box [the first > one, i.e. prior to this hypothetically separating inline element], > dividing the inline box into two pieces even if either side is empty", > and if it's considered clear that the two end boxes each take up a > line's worth of vertical space This last part is not clear, in fact. In Gecko they don't if they have nothing in them (where "nothing" can include collapsible whitespace and placeholders). > and have their border area rendered with > the inline's background& border They do, but in this sort of textcase: <span> <div/> <div/> </span> that middle inline ends up with a width of 0 (so the background,top,and bottom borders are invisible), and it doesn't render its left/right borders, of course. > What am I missing? I think the key issue is the vertical space. > Curiouser and curiouser: in the case of Gecko, such an inline is sometimes > considered separating and sometimes isn't "separating" in what sense? > in the following test case, Gecko > considers it to be separating where the separating inline has its own border > (between block3 and block4), but not where it doesn't have its own border > (between block2 and block3). That makes sense per the above, no? > Also of interest is that Gecko doesn't render > the parent inline's background& border for non-empty content that separates > two blocks: Er... are you testing this in Gecko 1.9.x or something? If so, don't. This section of CSS 2.1 wasn't implemented correctly until early in the 2.0 cycle; before that all the kids of the inline from the first block kid to the last block kid were wrapped in a single anonymous block, with the inline broken around that single block. -Boris
Received on Monday, 14 March 2011 15:35:52 UTC