- From: Tab Atkins Jr. <jackalmage@gmail.com>
- Date: Thu, 12 Feb 2015 08:33:30 +1100
- To: 那磊 <leif.na@hotmail.com>
- Cc: "www-style@w3.org" <www-style@w3.org>
On Thu, Feb 5, 2015 at 8:14 PM, 那磊 <leif.na@hotmail.com> wrote: > Dear working group, > > I found that it's weird between the description in the eighth paragraph of > section 6.1. Relative positioning and the followed EXAMPLE 2. > > Here is, cause 'one of them has to be ignored', but in the case 'left' wins, > the 'right' becomes -'left' instead of ignored. So, is it an > self-contradictory or confusion error in describing? > > And here, in the EXAMPLE 2, the three 'direction' should NOT appear here but > appear in another rule to constrain the right containing block, established > by the nearest ancestor block-level box, which did'n appeared in this > example, should it? Cause the 'direction' you mentioned belongs to the > containing block of the relatively positoned elements, according to the > definition of containing block, that must is the nearest ancestor box, > instead of the 'div.a8' itself(themselves). So, the presence of 'direction's > in the EXAMPLE 2 is inopportune. If my understand is right, I suggest you > can add a note of assumption first. > > For your information, the key words above are: 'ignored', 'containing > block', 'direction'. One of the directions is indeed ignored in overconstrained situations, but "ignoring" is equivalent to setting one to the negative of the other; when "position:relative" is set, "left: 20px" and "right:-20px" are exactly the same. The example is just trying to make the effect a bit clearer. ~TJ
Received on Wednesday, 11 February 2015 21:34:17 UTC