- From: Arron Eicholz <Arron.Eicholz@microsoft.com>
- Date: Wed, 26 Oct 2011 21:21:20 +0000
- To: "L. David Baron" <dbaron@dbaron.org>
- CC: "public-css-testsuite@w3.org" <public-css-testsuite@w3.org>
On Tuesday, October 25, 2011 11:17 PM L. David Baron wrote: > On Wednesday 2011-10-26 00:23 +0000, Arron Eicholz wrote: > > http://test.csswg.org/suites/css2.1/20110323/html4/z-index-020.htm > > > Looking at this logic it seems as if the second reference for the test > > case is incorrect for the test. Since the outline is painted in step > > 7.3 it paints before the inline-block and thus will be behind the aqua > > background. > > Based on attempting to read through Appendix E, I think you're right about > what Appendix E says. (That said, either I'm misunderstanding how appendix > E works or I found a mistake [1] in it, so I'm not particularly confident > tonight.) I'm not sure if the WG really intended to forbid the option of > drawing outlines in a step between toplevel step (7) and toplevel step (8), > though (a step like toplevel step (10)), particularly given that > http://www.w3.org/TR/2011/REC-CSS2-20110607/ui.html#dynamic-outlines > says: > # This specification does not define how multiple overlapping > # outlines are drawn, or how outlines are drawn for boxes that are > # partially obscured behind other elements. > > I think it's a reasonable alternative; it's the current behavior of both Gecko > and WebKit. > I think you misread steps in Appendix E. The Webkit and Gecko, renderings are incorrect per the reading and explanation I gave of Appendix E and the optional position to paint outlines of 7.3. The outline (fuchsia) paints in tree order and before the following inline-block (aqua). If you reduce the case to the following markup and the same CSS: <div class="test container"> <div class="outline"></div> <div class="inline">FillertextFillerText</div> </div> The steps that are important here are the following in this order (super, super simplified): Step 1 of HTML for HTML Step 4 of HTML for BODY Restart at Step 1 for BODY contents Step 8 of BODY for DIV (test container) Restart at Step 1 for DIV (test container) contents Step 7.3 of DIV (test container) for DIV (outline) Restart at Step 1 for DIV (test container) contents so you can process the next element Step 7.2.1 of DIV (test container) for DIV (inline) The only 2 implementations that are "correct" per the explanation I gave are IE (there is no reference that is what I am arguing) and Opera (the top reference). -- Thanks, Arron Eicholz
Received on Wednesday, 26 October 2011 21:37:32 UTC