- From: Alan Stearns <stearns@adobe.com>
- Date: Wed, 18 Jul 2012 15:17:16 -0700
- To: "www-style@w3.org" <www-style@w3.org>
On 7/18/12 8:40 AM, "fantasai" <fantasai.lists@inkedblade.net> wrote: >On 03/16/2012 11:02 AM, Vincent Hardy wrote: >> Hello, >> >> In section 4.1, the possible break points section says things like, >>about sibling boxes: >> >> "when the block flow direction of the sibling's parent is parallel Š" >> >> This refers to a notion of parent for layout boxes. I think this should >>be defined in the spec. In particular, in the case of regions, the >>content of a named flow does not necessarily have a common parent. We >>can have elemA and elemB moved to a named flow and they do not share a >>common parent. When laying them out, they will generate boxA and boxB. >>In that context, I think we should clarify that they are still sibling >>boxes for the purpose of layout and say that the block flow direction >>that is used is that of the first region (see section 4.1 of the regions >>draft, http://dev.w3.org/csswg/css3-regions/#the-flow-into-property, >>note on writing mode). > >We tried to work around this by using "containing block" instead of >"parent". >If that's not sufficient, then perhaps Regions needs to define that the >contents of a flow are wrapped in an anonymous box (and define what >properties >that box has). Various other aspects of layout are not well-defined if >this is >not done; it's not an issue specific to fragmentation. > >Wrt defining whether the boxes in a region flow are siblings: I believe >that >is certainly the job of Regions and not of fragmentation. > >~fantasai The change to "containing block" should be sufficient, as we do define the containing block for the named flow. I agree that CSS Regions should clarify that elements moved to a named flow are siblings. We mention this in two special cases, but I'll add text for the general case as well. Thanks, Alan
Received on Wednesday, 18 July 2012 22:18:34 UTC