[css3-sizing] comments on 'repudiate-floats' value

http://dev.w3.org/csswg/css3-sizing/#width-height-keywords says:

  # ‘repudiate-floats’ ISSUE 2: Less stupid name?
  #     In the inline axis, use the larger of the min-content
  #     measure, and the fill-available measure minus the size of
  #     any non-descendant floats in the same formatting context.
  #     (This is similar to the behavior of a box that establishes a
  #     formatting context next to a float.) In the block axis, this
  #     is identical to ‘auto’.
  # 
  #     ISSUE 3: Does this value work? Is it needed? 

I see a few issues with this (beyond the name):

 1. The behavior of boxes that establish a BFC is different in that
    they only reduce their width to accomodate floats that are
    *horizontally adjacent* to the block, not all floats within the
    containing BFC.  This requires multiple iterations of layout:
    determine the height, see if there are floats adjacent to that
    height that reduce the width, and if so, repeat.  This means
    that, in essence, this value is not just a computed value but
    almost essentially a used value.

 2. This value is pretty much useless unless there's simultaneous
    behavior to shift the block to the side so it doesn't intersect
    the floats.

Because of these (though primarily the second), I think we'd be
better off with an explicit way to make something establish a block
formatting context, and we should remove this value.

-David

-- 
𝄞   L. David Baron                         http://dbaron.org/   𝄂
𝄢   Mozilla                           http://www.mozilla.org/   𝄂

Received on Wednesday, 6 February 2013 19:40:31 UTC