Editorial: Cross-reference DOM "child" where appropriate in rendering

[zcorpan]

Fixes https://lists.w3.org/Archives/Public/public-whatwg-archive/2016Sep/0013.html

---

cc @fantasai @bzbarsky

Instances of "child" in the rendering section that I did not cross-reference DOM child:

  <p>For the purposes of the CSS ruby model, runs of children of <code>ruby</code> elements that are
  not <code>rt</code> or <code>rp</code> elements are expected to be wrapped in anonymous boxes
  whose <span>'display'</span> property has the value <span>'ruby-base'</span>. <ref
  spec=CSSRUBY></p>

  <p>For the purposes of the CSS table model, the <code>colgroup</code> element, if it contains no
  <code>col</code> element, is expected to be treated as if it had as many such children as its
  <code data-x="attr-colgroup-span">span</code> attribute <span data-x="rules for parsing non-negative
  integers">specifies</span>.</p>

Since these are talking about CSS model I suppose children here are referring to the CSS concept of children, but I'm not entirely sure.

[domenic]

LGTM with nits. Will let you decide whether to merge now or wait to see how the discussion on the mailing lists turns out.

[domenic]

data-noexport not necessary; we've just been cargo-culting it.

[domenic]

Maybe also mention this in the commit message

[bzbarsky]

The ruby/colgroup parts are simply nonsense, imo. The HTML spec should simply define the default UA styling for those (setting the display) property and say nothing else. I guess you still have to define what span means on colgroup, but that should be restricted to cases when the colgroup is in fact styled as a colgroup, and should explicitly talk about child table-column boxes somewhere, not child elements.

If I style my <colgroup> as display: table-row-group then none of the stuff quoted above should apply, of course.

[zcorpan]

Thanks, we'll track ruby/colgroup in #121 and #1800, respectively.