- From: Frédéric WANG <fred.wang@free.fr>
- Date: Thu, 23 Feb 2012 21:44:03 +0100
- To: www-math@w3.org
- CC: mathjax-dev@googlegroups.com
On 23/02/2012 21:25, Frédéric WANG wrote: > It seems that some people didn't get my point, so to be more accurate, in > > (Tree1) := > > <mrow> > <mo>|</mo> > <mspace height="100px" depth="100px"/> > </mrow> > > the | stretches to cover the height+depth of the mspace. That's the > vertical stretching rule for <mrow>: > > http://www.w3.org/TR/MathML3/chapter3.html#id.3.2.5.8.2 > > But Tree1 is an mrow whose children is one embellished op <mo>|</mo> > and a space-like element <mspace/>. Thus it is itself an embellished op: > > http://www.w3.org/TR/MathML3/chapter3.html#id.3.2.5.7.3 > > Now in > > (Tree2) := > > <math> > <mrow> > (Tree 1) > <mspace height="50px" depth="50px"/> > </mrow> > </math> > > the same vertical stretching rule applies: the core of the embellished > operator (Tree 1) stretches to cover the height+depth of the mspace. > > Hence there are two possibilities for stretching the <mo>|</mo> > element: one in Tree1 (height+depth=200px) and one in Tree2 > (height+depth=100px). Which one do we choose? > > I'm thinking I have to choose the outermost possibility (100px) or > stretch to the maximum possible size (200px)? > How do other rendering agents implement it? > > So I've just done some experiments: MathJax v2.0 beta: stretch to the maximum size of the mspace's MathPlayer 3.0 beta 1: stretch to the size of the innermost mspace. Firefox 10: stretch to the size of the outermost mspace. (probably like MathPlayer in earlier versions when embellished op support was not completely implemented) -- Frédéric Wang maths-informatique-jeux.com/blog/frederic
Received on Thursday, 23 February 2012 20:43:09 UTC