Re: Bug in test suite

On 21/05/2010, at 1:08 AM, William F Hammond wrote:

> I'm looking at the commutative diagram in the page
> http://www.w3.org/Math/testsuite/build/main/TortureTests/Complexity/ 
> complex1-full.xhtml
>
> The diagram is effected with an mtable having 3 cells in each of
> 3 rows.

That coding has a few other quirks that I'd like
to ask about.

Firstly
             <mo>,</mo>
             <mtext>&#x205F;&#x0200A;</mtext>
             <mtext>&#x205F;&#x0200A;</mtext>
             <mstyle fontstyle="normal">
               <mrow>
                 <mi>f</mi>
                 <mi>o</mi>
                 <mi>r</mi>
               </mrow>
             </mstyle>
             <mtext>&#x205F;</mtext>
             <mo stretchy="false">|</mo>
             <mi>q</mi>
             <mo stretchy="false">|</mo>
             <mo>&lt;</mo>
             <mn>1</mn>
             <mi>.</mi>
           </mrow>

The mathematical spacing is given here as <mtext> but
the actual word is given as an <mrow> with each letter
presented within a separate <mi> .

Is this what is supposed to happen when translating
   say  \mathrm{for}  instead of  \text{for}  ?


Also there is:

             <mfrac>
               <mn>1</mn>
               <mn>2</mn>
             </mfrac>
             <mtext>&#x2009;</mtext>
             <msubsup>
               <mi>g</mi>
               <mstyle scriptlevel="1">
                 <mrow>
                   <mi>&#x03BC;</mi>
                   <mi>&#x03BD;</mi>
                 </mrow>
               </mstyle>
               <mrow/>
             </msubsup>
             <mtext>&#x2009;</mtext>
             <mstyle fontstyle="normal">
               <mi>R</mi>
             </mstyle>

in which a \thinspace is given as

              <mtext>&#x2009;</mtext>

in a context where surely it is an "implicit multiplication"
kind of space, rather like how  &InvisibleTimes;  might be
presented.
  So why use <mtext> instead of  <mo>&#x2009;</mo> ?

This way of presenting thin-space is used repeatedly throughout
the example. Is it a MathML requirement?
Or just characteristic of the particular producer of this example?
Or a case where one of various alternatives has been used?

>
> It seems that the markup effect of the reference image is intended
> to be achieved using
>           <mspace height="100px" width="150px"/>
> in the middle cell rather than leaving the middle cell empty.
>
> Is this a normative technique?
>
> It works very well in one browser that I'm looking at and is rather
> grotesque in another (on a different platform with different fonts).
>
> It would seem better to be able to specify arrow lengths.

An attached image shows what I see in FireFox on a Mac.
Also there is an image of the Cauchy formula, where the
integral sign has not stretched.

>
> As to what "works" now, the rendering of this diagram will, I think,
> be acceptable on both of these platforms if (1) the middle cell is
> left empty and (2) long horizontal arrows, i.e., U-27F5 and U-27F6 are
> used.
>
>                                     -- Bill

Hope this helps,

	Ross

------------------------------------------------------------------------
Ross Moore                                       ross.moore@mq.edu.au
Mathematics Department                           office: E7A-419
Macquarie University                             tel: +61 (0)2 9850 8955
Sydney, Australia  2109                          fax: +61 (0)2 9850 8114
------------------------------------------------------------------------