Re: Contributing to MathML 3 ?

Dear Robert,

Am 16.05.2007 um 16:43 schrieb Robert Miner:
> Do you have a short list of issues for MathML 3 coming out of your
> JEuclid work?

So far only a few issued related directly to the the 2.0 spec, but as  
the 3.0 spec is based largely on it, here they are. I will have to  
look into the 3.0 spec into more detail before I can give reasonable  
feedback.

Mo:
by default, operators have a thickmathspace around them. This is ok,  
other than in situations where they have prescripts (mo is base in an  
mmultiscript) or postscripts (msub, msup, msubsuper, munderover with  
limits moved). In this case, there should be no space between the  
operator and its scripts, as this looks very strange.

Ms:
lquote and rquote are both "strings". However, in the context of  
escaping them it only makes sense for them to be "characters"  instead.

Mstyle/Scriptminsize:
by default is is 8pt. However, there are no other default font sizes.  
Assuming the default size was thought to be 12pt, maybe this should  
be mainfontSize / 3 * 2 instead?

Menclose:
What is the point of menclose notation "radical", as it is equivalent  
to "msqrt" ? I think it should be deprecated.

Mmultiscript:
introduces the new "none" element, which is handled / described  
nowhere else. Why not just use mspace, or an empty mi / mn / ...  
element?

The Testsuite:
Some of the .mml files from the testuite (as downloaded from) http:// 
www.w3.org/Math/testsuite/testsuite.zip are not valid MathML  
documents. Some still contain MathML 1 markup, others are not even  
loading using a standard XML parser.  Some example renderings are  
wrong. But I will have to report on this in a separate mail.

The Java DOM Binding:
downloaded from http://www.w3.org/Math/DOM/mathml2/mathml- 
dom_java.zip does not match the spec. It is apparently from the 1st  
edition of the MathML spec. Also, it lacks documentation. I have  
written an xslt-script to generate "proper" DOM interfaces, it is  
available at:
http://jeuclid.svn.sourceforge.net/svnroot/jeuclid/trunk/jeuclid-core/ 
generate-interfaces/
and I would be very much volunteering to "donate" it to receive a  
better usable standard Java DOM binding.

And, last but not least, a comment on <mu>:
What would be the exact difference of rendering <mu> and <mtext>?

> Robert


Max Berger
e-mail: max@berger.name

--
PGP/GnuPG ID: E81592BC   Print: F489F8759D4132923EC4  
BC7E072AB73AE81592BC
For information about me or my projects please see http:// 
max.berger.name