Re: Comments on MathML Last Call Draft

"What a firestorm! ... If I am missing anything major, maybe someone could 
post that."

Here is more (fire). Hope it helps.

Pankaj Kamthan

1. MathML Specification refers to both CSS1 (for example, and CSS2 
(for example, Shouldn't the reference be to CSS2  (which 
supersedes CSS1) throughout the MathML Specification?

2. Suggestion to change the name of definitionURL to definitionURI. Rationale: 
(1)  URIs are in the works to, among other things, extend the mechanism of 
URLs (which lack persistence). (2) It will include use of URLs, as well as, 
URNs (persistent). (3) XML Namespaces Recommendation uses URIs. 


In the foregoing, I'm using:
[prose with problems] -> [prose with suggested corrections]

Chapter 1

Extensible Style Language (XSL) -> Extensible Stylesheet Language (XSL) 

Chapter 2 

2.2.1 Presentation Examples, 2.2.2 Content Examples but 2.2.3 Mixed 
Markup Examples 

is explcitly ... -> is explicitly ...

references for most synbols is ... -> references for most symbols is
on a scale of x00 through xFF, ... -> on a scale of x00 through xff, ... 
(as #rrggbb and #5599dd use lowercase)
The phrase "explained in CSS subsection, below" at two places could 
provide a link to, and then one would not need to include ", below".
Several MathML attributes, ... defined by Cascading Style Sheets, Level 1 
Expanded form for CSS1 is not needed here. (The same for 2.3.5 and It should be placed at the first appearance with a link to the 
W3C Recommendation, and the acronym could be used thereafter.
This point of view is carried through to the MathML schema -> This point 
of view is carried through to the MathML schema.
... for all MathML attributqes, including ... -> ... for all MathML 
attributes, including ...

Chapter 3

"MathML expressions are often embedded in a textual data format such as 
HTML, ..."
May be XHTML 1.0 instead of HTML.
is intended to facilitate XSL processing -> is intended to facilitate 
XSLT processing

Chapter 4
definitionURL="www.vnbooks.com/VectorCalculus.htm" -> 

definitionURL="www.mathslib.com/vectorfuncs/plus.htm" -> 
definitionURL="www.vnbooks.com/ContDiffFuncs.htm" -> 
The expression
without the <fn>, as it is deprecated.

definitionURL="www.naturalnums.htm" -> 

definitionURL="www.primes.htm" -> 
definitionURL="www.openmath.org/cds/BesselFunctions.ocd"> -> 

definitionURL="www.uni.edu/universalconstants/Boltzmann.htm"> -> 
The default rendering (the image) of the Example in should be 
interchanged with that of the Example in
The default rendering is incomplete (it does not have any equality or 
"sine" term thereafter).
The default rendering is incomplete (it does not have any equality or 
"cosine" term thereafter). 

The "dot" between A and B in the default rendering should be in the 
The "dot" between A and B in the default rendering should be in the 
"ln" should be upright (and not slanted). Probably, the same in 

Chapter 5 

5.4 and other sections mention XSLT but there is no reference to what the 
acronym stands for (XSL Transformations) or what it means (reference to 
the XSLT Recommendation). 

In general, the interaction of content an presentation ... -> In general, 
the interaction of content and presentation ... 

It should be mentioned that "mx" is a non-MathML namespace prefix (and 
that a-factorial, rank, etc. are from that namespace).

Chapter 7

"Indeed, XHTML already renders properly in HTML 4 user agents."

Yes, but only when the Appendix C. HTML Compatibility Guidelines are 

"This is primarily useful when referencing separate files containing MathML 
markup from an EMBED or OBJECT element."
EMBED element is not part of standard HTML 4.01. 

Appendix A

Normally. however, a MathML ... -> 
Normally, however, a MathML ...

... treat the whitespace whitespace characters ... -> 
... treat the whitespace characters ... 

MathML2 DTD SYSTEM identifier is given to be SYSTEM "mathml2.dtd".
Is that canonical?

A.1 mentions "A zip file of the full DTD including entity declarations 
is provided for reference." It is unclear where, as the downloadable 
version of MathML 2.0 doesn't include it nor does it provide a pointer.

Appendix G

Roger Hunter is not mentioned here but is in the list of authors 
(overview.html) and E.1. 

Appendix H

It is incomplete. It does not mention: The addition of 5.3 Parallel 
Markup. The addition of 5.4 (which is a sort of renaming 5.3 from MathML 
1.01). Reordering of sections in Chapter 6 (6.2 became 6.1.3). Unavailability 
of (and removal of links for) Basic and Extended glyphs.