On 21 October 2010 the W3C made an announcement about an “important standard for making mathematics on the Web more accessible and international, especially for early mathematics education“. The press release described how “MathML 3 is the third version of a standard supported in a wide variety of applications including Web pages, e-books, equation editors, publishing systems, screen readers (that read aloud the information on a page) and braille displays, ink input devices, e-learning and computational software.”
Does it work? In order to investigate I installed the Firemath extension for FireFox and the MathPlayer plugin for Internet Explorer. I then viewed the MathML Browser Test (Presentation Markup) page using FireFox (v 4.0), Chrome, Internet Explorer (v 8) and Opera (v 10.61). The results shown using Internet Explorer version 8 are shown below, with the first and second columns containing an image of how the markup has been rendered in TeXShop and FireFox with STIK Beta Fonts and the third column showing how the markup is rendered in the browser the user is using.
A quick glance at the display on all four browsers shows that the support seems pretty good [Note following a commented I received I have noticed that the page isn't rendered in Chrome) - added 2 November 2010]. However it would take a mathematician to ensure that the renderings of mathematical formula are acceptable.
It should also be noted that MathML 3 is part of HTML5. This means that embedding maths in Web documents should become easier, with direct import from HTML to mathematics software and vice versa.
In order to encourage takeup the W3C Math home page provides links to “A Gentle Introduction to MathML” and “MathML: Presenting and Capturing Mathematics for the Web” tutorials with “The MathML Handbook” available for purchase.
I think this announcement is of interest in the context of institutional planning for migration of document formats to richer and more open environments provided by HTML5 and associated standards such as MathML, CSS 3. etc.
Will we start to see documents containing MathML markup being uploaded to institutional repositories, I wonder? And should this format be preferred to PDFs for scientific papers containing mathematical markup?