A bit more on the TeX part of this argument. Over a decade ago my company polled publishers that accept submissions from authors of content containing math. Although not a scientific poll, the results were overwhelming. Approximately 85% of all submissions were in MS Word format with equations written using its Equation Editor (my company's product at the time, licensed to Microsoft).
It has been pointed out already here but let me emphasize that most math is not done by mathematicians. All K12 students must learn math so there's a huge industry to serve them. Half the departments at a typical university use math in their teaching (not just science and engineering but anything with statistics and business, economics). Mathematicians are a tiny fraction of the whole. I guarantee you that virtually all K-12 teachers have never heard of TeX. Ok, with that dead horse beaten let me turn to MathML. It was mentioned that MathML lacks open source tools to work with it. I believe one of the main reasons for this is the lack of browser support for MathML. It is ironic that most STEM publishers' internal workflows are based on XML and MathML for math but they can't deliver it to browsers. MathJax changes that of course which is why we now see the overwhelming interest in MathJax and its ability to render MathML. It would be truly ironic if MathJax success killed MathML. MathJax does a truly heroic job of formatting mathematics. It is amazing to me that a JavaScript library can do so well. That said, it is a stand-in for true MathML support. It lacks access to fonts and character metrics as well as layout information. The fact that it works so well is due to the brilliance of Davide Cervone and the rest of the MathJax team and not because it is the right way to render MathML in a browser. That it is not truly integrated into the browser results in all sorts of struggles. As witness, see all the postings in its forums from people who run into trouble creating dynamic web pages containing math. They get into all sorts of tangles involving DOM changes, rendering, event processing order, etc. Those of us that work in MathML and equation editing constantly run into the misconception that an equation rendering is more like an image than text. I think this is due to the fact that many document processing systems through history have had to handle math as an image because they don't support math directly. Mathematics is really just a fancy text format. Think subscripts and superscripts on steroids. Some chemistry notations fit into this mold as well but music and some of the other things mentioned in this thread are not. There is one easy way to tell if a notation is fancy text. Do books include the notation inline in paragraphs or not. Even with a block (or display) equation it flows with the text. It was stated that Mozilla's two main competitors don't support MathML. While that is literally true, Internet Explorer has had the benefit of MathML display via my company's MathPlayer plug in for years. When we introduced it, Microsoft added some APIs specifically to allow us to do a better job integrating its rendering into the surrounding text. Most screen readers interface with MathPlayer to make math accessible to people with various disabilities. Please don't turn your backs on MathML just as it is coming into its own. MathJax is a good stand-in for missing MathML support but it is does not eliminate the need for native support. Instead, it makes its absence all the more glaring and the need to fix the problem all the more urgent. _______________________________________________ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
