- From: Ric Wright <rkwright@geofx.com>
- Date: Wed, 06 Sep 2017 13:55:16 -0500
- To: W3C Publishing Business Group <public-publishingbg@w3.org>
- Message-ID: <D5D5A7F5.5D2BA1%rkwright@geofx.com>
(This question is a little self-serving (well, maybe a lot self-serving :-) but just trying to leverage the deep experience and knowledge of this group. Please feel free to ignore this post). I know itıs early days and it could be construed as more of a EPUB-specific issue, but Readium is hard at work (and making good progress) on Readium-2 which is envisioned as ultimately being a successor to the current Readium-1. One aspect we have been debating concerns support for MathML <https://www.w3.org/Math/> . Do note that we are talking here of the rendering of MathML, not the semantic markup. In Readium-1, we use MathJax <https://www.mathjax.org/> to provide support for MathML by injecting SVG. We do this no matter which browser engine is being used and ignoring whether MathML is even present in the EPUB. Both decisions arose largely from expediency. We use the SVG output form from MathJax which is all paths, so completely inaccessible, but works very well. Now, with Readium-2 (R2), we are considering two options: * Only inject MathJax IFF the EPUB/WP has MathML (EPUB requires it be declared in the manifest) and the browser is considered poor at rendering MathML (which at present is all but Safari). * Drop support for MathML entirely. There are arguments to be made on both sides. MathML is required by the EPUB spec and R1 supports it so why go backwards? OTOH, how much is MathML really used? Plus, given the architecture of R2 it would not be too difficult to simply provide implementers the instructions of how to inject the necessary support so it could be an optional feature. This discussion is of course primarily a reading-system issue, but what we would very much appreciate feedback from those willing to provide it is: * Given that the MathML spec seems to be, well, a bit stale (3.5 years since last update) what is the long-term view of its viability? * Given the difficulty of making it accessible, does it make sense to try and provide markup that is accessible or just throw in the towel and provide bitmaps or use a tool like MathJax on the server (or production time) and generate SVG or HTML/CSS then? * What is current publishing views on these questions? Are the major publishers and/or journals actually using MathML? It is not easy to find examples of the use of MathML (and most are very old) which suggests that it is not widely used. Thoughts? Thanks in advance for any feedback. Ric
Received on Wednesday, 6 September 2017 18:56:52 UTC