- From: Doug Schepers <schepers@w3.org>
- Date: Mon, 24 Aug 2015 17:43:14 -0400
- To: Peter Krautzberger <peter.krautzberger@mathjax.org>, W3C Digital Publishing IG <public-digipub-ig@w3.org>
Hi, folks– MathML was developed a long time ago, in isolation from HTML and CSS. I think we could really benefit from taking a look at _why_ browser vendors aren't implementing MathML, learning those lessons, and making a new version of MathML that's easier to implement and maintain, easier to author, easier to style, and is better integrated into HTML5. I've voiced this opinion to Peter before. I don't think we differ much, and I think the years of experience the MathJax team and others have in implementing MathML in modern browsers could yield a lot more benefit than trying to push browser vendors to implement MathML as it is today. Regards– –Doug On 8/24/15 3:47 PM, Peter Krautzberger wrote: > Hi, > > This is a bit longer. Sorry about that. > > 1. Ivan suggested I should sum up what I've been involved in. > > * In 2013, when Chrome dropped WebKit's MathML code, we coordinated with > our MathJax sponsors to reach out to Google management (non-publicly) to > encourage the Chrome team to reconsider. > * in 2013/14, MathJax looked into working on Gecko and WebKit directly; > we could not find funding and we could not get WebKit companies to even > have a discussion about long-term code ownership (let alone architecture > or any other guidance). I'm aware of several efforts to find funding > since then. > * in 2015, the MathWG reached out to browser vendors to start a > discussion of the overall situation; we put this on hold for lack of > availability from the browser end > > 2. Regarding the discussion so far > > * re Paul Belfanti: "publishing at volume". > > I think from a browser vendors perspective, there will never be enough > volume. > > * re Bill Kasdorf > +1 to all points. But also: publisher investments in MathML have been > mostly internal. Not many have invested in polyfills, web development > tools, web standards development, or browser development. Similarly for > third party vendors. > > * re Paul Topping. > > > I used all my votes for MathML but I am sure it didn't get many votes > compared to other features. > > Actually, some features have fewer votes than MathML and get moved > towards implementation. > > > After MathJax, the problem is largely solved, at least from their > perspective. > > This does not match my experience. Frequently, web developers are using > MathJax's shortcomings as an excuse to dismiss MathML. > > * re Deborah Kaplan > > > I want to stress that the push for native MathML is not out of any > desire to get rid of MathJax. > > That's ok ;-) the MathJax project was designed to eliminate itself by > encouraging browser support. But we'd also be happy to enable what comes > after MathML 3. > > > It is out of the need for publishers to be able to distribute native > math that can be read by screen readers without users having to do > anything special. > > This strikes me as orthogonal to native MathML rendering. Both exposing > MathML to a11y interfaces and finding MathML-enabled AT are quite > different (but equally messy) problems. > > Simply reading out MathML via MathML-enabled AT is theoretically easy; > try the example at https://github.com/mathjax/MathJax-docs/issues/111. > > The much bigger problem is AT for visual users, e.g., sync-highlighting > and exploration (see comment on ARIA below). > > > Rather, the greater concern is non-browser-based reading systems, > frequently used for textbooks, which have legal requirements for > accessibility. > > How do you see browser implementations for MathML help here? > > * re Liam > > re (1) you make it sound easy ;-) > > re (3) is actually a huge, messy, overlooked issue; see my use cases on > font metrics APIs for Houdini. > > re (4) I don't really think this is a big issue these days (although > third party vendors are a different issue). TeX conversion has many > options, Microsoft's own equation editor and Windows handwriting > recognition can produce MathML, MyScript has great apps. Although > high-quality MathML Editors are virtually non-existent (esp. if your > focus is publishing on the web) -- that probably means something? > > re (5) the examples are not quite right, I think, (long division etc is > actually pretty easy in MathML 3 markup). But otherwise, yes. > > re (a) math content becomes quickly tabular and there's not a lot room > for reflow in tables. But here's something fun > https://github.com/mathjax/MathJax-RespEq > > re (c) the state of MathML accessibility is pretty messy, on both the > content, browser, OS, and AT end. > > > 3. Let me add a couple of entirely personal points. > > I think the main problem is that nobody really knows what it would > technically require to implement good math layout in browser engines or > what it would take to implement good MathML support. > > * Browser vendors have never worked on MathML. > * MathML in Gecko and WebKit has been almost exclusively implemented > by unpaid volunteers > * Only Mozilla devs has actually supported its volunteer; they are > the only knowledgeable browser devs regarding MathML. > > * In my experience, both browser vendors, publisher, and third party > vendors have surprisingly little knowledge of > * MathML as a markup language > * math layout in general > * MathML layout specifically > * math layout using OWP tech > * math accessibility using OWP tech > > * third party vendors have a bit more knowledge of the first two but I > sometimes see an incentive to keep things complicated (read: expensive). > > * The ARIA spec has a massive hole where mathematics should be. > > * Neither browsers nor publishers are active in the MathWG (ok, maybe 1 > exception). > > * The MathML spec needs some work in an HTML5 context > * MathML has been "out of sync" (for lack of a better term) with > HTML/CSS/SVG for a while > * MathML duplicates HTML/CSS features (sometimes the features are not > compatible though never critically so imho) > * MathML hides quite a few complex layout features behind math > syntax, complicating the implementation of both > * ContentMathML has failed outside of CAS. > > Another key problem is: MathML is neither necessary nor sufficient for > math layout on the web. Nowadays, HTML/CSS implementations are good > enough for (high quality) math/MathML layout. I'm not speaking of > client-side JS-driven rendering but of (server-side) HTML+CSS generation > that looks the same on all recent browsers. (MathJax will introduce such > a technology in our next release. It's not pretty markup code and it's > not perfect but it works.) > > Personally, I don't think MathML will be implemented in browsers (though > I'd be extremely happy to be wrong here and will support any effort in > this group). I do think there are more realistic alternative goals such > as improving CSS implementations incrementally and developing standards > for exposing underlying data such as MathML to tools such as AT. > > Again, sorry about the length. > Peter. > > On Mon, Aug 24, 2015 at 7:29 PM, Deborah Kaplan > <dkaplan@safaribooksonline.com <mailto:dkaplan@safaribooksonline.com>> > wrote: > > On Mon, 24 Aug 2015, Paul Topping wrote: > > Convincing them to implement EPUB would be a worthwhile cause. > > > Yes, absolutely. > > Isn't the bigger problem convincing students and teachers to > adopt etextbooks? > > I'm sure many of you here know the actual data, and anecdotes are > not data. But I do know that when I was a graduate student two years > ago, I had two choices: print, or the official electronic textbook > through our vendor, which was an epub which could only be read > through the vendor's reading system. > > At that institution, in any case, the students and the teachers > aren't making the decisions. If you buy a textbook, you get the > electronic version as is delivered, which is usually EPUB. > > But again, I'm sure many people on this list have the actual data. > > (There was non-terrible accessibility on the reading system (since > it is an educational platform with compliance requirements), so it > was minimally usable without a mouse. But the textbooks were > unedited OCR, so I wouldn't have wanted to use them with a screen > reader. > > That's a content problem, though. Unedited OCR isn't going to end up > with MathML markup no matter what the reader supports.) > > Deborah > >
Received on Monday, 24 August 2015 21:43:18 UTC