W3C home > Mailing lists > Public > www-math@w3.org > February 2013

Re: Help get math turned back on in Chrome

From: Peter Murray-Rust <pm286@cam.ac.uk>
Date: Sat, 16 Feb 2013 23:20:59 +0000
Message-ID: <CAD2k14MwuNh6kWEjfkP0aqpw5b5oVU7xaEFhwcO49aQx1VYL-w@mail.gmail.com>
To: www-math@w3.org
[Please excuse any ignorance of the deployment of MathML tools and the
operation of the community].

First to state that I and friends are passionately committed to MathML.
I've followed it for 20 years (and had the privilege of being asked to
speak to the MKM folks) and now feel that Content-MathML is fit for my
purposes. I have been developing Chemical Markup Language (CML) for 20
years and hope that experience from that may be useful. I am also keen to
see domain semantics (CML) integrated with MathML (i.e. the MathML
recognizes CML semantics and can action it). We have built a prototype
that, soon, I'd like to announce to this list and get feedback.

I have always looked with great envy on the MathML community. You have an
active public respected lead/spec from W3C and you have many organizations
including companies which provide high-quality implementations, including
Microsoft Office. By contrast in chemistry we have active indifference (and
sometimes antagonism) from chemical software software companies. We do not
expect any support from Google, Microsoft, Mozilla or any non-domain
companies. [I mention with gratitude significant funding from MSResearch to
develop a prototype Chemistry add-in for Word but this was never going to
be a mainstream product.].

As a result the CML community has to build everything itself, bottom-up,
F/OSS. And we are doing it and we are gradually starting to become a de
facto approach. We have to live with whatever the state pf the browsers is.

My questions are:
* is there a sufficiently good F/OSS plugin or plugins for MathML?
* is it technically possible to deploy it/them in browsers?
* are these deployments browser-independent?

If the answer to these are yes, then why should we not deploy plugins
ourselves?

Again, pardon any simplicity ...

The browser is given a document, either XHTML/HTML5, SVG or MathML with
proper namespacing of the MathML. The task for a browser is to:

   1. * recognize the MathML by namespace (possibly as descendants)
   2. * render it virtually (ignoring non-MathML descendants)
   3. * negotiate the size  of the target rectangle (float) in the
   displayed XHTML.
   4. * use Canvas or SVG to render within the target document


I appreciate this is non-trivial, but it's not impossible. We will be
exploring this for chemistry (probably starting with SVG). So long as we
are namespace-controlled then steps 2,3,4 should be common to both our
efforts. I would have thought it was a  reasonable project for a bright MSc
student (e.g. in the MKM community).

Is there a F/OSS MathML hacker community? Because it would be a very
powerful, rapid, way of tackling the problem. I can imagine that an Open
solution running as a plugin in a closed browser creates some problems, and
I may have underestimated them.

And if there is a fundamental flaw in the above I really need to know!

-- 
Peter Murray-Rust
Reader in Molecular Informatics
Unilever Centre, Dep. Of Chemistry
University of Cambridge
CB2 1EW, UK
+44-1223-763069
Received on Saturday, 16 February 2013 23:21:27 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 16 February 2013 23:21:27 GMT