RE: [Filter Effects][css3-transforms] Using MathML for formulas

While you certainly can take your own copy of MathJax as it is open
source, the MathJax Consortium does maintain a CDN that many of its
users point their pages at. You can refer to a fixed version of MathJax
in order to control when to move to a new version and to test it on your
own content before going live.

 

Paul

 

From: Rik Cabanier [mailto:cabanier@gmail.com] 
Sent: Monday, April 23, 2012 3:56 PM
To: Dirk Schulze
Cc: Paul Topping; public-fx@w3.org
Subject: Re: [Filter Effects][css3-transforms] Using MathML for formulas

 

 

On Mon, Apr 23, 2012 at 2:44 PM, Dirk Schulze <dschulze@adobe.com>
wrote:


On Apr 23, 2012, at 2:29 PM, Paul Topping wrote:

> I'm confused. What script and demo site are you referring to? What
does
> the popup say?

http://www.mathjax.org/demos/ tested on IE 9 without MathPlayer plugin.


>
> MathJax is very configurable, by the way. I understand about your
> reluctance to having a W3C spec require JavaScript but, IMHO, one
can't
> do much on the web these days without JS enabled. How many people
really
> run with JS disabled these days anyway?

I totally agree. But something that applies to general websites doesn't
necessarily apply to specifications. Some specifications are still
written in pure ASCII text, just to provide long term achievement (not
at W3C). And JavaScript seems to be one of the things that belong to
this kind of category that might make it harder to provide long term
achievement. I am not sure if the W3C has any preferences to this topic.
However MathJax might be different, since the content does not
necessarily rely on the JS library.

 

It would be good to know what the W3C policy is.

Obviously, we would have to use our own snapshot of MathJax so we can be
sure that there are no breaking changes.

 

It would make our lives easier and the specs prettier.

 

Rik 

 

	>
	> Paul
	>
	>> -----Original Message-----
	>> From: Dirk Schulze [mailto:dschulze@adobe.com]
	>> Sent: Monday, April 23, 2012 2:23 PM
	>> To: Paul Topping
	>> Cc: Rik Cabanier; public-fx@w3.org
	>> Subject: Re: [Filter Effects][css3-transforms] Using MathML
for
	>> formulas
	>>
	>>
	>> On Apr 23, 2012, at 2:04 PM, Paul Topping wrote:
	>>
	>>> Screen readers work with MathPlayer, the IE plugin that my
company
	>> gives
	>>> away for free. (I could not tell you which ones and which
versions
	>>> work.) If MathPlayer is installed, IE uses it to display
MathML and
	>>> screen readers use it to turn math into text to be spoken.
If the
	>> page
	>>> uses MathJax, and MathJax detects IE+MathPlayer, it defers
to
	>> MathPlayer
	>>> to display the math and screen readers will work fine on the
math.
	>>
	>> I know MathPlayer. I just checked MathJax on IE. The problem
is that
	>> the script suggest that the MathPlayer plugin is needed to
display the
	>> formulas on the demo site, even if that is obviously not the
case! As
	>> much as I like MathPlayer and the output of MathJax, this
behavior
	>> seems not to be applicable for a specification of the W3C. It
is
	>> strange that I didn't get a popup on the main page of the
project on
	>> the first load. So it looks like this popup can be blocked
(without
	>> changes to the code itself)?
	>>
	>> Greetings,
	>> Dirk
	>>
	>>
	>>
	>>>
	>>> Paul
	>>>
	>>>> -----Original Message-----
	>>>> From: Dirk Schulze [mailto:dschulze@adobe.com]
	>>>> Sent: Monday, April 23, 2012 1:54 PM
	>>>> To: Paul Topping
	>>>> Cc: Rik Cabanier; public-fx@w3.org
	>>>> Subject: Re: [Filter Effects][css3-transforms] Using MathML
for
	>>>> formulas
	>>>>
	>>>> I would like to know more about MathJax before I come to a
	>> conclusion
	>>>> for my self. MathJax is a Javascript library. Content in
	>>> specifications
	>>>> should not rely on JS at all in my eyes.
	>>>>
	>>>> But if I understood it correctly, than you can add formulas
in pure
	>>>> MathML and the library would take care about the rendering?
It
	> would
	>>>> still mean that you cannot read the specification on
disabled
	>>>> JavaScript and missing MathML support.
	>>>>
	>>>> Is there a benefit to the idea of Aryeh to use CSS and
check for
	> the
	>>>> MathML namespace? My priority is the accessibility. If we
can use
	>> pure
	>>>> MathML, the specs might be more accessible. The problem
with
	> Aryeh's
	>>>> solution is, that current screenreader don't inspect
elements which
	>>> are
	>>>> not displayed on the screen (e.g display:none). Therefore,
on IE
	> the
	>>>> MathML code wouldn't be accessible for screen reader users
since
	> the
	>>>> code does not get displayed. Would that be different with
MathJax?
	>>>>
	>>>> Greetings,
	>>>> Dirk
	>>>>
	>>>>
	>>>> On Apr 23, 2012, at 1:31 PM, Paul Topping wrote:
	>>>>
	>>>>> It is not as fast as if MathML were supported directly in
the
	>>> browser
	>>>> but JavaScript performance is getting faster all the time.
It also
	>>>> depends a lot on what browser you are talking about and
what
	> device.
	>>> It
	>>>> is very fast in IE with MathPlayer installed as it defers
to
	>>> MathPlayer
	>>>> to draw the MathML. It is much slower on an iPhone, for
example.
	>> Best
	>>>> thing to do is try it on a sample page. There are also
links to
	>> other
	>>>> sites that use MathJax listed on the MathJax site. You
might find
	>> some
	>>>> analogous content.
	>>>>>
	>>>>> Paul
	>>>>>
	>>>>> From: Rik Cabanier [mailto:cabanier@gmail.com]
	>>>>> Sent: Monday, April 23, 2012 1:26 PM
	>>>>> To: Paul Topping
	>>>>> Cc: public-fx@w3.org
	>>>>> Subject: Re: [Filter Effects][css3-transforms] Using
MathML for
	>>>> formulas
	>>>>>
	>>>>> This is great! Thanks for pointing this out.
	>>>>> Do you know if the performance is good? Some spec pages
are quite
	>>>> long and I think mathjax has to process all the text.
	>>>>>
	>>>>> Rik
	>>>>>
	>>>>> On Mon, Apr 23, 2012 at 12:52 PM, Paul Topping
<pault@dessci.com>
	>>>> wrote:
	>>>>> You may be interested in the MathJax project
(www.mathjax.org). It
	>>> is
	>>>> an
	>>>>> open source JavaScript engine for displaying MathML and
LaTeX
	>>>> equations
	>>>>> in all modern web browsers and ebook readers. It
essentially fills
	>>>> the
	>>>>> gaps in browser support of MathML.
	>>>>>
	>>>>> Paul Topping
	>>>>> Design Science, Inc.
	>>>>>
	>>>>>
	>>>
	>