Re: MathML 3.0 far enough along for implementation?

For presentation MathML, we are working on simplfying/unifying some of the
length measurements.  We are reworking the elementary math (mcolumn, mline)
based on attempts at implementing it.  The rest should be stable.

We are very much interested in getting feedback from implementers about
MathML3.  Having two or more implementations is a requirement of the W3C for
moving the spec forward.  You can record how will your implementation does
at the testsuites (http://www.w3.org/Math/testsuite/) by using (
http://devdemo.activemath.org/testsuite/main/) linked off the testsuite
pages.  Those will then get added to the test results page:
http://www.w3.org/Math/testsuite/results/tests.html.

As with any spec before it is approved, it might change.  However, any
change is likely to be small at this point in time.

Good luck,

Neil Soiffer
Senior Scientist
Design Science, Inc.
www.dessci.com
~ Makers of Equation Editor, MathType, MathPlayer and MathFlow ~


On Fri, Jan 23, 2009 at 12:34 AM, Paul Libbrecht <paul@activemath.org>wrote:

> Interesting and motivating,
>
> however i would insist to ask on implementations of which part of MathML
> you intend to go. For "MathML presentation"I think the answer is yes, for
> MathML content, things are getting clearer.
>
> For such things are schemas I'm afraid we're just a bit behind... but it
> may be a matter of a few weeks or less, RelaxNG being the priority thus far.
>
> So, if you accept to make comments on the draft, I think it would be a
> fruitful thing to do for both MathML-3 (your comments, the consideration of
> another implementation) and your project (modernity, maturity, some more
> interoperability dimensions).
>
> hope it helps
>
> paul
>
> Le 22-janv.-09 à 21:45, Scott Hudson a écrit :
>
>  I have a client that is in the midst of implementing a new content model.
>> They are currently including MathML 2.0, but I wonder if the MathML 3.0 spec
>> is far enough along and backward-compatible enough to recommend implementing
>> against this spec? The immediate advantage, is that both schemas would be
>> encoded in RelaxNG.
>>
>
>