Re: MathML spec re: Obsoleting some specifications

Hi,

Chaals wrote:

> The implication is effectively that it gets a banner, saying we think you
should be looking at something else if you want to implement MathML.

(Where "something else" refers to MathML 3 as opposed to MathML 1 and 2.)

This sounds like a very good idea.

> If there is some significant level of MathML 2 support in deployed stuff
that people use, that isn't matched by MathML 3 support then it may make
sense to keep both [...]

>From our experience at MathJax, there are extremely few cases of MathML 2
content that is not compatible with MathML 3 (and causes issues). My
unsubstantiated estimate would be that we have one case per year appearing
on our radar. I can't recall any of these being a major problem to resolve.

So marking MathML 1 and 2 as obsolete seems very sensible.

Best,
Peter.
-- 

Peter Krautzberger, MathJax Manager

mathjax.org <http://www.mathjax.org/> | fb.com/mathjax | @mathjax
<http://twitter.com/mathjax>

Support MathJax - Become a Sponsor!
<http://www.mathjax.org/sponsors/mathjax-sponsorship-program/>


On Thu, Apr 27, 2017 at 6:22 AM, <chaals@yandex-team.ru> wrote:

> Dear mathemagicians and others,
>
> this discussion is taking place on the W3C TAG mailing list, and your
> input might be helpful there...
>
> cheers
>
> Chaals
>
> -------- Пересылаемое сообщение  --------
> 27.04.2017, 00:48, "chaals@yandex-team.ru" <chaals@yandex-team.ru>:
>
> Hi David,
>
> 27.04.2017, 00:23, "David Carlisle" <davidc@nag.co.uk>:
> >   > MathML 1 and 2 - there is a version 3
> >
> >  As you note the Math WG is currently in cold storage however you could
> >  flag this on the www-math list.
>
> I'll send a pointer to this discussion onto the www-math list.
>
> [...thanks for explaining in more detail...]
> >  That would leave MathML 2.0 2nd edition and MathML 3.0 2nd edition.
> >
> >  In these versions MathML 2 is with only very minor exceptions a
> >  compatible subset of MathML 3, and so it possibly makes sense to leave
> >  that so any implementations that don't implement the additional features
> >  can claim they support MathML 2 rather than support some MathML 3
> >  subset, but they could probably claim that anyway. I'm not clear to be
> >  honest what are the implications of marking a spec obsolete, is it just
> >  that it gets a banner added in place saying that it is obsolete (and
> >  pointing at the newer spec) or is there more to it?
>
> The implication is effectively that it gets a banner, saying we think you
> should be looking at something else if you want to implement MathML. The
> specs are still formally W3C Recommendations, and the text is of course
> available, so if anyone wants to check their conformance specifically to
> e.g. MathML 1.0 they can do so, and there is a recognition that
> "obsoletion" might happen prematurely, e.g. because for some reason a whole
> new industry develops around MathML 1 - so it is meant to be relatively
> straightforward to reverse in such cases.
>
> If there is some significant level of MathML 2 support in deployed stuff
> that people use, that isn't matched by MathML 3 support then it may make
> sense to keep both of them, in the same way that Martin noted for the XSLT
> example.
>
> cheers, and thanks for the input. I hope we get more from people who
> specifically know MathML…
>
> chaals
>
> --
> Charles McCathie Nevile - standards - Yandex
> chaals@yandex-team.ru - - - Find more at http://yandex.com
> -------- Конец пересылаемого сообщения --------
>
> --
> Charles McCathie Nevile - standards - Yandex
> chaals@yandex-team.ru - - - Find more at http://yandex.com
>
>