# Re: [mathjax-dev] New menclose notation coming in MathML 3, 2nd edition

From: Neil Soiffer <NeilS@dessci.com>
Date: Mon, 30 Sep 2013 13:49:20 -0700
Message-ID: <CAESRWkAmN=AeC+enYfppdBO7NDdZQO+AqgfYnOs73Jpz6g4sNQ@mail.gmail.com>
To: Frédéric WANG <fred.wang@free.fr>

I apologize so much time has elapsed, but the working group took a long
time to come to some consensus.  The two main issues that we needed to
resolve were mirroring (bidi) and naming that can be extended.

It's clear that for the \cancelto notation, the symbol should mirror.
However, the diagonalstrikes don't mirror, and it's not clear that once
extended to all possible (single line) arrows, that all the arrows should
mirror.  Furthermore, if someone is authoring in a RTL language, it is
highly likely the characters they use will come from that language so at
least the cdata of the leaves in the MathML would need to change. Requiring
the author to also change to an menclose attr value is not a big deal if
they want their example to work for an LTR language.  In the end, the group
felt we shouldn't change the spec for how "updiagonalstrike" and
"downdiagonalstrike" behave and that arrow notations should not flip in an
RTL context.

For naming, we considered lots of alternatives and concluded there was some
precedence with unicode naming and with TeX naming that compass points
along with up/down/left/right were used.  So *if* a full set of names were
to be used, they should follow the conventions:

uparrow
rightarrow
downarrow
leftarrow

northwestarrow
southwestarrow
southeastarrow
northeastarrow

updownarrow
leftrightarrow
northwestsoutheastarrow
northeastsouthwestarrow

That means for the cancelto notation, the name should be "northwestarrow".
Given that mathjax and mozilla support a different name, I suggest just
adding this new alternative to accepted names.

The spec mentions the naming convention, but only lists "northwestarrow" as
a recommended name to support.

The draft-spec is at:
http://www.w3.org/Math/draft-spec/chapter3.html#id.3.3.9.2.

Neil Soiffer
SeniorScientist
Design Science, Inc.
www.dessci.com
~ Makers of MathType, MathFlow, MathPlayer, MathDaisy, Equation Editor ~

On Mon, Sep 9, 2013 at 12:27 AM, Frédéric WANG <fred.wang@free.fr> wrote:

>  I am the one who proposed this notation, so perhaps I should say a word
> on this.
>
> First, as I said I've never seen this notation in France so I can't
> comment about its usefulness other than believing you that it is (or was)
> used at least in K12 math in the US. It's probably not fundamental to have
> this notation (even the doc of the cancelto says that striking formulas is
> a poor notation), but at least it was implemented independently in two
> different projects (LaTeXML and MathJax) with essentially the same method:
> menclose updiagonalstrike + a superscript element. So it was important
> enough to need an implementation and the MathML code to choose seems
> somewhat natural since the two developers had the same idea. I don't like
> that menclose notation names don't take the overall direction into account
> but here that seemed irrelevant if, as I believe, it's not used in RTL math
> (the W3C Arabic Math notes doesn't mention it). So I thought it was OK to
> add one menclose notation for this particular arrow, but I definitely don't
> see the need for other striking arrows if there is no use case.
>
> The reason why I proposed this notation is that MathJax used to have a
> non-standard class on the menclose element:
>
>  <menclose notation="updiagonalstrike" class="MJX-arrow">...</menclose>
>
> to tell that the updiagonalstrike should have an arrow head. Hence MathJax
> rendered the notation visually the same as the cancelto command but other
> rendering engines like MathPlayer or Gecko just renders an
> updiagonalstrike. Moreover, other TeX to MathML converter like LaTeXML are
> not aware of MathJax's non-standard CSS classes and can not achieve the
> same rendering via standard MathML (for example Bruce only used an
> updiagonalstrike). Obviously, I don't like these specific hacks in MathJax
> and I hope these nonstandard MJX-* classes will be implemented via standard
> means in the future. In the case of cancelto, the menclose notation is
> open-ended so I thought it would be easy to add a new notation for that.
> Now MathJax and LaTeXML uses
>
> <menclose notation="updiagonalstrike updiagonalarrow">...</menclose>
>
> and this will be supported by Gecko 24 (to be released next week). I'd
> like this to be in MathPlayer too, but even if that's not the case,
> MathPlayer will just render the notation "updiagonalstrike" as it has
> always done, which is understandable...
>
> Le 08/09/2013 20:24, Neil Soiffer a écrit :
>
> The fact that it was one of the most requested features for MathJax is a
> strong argument for adding it to MathML.  I still remain dubious about its
> use though.  My kids have gone through/are going through K12 math and I've
> never seen arrows used as part of crossouts in any of the books or online
> materials they used/we've looked at, either for cancellations or for
> carries/borrows. But that's just anecdotal evidence which is why I was
> hoping that Bruce Miller or Michael Kohlhase could give some input as to
> usage based on their arXiv work since they will see the macro and will have
> needed to implement it (or not) for LaTeX2XML.
>
> You also touched on a hot button topic for me:  I'm strongly against the
> idea that hacking solutions with mpadded or playing games with tables to
> achieve a particular presentation is acceptable. It means the result is not
> accessible and a certain percentage of the population is excluded from
> using the material. I'm not arguing for semantic usages, I'm just arguing
> that the presentation be able to be audibly described to the user in a
> meaningful manner.  E.g,. 'x minus 1' with a horizontal strike across it"
> is understandable and but not semantic.  However, playing games with mover
> and mpadded: "x minus 1 with a line above it that has a negative depth of
> 15 pixels" is gibberish and if the mpadded is not spoken, (... with a line
> above it), the result is very misleading.
>
>     Neil
>
>
>
> On Saturday, September 7, 2013 8:46:36 AM UTC-7, dpvc wrote:
>
>> > The original use for the notation pointed to the TeX package that
>> implemented "cancelto".  But this command has another argument which the
>> arrow points to.  That doesn't fit into the menclose notation.
>> >
>> > As you can see, I'm rather skeptical that adding this is a good idea.
>>  Perhaps Bruce or Michael can look at arXiv files and tell us how many
>> times cancelto is used in those files.  Maybe it is much more common than I
>> think it is.
>>
>> Before the cancel package was added to MathJax, it was one of the most
>> requested extensions that we had.  It is used frequently at the K-12 level,
>> but not so much in higher level mathematics, so I would be surprised if
>> arXiv showed much use of it.  You would need to look at educational
>> material rather than research papers.
>>
>> While I understand your desire for symmetry in providing all possible
>> arrows, I don't see the mathematical need for anything but the lower-left
>> to upper-right arrow, as I am not aware of any other usage.  There
>> certainly may be, but they would be far less prevalent than this.  And to
>> propose a lot of additional notations for which there are no ready use
>> cases, and then use that to suggest that the one notation that DOES have
>> use cases should not be implemented seems a bit backhanded to me.
>>
>> As for those other notations, I would say that vertical and horizontal
>> arrow notations are not needed because there are already means of obtaining
>> those arrows (even stretched to the proper sizes, and they can be placed
>> above or below with mover or munder, to the left and right via
>> juxtaposition, and struck through vertically or horizontally with mpadded),
>> while there is no mechanism for obtaining diagonal arrows at arbitrary
>> sizes.  If you want to say that you need such arrows in all directions, and
>> with one or two arrowheads because they would be useful in commutative
>> diagrams, then I could see that (though it would be awkward to use menclose
>> for that), and this would lead to 6 new notations (or 4 if you just to the
>> heads separately); but suggesting that the need for a cancelation arrow
>> necessitates 23 other arrows seems to me to be taking the desire for
>> symmetry too far.
>>
>> Davide
>
>  --
> You received this message because you are subscribed to the Google Groups
> "MathJax Development" group.
> To unsubscribe from this group and stop receiving emails from it, send an