Re: [html-bidi] Any bidi requirements for MathML (or MathMl in HTML)?

Oops. I missed the following in Bert's message:

> Unlike in HTML, each token is "directionally isolated" (in the terms of
> html-bidi WD), so a BDI attribute is not needed.

That's great! My previous message about what happens when RTL variable names
are used in an LTR formula should be ignored.

To check if there are any less obvious bidi needs note yet covered in
MathML, I will need to read its spec thoroughly.

Aharon

On Wed, Aug 18, 2010 at 10:08 AM, Aharon (Vladimir) Lanin <aharon@google.com
> wrote:

> > Some mathematicians still need to write formulas with Arabic letters,
> e.g. for publications.
>
> What happens when a formula, intended to be displayed left-to-right happens
> to have two RTL variable names around an operator? For example, in HTML,
>
> <div dir=ltr>(A - B) * 3 = C</div>
>
> (where capitals are RTL text, once again), the display will be garbled to:
>
> (C = 3 * (B - A
>
> So, in HTML, ubi (the new name for bdi) would indeed be useful if RTL
> variable names are used in formulas:
>
> <div dir=ltr>(<span ubi>A</span> - <span ubi>B</span>) * 3 = <span
> ubi>C</span></div>
>
> will be displayed correctly as:
>
> (A - B) * 3 = C
>
> Of course, when a formula has all-RTL variable names, the reader may not be
> sure if it is supposed to be read RTL (unless it has inherent directional
> hints like roots or exponents). For this reason I would imagine that Hebrew
> authors will stick with Latin and Greek letters for the variables.
>
> Aharon
>
>
>
> On Wed, Aug 18, 2010 at 5:40 AM, Najib Tounsi <ntounsi@gmail.com> wrote:
>
>> In Morocco, mathematic formulas, at least in school books, and written
>> with Latin letters and thus are LTR, although mathematics are teached  in
>> Arabic. Some mathematicians still need to write formulas with Arabic
>> letters, e.g. for publications.
>>
>> I think Arabic "requirements" in MathML were mentioned in a W3C note:
>> Arabic mathematical notation, W3C Interest Group Note 31 January 2006 (
>> http://www.w3.org/Math/Group/Notes/draft-notes/arabic.xhtml)
>>
>> This Note has allowed clarification of the relationship with bidirectional
>> text.
>>
>> The Note is also mentionned in MathML 3.0 specification, where the section
>>  "3.1.5 Directionality " is devoted to RTL issues in general. (
>> http://www.w3.org/TR/2010/PR-MathML3-20100810/chapter3.html#presm.bidi)
>>
>> This said, Bert Boss may be right:  a BDI attribute should not be  needed.
>>
>> Regards, Najib
>>
>>
>>
>>
>> CE Whitehead wrote:
>>
>>> Hi, Ehsan, Aharon, all:
>>>
>>>
>>>> From: ehsan@mozilla.com
>>>> Date: Tue, 17 Aug 2010 18:06:09 -0400
>>>> To: aharon@google.com
>>>> CC: addison@lab126.com; bert@w3.org; public-i18n-bidi@w3.org
>>>> Subject: Re: [html-bidi] Any bidi requirements for MathML (or MathMl in
>>>> HTML)?
>>>>
>>>> The situation is exactly the same for Persian: math formulas are
>>>> written from left-to-right. I suspect that the same is true for
>>>> Arabic as well, but I'll defer the final confirmation to a native
>>>> speaker.
>>>>
>>>> --
>>>> Ehsan
>>>>
>>>>
>>>
>>> Arabic digits were discussed in a thread at internationalization:
>>> http://lists.w3.org/Archives/Public/www-international/2009AprJun/0081.html
>>>  Apparenty phone numbers are written left to right and numbers have their
>>> most significant digit on the left; not sure that they are always typed left
>>> to right however.
>>>  I should wait to see if Najib has anything to add to this since this was
>>> a 2009 thread.
>>>  Best,
>>>  C. E. Whitehead
>>> cewcathar@hotmail.clom
>>>
>>>
>>>
>>>>
>>>>
>>>> On Tue, Aug 17, 2010 at 4:28 PM, Aharon (Vladimir) Lanin
>>>> wrote:
>>>>
>>>>
>>>>> Traditionally, formulas in Hebrew mathematical publications are always
>>>>> written left-to-right. The Israel Standards Institute is in the midst
>>>>> of
>>>>> considering the question of whether to break with his tradition.
>>>>> Although a
>>>>> final decision has not yet been made, it looks like the answer will
>>>>> basically be no: math will continue to be written left-to-right.
>>>>>
>>>>> For Arabic, on the other hand, the answer may be quite different, but I
>>>>> know
>>>>> little about it.
>>>>>
>>>>> Aharon
>>>>>
>>>>> On Tue, Aug 17, 2010 at 5:52 PM, Phillips, Addison
>>>>> wrote:
>>>>>
>>>>>
>>>>>> Hi Bert,
>>>>>>
>>>>>> Thanks for the note. I have added this to our WG agenda for this week.
>>>>>>
>>>>>> Addison
>>>>>>
>>>>>> Addison Phillips
>>>>>> Globalization Architect (Lab126)
>>>>>> Chair (W3C I18N, IETF IRI WGs)
>>>>>>
>>>>>> Internationalization is not a feature.
>>>>>> It is an architecture.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>> -----Original Message-----
>>>>>>> From: public-i18n-bidi-request@w3.org [mailto:public-i18n-bidi-
>>>>>>> request@w3.org] On Behalf Of Bert Bos
>>>>>>> Sent: Tuesday, August 17, 2010 6:33 AM
>>>>>>> To: public-i18n-bidi@w3.org
>>>>>>> Subject: [html-bidi] Any bidi requirements for MathML (or MathMl in
>>>>>>> HTML)?
>>>>>>>
>>>>>>> This is a somewhat vague question, and also rather late, given that
>>>>>>> MathML3 is as good as finished (it's a Proposed Recommendation),
>>>>>>> but I
>>>>>>> just wondered if anybody has thought about math...
>>>>>>>
>>>>>>> The html-bidi[1] draft talks about requirements on HTML and CSS. It
>>>>>>> doesn't mention MathML, although the current MathML also has bidi
>>>>>>> support. And in the current plans for HTML5, MathML will be an
>>>>>>> integral
>>>>>>> part of the HTML language.
>>>>>>>
>>>>>>> Are there no additional requirements on MathML?
>>>>>>>
>>>>>>>
>>>>>>> The problem of bidi isolation at least doesn't seem to occur in
>>>>>>> MathML:
>>>>>>>
>>>>>>> Section 3.1.5[2] of MathML3 describes the bidi features of MathML.
>>>>>>> It is
>>>>>>> possible to set the overall direction of a formula as well as the
>>>>>>> base
>>>>>>> direction of individual tokens.
>>>>>>>
>>>>>>> Changing the overall direction acts almost like a mirror, e.g., "a
>>>>>>> + b"
>>>>>>> is written as "b + a" and superscripts (a²) are written on the left
>>>>>>> of
>>>>>>> their base (²a).
>>>>>>>
>>>>>>> Tokens are symbols (+, =), numbers (12, 0.5), letters (x, y), words
>>>>>>> (sin, cos), or phrases ("such that," "theorem 1"). Setting a
>>>>>>> direction
>>>>>>> on tokens should rarely be needed, the inherited direction and the
>>>>>>> Unicode bidi algorithm are usually enough.
>>>>>>>
>>>>>>> Unlike in HTML, each token is "directionally isolated" (in the
>>>>>>> terms of
>>>>>>> html-bidi WD), so a BDI attribute is not needed.
>>>>>>>
>>>>>>>
>>>>>>> [1] http://www.w3.org/TR/2010/WD-html-bidi-20100304/
>>>>>>> [2] http://www.w3.org/TR/MathML3/chapter3.html#presm.bidi
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Bert
>>>>>>> --
>>>>>>> Bert Bos ( W 3 C )
>>>>>>> http://www.w3.org/
>>>>>>> http://www.w3.org/people/bos
>>>>>>> W3C/ERCIM
>>>>>>> bert@w3.org 2004 Rt des Lucioles / BP
>>>>>>> 93
>>>>>>> +33 (0)4 92 38 76 92 06902 Sophia Antipolis Cedex,
>>>>>>> France
>>>>>>>
>>>>>>>
>>>>>>
>> --
>> Najib TOUNSI (tounsi at w3.org)
>> W3C Office in Morocco (http://www.w3c.org.ma/)
>> Ecole Mohammadia d'Ingénieurs, BP. 765 Agdal-RABAT Morocco
>> Phone : +212 (0) 537 68 71 50  Fax : +212 (0) 537 77 88 53
>> Mobile: +212 (0) 661 22 00 30
>>
>>
>>
>