Re: leading/trailing spaces of operators

On 15/10/2012 13:48, Frédéric WANG wrote:
> I just read David Carlisle's private answer


in case anyone is wondering I'll make that private answer public, but
note this is a personal response not a definite working group position:-)

-----------------------------------------------------------------
>
> moving to internal list but CC'ed to Frédéric for information.
> someone should post to public list if it's clear what the answer
> is:-)
>
> David
>
>
> possible answers inline
>
>
>>
>> -------- Original Message -------- Subject: leading/trailing
>> spaces of operators Resent-Date: Mon, 15 Oct 2012 10:39:36 +0000
>> Resent-From: www-math@w3.org Date: Mon, 15 Oct 2012 12:41:07 +0200
>> From: Frédéric WANG <fred.wang@free.fr> To: www-math@w3.org
>> <www-math@w3.org>
>>
>> Hi,
>>
>> One contributor to the Mozilla project is working on some
>> improvements regarding the parsing of length values. We have two
>> questions about leading/trailing spaces in <mo>:
>>
>> - Can they be negative?
>
> yes?
>
>> This is allowed by the RelaxNG schema and the spec does not seem to
>> say anything against it. However the section 3.1.8.2 contains
>> "MathML elements that permit 'negative spacing', namely mspace,
>> mpadded, and mtext" which would suggest that mo is excluded
>> (although I'm not sure why mtext is there).
>
> I think "mtext" should be replaced by "mo" there
>
>>
>> - I think I already asked it before, but how relative values
>> should be interpreted? Or more precisely, what is the default value
>> that should be used as a reference? I think it is the one set by
>> the dictionary but I just want to be sure. For example:
>
> well....
>
> 2.1.5.2 Length Valued Attributes says the reference value is the
> default value as specified in the table of attributes, and for rspace
> on mo that is set by dictionary or thickmathspace as default
> default.
>
> so....
>
> <mstyle rspace="110%"><mo rspace="10px">+</mo></mstyle>
>
> here I think the value is 10px as the explicit attribute overrides
> the value inherited from mstyle.
>
>
>>
>> <mstyle rspace="110%"> <mi>a</mi> <msub> <mo rspace="10px">+</mo>
>> <mi>A</mi> </msup> <mi>b</mi> </mstyle>
>
> here msub (which doesn't directly take an rspace attribute) is only
> being considered at all by way of it being an embellished operator in
> which case it is supposed to take its rendering attributes from the
> contained mo.
>
> So I _think_ it should take the effective value of 10px, the 110%
> relative value not being used.
>
>>
>> As I read the spec, the spacing around the <msub> should be 110%
>> of the default value "thickmathspace", not 110% or the current
>> value "10px".
>
>
> David



________________________________________________________________________
The Numerical Algorithms Group Ltd is a company registered in England
and Wales with company number 1249803. The registered office is:
Wilkinson House, Jordan Hill Road, Oxford OX2 8DR, United Kingdom.

This e-mail has been scanned for all viruses by Star. The service is
powered by MessageLabs. 
________________________________________________________________________

Received on Monday, 15 October 2012 12:57:16 UTC