W3C home > Mailing lists > Public > www-math@w3.org > May 2012

Re: unrecognized actiontype in maction

From: David Carlisle <davidc@nag.co.uk>
Date: Tue, 29 May 2012 12:29:31 +0100
Message-ID: <4FC4B31B.206@nag.co.uk>
To: Frédéric WANG <fred.wang@free.fr>
Cc: www-math@w3.org
On 29/05/2012 10:30, Frédéric WANG wrote:
> On 29/05/2012 11:13, David Carlisle wrote:
>> On 28/05/2012 17:17, Andrii Zui wrote:
>>> Hello everyone!
>>>
>>> While working on my summer project with MathML maction, there
>>> was discovered an uncertainty in the REC. In section 3.7.1.1
>>> <http://www.w3.org/TR/MathML3/chapter3.html#id.3.7.1.1> it says:
>>>
>>> By default, MathML applications that do not recognize the
>>> specified actiontype should render the selected sub-expression as
>>> defined below.
>>>
>>>
>>> and then it goes on to describe the actiontype attribute and the
>>> effect of the selection attribute.
>>>
>>> The question is, how to deal with the selection attribute when we
>>> get an unrecognized actiontype? Should the selection attribute on
>>> maction be considered by default?
>>>
>>> -- Andrii Zui
>>
>> I think that is fairly unambiguous that "the selected
>> sub-expression as defined below" means "the child specified by the
>> selection attribute" so yes, my reading is that the selection
>> attribute should always be used, and if it refers to a missing
>> child then that is an error.
>>
>> David [personal response not checked with WG]
> There are four predefined action types: "toggle", "statusline",
> "tooltip", "input". For the first one, the expected behavior with
> respect to the "selection" attribute is clearly defined. However, for
>  the three others this attribute does not seem to make sense. Firefox
>  always took into account the selection attribute whatever the
> specified action type, but Andrii changed this and now this attribute
> is ignored for all but toggle.

> So the real question is: what to do when the actiontype is not among
> those defined in the spec, or not specified at all? Andrii is
> preparing a patch to do this:

It seems to me that the spec is fairly unambiguous that the child
specified by the selection attribute (or 1st child by default) should be
rendered unless you are handling a "known" actiontype whose defined
behaviour ignores the selection attribute. So if you are asking for
clarification of what the spec says, I think (although I will get WG
confirmation) that is what the spec says. If you are asking for a spec
change via errata that's also possible of course I wouldn't really want
to comment on that before WG discussion.

>
> - If the actiontype is not specified, is is an error: display an
> "invalid markup" message (as Firefox does e.g. when an msup has only
> one child)

That would be an incompatible (but not necessarily impossible) change to
the spec, since it currently says unknown values should be ignored.

> - If the actiontype is unknown, take into account the selection
> attribute (that may help people to implement via javascript custom
> actiontype like "menu" described in MathML 1)

This is I think clearly what the current spec says should happen.
>
> Does that make sense?



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 Tuesday, 29 May 2012 11:30:00 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 29 May 2012 11:30:01 GMT