W3C home > Mailing lists > Public > public-xhtml2@w3.org > May 2008

Re: [XHTML Role] SVG WG LC comments (PR#8038)

From: Mark Birbeck <mark.birbeck@webbackplane.com>
Date: Sat, 17 May 2008 11:43:38 -0700
Message-ID: <ed77aa9f0805171143x1979b486y48ebfc0ce781cd80@mail.gmail.com>
To: "Shane McCarron" <shane@aptest.com>, ed@opera.com
Cc: public-xhtml2@w3.org

Hi Shane/Ed,

>> Whereas XHTML typically has elements to markup text, the text elements in
>>  SVG
>> provide no idea about their semantical meaning. Therefore it could be
>> quite
>> useful to have some role values for 'paragraph', 'section', etc. This is
>> especially useful for the interpretation of documents, if there is no
>>  visual
>> rendering available due to the capabilities either of the user or the
>> user-agent and improves accessibility of non (X)HTML markup languages.
>> Such
>> attributes might help to align prose text in SVG in a better
>> understandable
>> way in such an additional text view.

Something which was part of the early design concepts from @role which
has got a little lost, is that *elements* from any language can
provide a handy source of role values too.

To reconstruct the logic:

A role value is simply a URI, or resource. The reason for this is so
that we get an extensibility hook that puts us into the world of RDF.

Now, some values of @role will need to be invented. This might be
because they simply don't exist, or because we want the values to be
'cross-cutting', and apply to many mark-up languages.

But there are many values that already exist, that are suitable. For
example, XForms has a 'hint' element, that can apply to a form

  <xf:input ref="surname">
    <xf:hint>Please enter your surname or family name</xf:hint>

It's pretty straightforward to take the semantics of 'XForms hint' and
apply it to other situations. For example, an Ajax library could pick
up this hint and do something with it in an (X)HTML document, even
without XForms:

  <input name="surname" />

  <div role="xf:hint">
    Please enter your name

So, my suggestion for paragraph, section, header, etc., in SVG, would
be to use XHTML values as roles:

  <svg:text role="xh:h1">Metadata</svg:text>
  <svg:text role="xh:p">
    Metadata is data about data...which is also data...kind of
    turtles all the way down...

Note that a voice system would be able to work with any mark-up
language, simply by knowing XHTML role values.



Mark Birbeck, webBackplane



webBackplane is a trading name of Backplane Ltd. (company number
05972288, registered office: 2nd Floor, 69/85 Tabernacle Street,
London, EC2A 4RR)
Received on Saturday, 17 May 2008 21:01:10 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:30:30 UTC