Re: Integration of HTM

Maciej, you are making an assumption here. It may not be that the author
wishes to apply a style sheet through a UI that allows the users to walk
through all the alternatives and make a selection. We are having to do that
today in some IBM products for complex visualizations and it unnecessarily
complicates the UI for people that don't need some of the UI alternatives.
A third party may apply their own style sheet on any web page to choose the
content. It may also be a third party AT. Rather than populate a UI with a
variety of options the user has to choose from, we simply annotate the
markup with the resource meta data and trigger its rendering using a
generic style sheet.

If you don't annotate your content with pre-defined meta data your UI
author would need to provide a user interface of options that the user has
to do a manual selection of. There are set of standards in IMS Global
Learning Consortium that define resource meta data applicable to each
modality type.

Could we be doing what you are suggesting using CSS bound to non standard
resource meta data on a one off basis - and leave the selection up to the
UI designer? No disagreement, but this is an unnecessary burden on the
author and UI designer.


Rich Schwerdtfeger
Distinguished Engineer, SWG Accessibility Architect/Strategist

             Maciej Stachowiak                                             
             Sent by:                                                   To 
             public-html-reque         Richard                             
                                       Ian Hickson <>,         
             02/04/2010 10:51          ""                
             AM                        <>                
                                       Re: Integration of HTM              

On Feb 3, 2010, at 5:32 PM, Richard Schwerdtfeger wrote:

      It will be done using CSS but the group wants to bind it to
      pre-defined attributes in the DOM and not IDs as ID values can not be
      guaranteed. If a user wishes to bind a media query to an arbitrary
      document which ATs often do it is necessary to have pre-defined
      attribute names.

The media query style rules to hide or show particular parts of the DOM.
They are intended to be written by the content author and would be put in
the author stylesheet. They do not "bind to IDs", they bind using selectors
to whatever the author chooses. I feel like we are talking past each other
on this and have different mental models of how CSS works. You can use CSS
selectors to address any element in the DOM. It is not necessary to choose
a specific way to do so.

      The author simply then needs to annotate the meta data in HTML using
      these standard attributes and trigger what is displayed through a
      media query.

I don't see the need for the attributes. They would also be strictly less
powerful than media queries, which can use logical operators to apply rules
for multiple media query properties. If anything, I would say we should
have a visible-if-media="" global attribute which takes a media query as a
value. But event that strikes me as unnecessary.


      Rich Schwerdtfeger
      Distinguished Engineer, SWG Accessibility Architect/Strategist

      <graycol.gif>Ian Hickson ---02/03/2010 05:45:41 PM---On Wed, 3 Feb
      2010, Richard Schwerdtfeger wrote:
                         Hickson <                                         
                         .ch>      <ecblank.gif>                           
                         02/03/201                    Richard              
                         0 05:45                      Schwerdtfeger/Austin 
                         PM                           /IBM@IBMUS           
                                                      Re: Integration of   

      On Wed, 3 Feb 2010, Richard Schwerdtfeger wrote:
      > As we discussed, the use of media query of alternative content I am

      > working to pull over a standard set of attributes from the IMS
      > For All specification. It was suggested that we preface these with
      > aria-, however these are not part of the aria specification and
      > preceding these with an aria- dash would not give credit to the IMS

      > Access For All effort.
      > [...] For example, one attribute would be AdaptationType and we
      > define an equivalent CSS Media query property for it.

      I don't think this media-specific approach is a good approach for
      HTML. I
      think media-specific features should be done in CSS, so that an HTML
      document can remain device-independent.

      Ian Hickson               U+1047E                )\._.,--....,'``.
      fL       U+263A                /,   _.. \
      _\  ;`._ ,.
      Things that are impossible just take longer.

(See attached file: pic16439.gif)

Received on Thursday, 4 February 2010 18:18:02 UTC