W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > April to June 2000

Re: defining CLASSes

From: Charles McCathieNevile <charles@w3.org>
Date: Thu, 6 Apr 2000 00:21:54 -0400 (EDT)
To: Al Gilman <asgilman@iamdigex.net>
cc: WAI GL <w3c-wai-gl@w3.org>, WAI UA group <w3c-wai-ua@w3.org>, Dan Brickley <danbri@w3.org>
Message-ID: <Pine.LNX.4.20.0004060018160.24622-100000@tux.w3.org>
Alternatively, if you include a link rel="metadata" that explains where to
find out what the URI values you have used are, and you have XHTML (which
half-decent HTML can be automagically converted into) and you have a smart
parser that understands all this stuff, then you could use the linked
metadata to make assertions about things of class
http://somewhere.com/classes/etc...

After all, HTML is a whole lot of special case processing already. If it is
real XML then you can play smarter games with dumber parsers. But the result
is that you could use known or discoverable semantics to describe the role of
your nav elements in a machine-readable way.

cheers

chaals

On Wed, 5 Apr 2000, Al Gilman wrote:

  At 02:52 PM 2000-04-05 -0400, Charles McCathieNevile wrote:
  >To make it available to humans, use the title attribute. To do some kind of
  >semantic processing on it, use a class attribute.
  >
  >And interesting idea is to use a URI as the value of the class attribute -
  >then if that points to some data that has expression in RDF we can make use
  >of it in all kinds of helpful ways.
  >
  
  For what it's worth, the apparently best way to do this and support
  graceful migration to XML-based dialects is to have a dictionary of class
  definitions and to use an RDF schema referenced from the PROFILE parameter
  in the HTML HEAD section to say "for this document, META and CLASS literals
  follow the following usage [more RDF stuff...]"  In this gloss or data
  dictionary the standard [e.g. Chicago or PMLA Style manual] dictionary of
  classes would be cited.  Then the same external definitions can be cited in
  a parallel gloss incorporated in an XML document vernacular definition
  (a.k.a. module or profile) and the base semantics can be expressed in XML
  element type names and HTML class tokens and these don't even have to
  textually match where there are conflicts, but the intent of the
  classification will be crystal clear for a tool (such as a Braille
  transcription service at a fixed-base operation) which has the time and
  resources to trace through multiple layers of references.
  
  You don't want to expect HTML implementations to follow URI-references
  unless they are in attributes already typed to %URI.
  
  On the other hand, I find it a tolerable invasion of HTML spec 'space' to
  extend the purview of the cited PROFILE(s) to govern the interpretation of
  CLASS literals as well as META NAME literals.
  
  [And this works with HTML 3.2 ...<evil grin>]
  
  I don't think we are in a position to demand that User Agents dereference
  and apply a schema cited in html:head.profile but by the same token, it is
  the natural place [other than an new LINK REL] to put it, and it is
  available for "friendly applications that want to try harder" to use.  They
  just have to be reasonably mnemonic if interpreted at face value without
  the precise definitions in the profile definition.
  
  Here I go into broken record mode.  
  
  When the general Web community is consensed on how to cite and when to
  apply schemas in the interpretation of Web pages, we can then see what we
  can get as far as schema application and who we need it from and so how far
  it can be extended into legacy formats.  If a browser or servlet is already
  set up to apply any URI-referenced schema, then it is probably a piece of
  cake to put in its business rules to check html:head.profile or
  html:link(rel=class-definitions) for such a schema.
  
  Al
  
  
  
  >Charles McCN
  >
  >On Wed, 5 Apr 2000, Jon Gunderson wrote:
  >
  >  >IJ: I don't know that you need title. The lack of a name attribute
  >  >means that no image can use this map as a graphical image map. So
  >  >it's purpose could only be to contain some other navigation mechanism.
  >  
  >  Resources may have more than one group of navigation links.  For example a
  >  site navigation bar (between URIs) and an internal table of contents (with
  >  a URI) it would be nice to have a mechanism to distiguish between the two
  >  groups of navigation links.
  >  
  >  Jon
  >  
  >  Jon Gunderson, Ph.D., ATP
  >  Coordinator of Assistive Communication and Information Technology
  >  Chair, W3C WAI User Agent Working Group
  >  Division of Rehabilitation - Education Services
  >  College of Applied Life Studies
  >  University of Illinois at Urbana/Champaign
  >  1207 S. Oak Street, Champaign, IL  61820
  >  
  >  Voice: (217) 244-5870
  >  Fax: (217) 333-0248
  >  
  >  E-mail: jongund@uiuc.edu
  >  
  >  WWW: http://www.staff.uiuc.edu/~jongund
  >  WWW: http://www.w3.org/wai/ua
  >  
  >  
  >
  >--
  >Charles McCathieNevile    mailto:charles@w3.org    phone: +61 (0) 409 134 136
  >W3C Web Accessibility Initiative                      http://www.w3.org/WAI
  >Location: I-cubed, 110 Victoria Street, Carlton VIC 3053
  >Postal: GPO Box 2476V, Melbourne 3001,  Australia 
  > 
  

--
Charles McCathieNevile    mailto:charles@w3.org    phone: +61 (0) 409 134 136
W3C Web Accessibility Initiative                      http://www.w3.org/WAI
Location: I-cubed, 110 Victoria Street, Carlton VIC 3053
Postal: GPO Box 2476V, Melbourne 3001,  Australia 
Received on Thursday, 6 April 2000 00:22:03 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:47:04 GMT