Frames, was Re: Lower priority of some checkpoints

I disagree. Frames are a spatial metaphor, and make extremely good sense
in a visual setting. however in a voice setting they appear to be a
slightly artificial construct, and in small devices such as PDAs and
magnified screens they will not necessarily make sense. Although they can
be interpreteed by User Agents (and should be) this is only sometimes the
case. (I will argue the legacy solutions issue in a seperate email.) Where
it is not the case, people who rely on such User Agents cannot get access
to the content.
Ergo, I think it is priority one.

Charles McCathieNevile

On Wed, 6 Jan 1999, Daniel Dardailler wrote:
  Looking at the checkpoint list in
  I advise we lower the priority (from P1 to P2) of 
    For frames, provide a fallback page for pages that contain frames
    (e.g., by using NOFRAMES in HTML at the end of each frameset). 
  That's on the basis that these can simply be (e.g. lynx does it)
  handled by the User Agent.
  I know, some UAs will not handle them and so some combinations of
  these UA and AT will create inaccessible problems, but more or less
  everything theorically fall in that area (i.e. a UA has to comprehend
  it for the user to digest it, the ALT text for instance) and so this
  is a judgment call on the face of what's happening in the field (which
  is why I mentioned lynx), and this is my position.
    For frames, ensure that the source of each frame is a markup file,
    such as HTML. 
  Isn't it why we added a longdesc attribute to FRAME in HTML4, so that
  one can actually do that and be accessible ?
  Maybe it should say "unless you provide a longdesc to what this non
  markup file is about".

Received on Wednesday, 6 January 1999 11:18:36 UTC