RE: Proposal for 1.5 success criteria

User Agent should be in the glossary (I think it is already) and we
should link to it from here.

It would be useful to find or create a short summary of the types of
configuration that a UAAG compliant browser should support.  


-----Original Message-----
From: Gregg Vanderheiden [mailto:GV@TRACE.WISC.EDU] 
Sent: Tuesday, December 18, 2001 1:26 PM
To: GLWAI Guidelines WG (GL - WAI Guidelines WG)
Subject: RE: Proposal for 1.5 success criteria

Can we think of a way to state this in specific terms. 

I am worried that if I were a page author I wouldn't know what a UAAG
compliant user-agent was?   (or even what a user-agent was).

Gregg



-- ------------------------------ 
Gregg C Vanderheiden Ph.D. 
 
> -----Original Message-----
> Behalf Of Cynthia Shelly
> 
> How about this:
> A user can change the presentation to meet his/her needs, by
configuring
> his UAAG-compliant user agent.
> 
> It needs some work on the wording, but you get the idea.
> 
> The tool to test this is a browser.  You open the page in a browser,
> play with the configuration settings, and see if you can change the
> presentation without breaking the page.
> 
> In a separate test case document, we can list things to try, like:
> 
> Test case:
> In Internet Explorer, choose view/text size/largest
> 
> Expected result:
> The fonts get bigger
> The page is still usable
> 
> We could also create a test user-defined stylesheet, and directions on
> how to apply it.
> 
> Test Case:
> Apply the stylesheet at <uri>
> [directions for how to do that]
> 
> expected result:
> a bunch of stuff changes [need to define what, based on the stylesheet
> we supply]
> the page is still usable
> 
> The author would need to go through the test cases, and verify that
the
> results were reasonable.
> 
> Remember, testable doesn't mean machine-testable.
> 
> 
> I don't know of a tool that can test #2, but I think this requirement
is
> specific enough that someone could write one for a given technology.
> 
> I don't think that was true for the old success criteria.  What is
> sufficient markup?  How do I tell if content and presentation are
> separate?  And, what is content, anyway?
> 
> -----Original Message-----
> From: Gregg Vanderheiden [mailto:GV@TRACE.WISC.EDU]
> Sent: Monday, December 17, 2001 11:03 AM
> To: Cynthia Shelly; w3c-wai-gl@w3.org
> Subject: RE: Proposal for 1.5 success criteria
> 
> Boy,
> 
> This just shows how hard this one is.
> I don't think we can have a checkpoint that asks the author to guess
at
> the capabilities of a user.  In this case do we mean
> "a" user (i.e.  any one user) ?
> or
> "all" users (i.e. any user)?
> 
> The first means nothing since they could pick someone who can see,
hear,
> etc.
> 
> The second one asks for a conclusion based on knowledge the person
> doesn't have.
> 
> Can we do this in a way that doesn't require any knowledge of the user
> or his/her needs?  (which will be the case for most authors).
> 
> RE the second criterion
> 
> Do we have a tool that an author could use to test this?   I don't
know
> of many people who could answer this by just looking at a page.
> Especially if they created it with a Visual Authoring Tool and didn't
> know HTML.
> 
> Cynthia --  I think these are progress but do you see the problems I'm
> referring to?
> 
> Gregg
> 
> -- ------------------------------
> Gregg C Vanderheiden Ph.D.
> 
> 
> 
> > -----Original Message-----
> > From: w3c-wai-gl-request@w3.org [mailto:w3c-wai-gl-request@w3.org]
> On
> > Behalf Of Cynthia Shelly
> > Subject: Proposal for 1.5 success criteria
> >
> > Here's my action item from the 6th - reworked success criteria for
1.5
> >
> > You will have successfully separated content and structure from
> > presentation if:
> > 1.	A user can change the presentation to meet his/her needs, for
> > example by applying a different stylesheet
> > 2.	The following can be derived programmatically from the content:
> > a.	A logical, linear reading order
> > b.	Hierarchical elements, such as headings, paragraphs and lists
> > c.	Relationships between elements, such as cross-references and
> > associations between labels and controls
> > d.	Emphasis
> >
> >
> > I've taken out the stuff about markup and data models.  This is
mostly
> > because I don't think it matters how the structure is made
> > programmatically available, as long as it *is* made programmatically
> > available.  This approach is also more flexible for future
> technologies,
> > and a lot less wordy.  I added #1 because I felt that user control
> > needed to be made more explicit.
> >
> > Let me know what you think,
> > Cynthia

Received on Friday, 21 December 2001 13:33:49 UTC