Re: GL 3.2 proposal and summary

>All issues surrounding extreme change of context has to do with pop-up
>windows and, to a much smaller degree, content updates. The examples
>named a couple of caption identification for speaker change in audio
>presentation, but those should belong to GL 1.3 and UA. None of the open
>issues has anything to do with captioning. Of the 48 open issues, 21 are
>related to extreme change of context. Almost all of them are about
>pop-up windows. Therefore, I propose eliminating the term "extreme
>change of context" and directly address the pop-up windows issues. This
>will remove the ambiguity.
>
>Next, many of the issues relating to "extreme change of context" would
>be solved if user agents can inform users about the pop-up windows. This
>is a more efficient method to address the problem. We need to pass on to
>UAAG. In particular, UAAG 1.0 GL 10.5 should have informed the users of
>pop-up windows that are expected. It may not solve the unexpected pop-up
>windows such as pop-up advertising. In all cases, all pop-up windows can
>be programmatically determined anyway. I know of no way for pop-up
>windows to fail level 1. Additionally, the pop-up windows can now be
>blocked by all new browsers. In conclusion, I propose to remove the SC
>related to "extreme change of context".

The goal of the level 1 success criterion was to make sure that the author 
had provided suitable information, and was using a suitable technology, so 
that a user agent could either inform the user about pop-up windows or 
prohibit them. It does seem as if "programmatically determined" is not 
doing the trick here. Maybe we need to change this success criterion to the 
desired functional outcome, that the user be advised about or be able to 
disable "extreme change of context".

>There is a left-over issue of content update that I would post to the
>group.  Let's say we have a web page for NBA (professional basketball)
>statistics.  To reach the statistics of an individual team, the user
>pick from a drop down list of teams.  The content is updated based on
>the selection.  Is it necessary to have a separate submission button to
>warn of content change before the content update should occur?  Should
>tabbing out of the selection list be enough?  Another example can be
>real time content such as weather, stock quote, auction price, and time
>sensitive game, in which content is updated constantly.  All of these
>content updates are expected can be programmatically determined.  If we
>leave anything from "extreme change of context", this may be it.  But I
>think this is also an user agent issue.  But I would like to hear from
>everybody.
This issue of content changing, either in response to some external event 
or as a side effect of some user action, is a serious accessibility issue. 
Blind users often have no idea that anything has changed at all, even 
thought it is possible for them to discover it if they explore. Sometimes, 
the act of exploring will cause more side-effect changes, so that it is 
never possible to read the first set of new content at all. Again, 
"programmatically deteremined" may be the wrong condition, but I think we 
need some success criterion that addresses this.

><sc>L2 SC6</sc>
>
><proposed>No number 6 success criterion</proposed>
>
><current>The destination of each link
><<http://www.w3.org/TR/WCAG20/#linkdef>http://www.w3.org/TR/WCAG20/#linkdef> 
>is identified through words or
>phrases that either occur in the link or can be programmatically
>determined. </current>
>
><rationale>This is already covered by UAAG.</rationale>
I don't understand how this is already covered by UAAG. This seems like a 
case where the words to describe the destination can only come from the 
author, especially if the link text itself is not sufficient.

Received on Tuesday, 24 May 2005 23:24:05 UTC