Re: Comments/questions about checkpoint 9.3 (configuration of eventnotification)

An event that doesn't cause a change need not be notified - I am thinking
about changes to the document (am I on the right tram here?).

A change to the UI may not be obvious to somebody using a serial access mode
on a large document (50 characters opf braille doesn't give a very big
picture of 90 lines of document. And the UA guidelines are a fair bit longer
than 90 lines, as is the checklist.

Charles McCN

On Wed, 26 Jul 2000, Ian Jacobs wrote:

  Charles McCathieNevile wrote:
  > 
  > I guess there is still an important issue of whether we require AT to access
  > through an API. (Rich, I'm fishing for comment here <grin/>). If not, then
  > producing the content through the UI is how the user is going to find out
  > what happened. I thought we had a seperate checkpoint that required taht, adn
  > the configuration was to allow the user to turn that off.
  
  No. Here's why:
  
  1) Some events cause no change in the UI. For example, an executed
  script
     that does a calculation and changes content but does not change the
  UI.
  
  2) Other events cause changes to the UI. Since those changes are obvious
     to a user of the native UI, we don't require the UA to provide
     redundant indication of the change in the UI.
  
  3) We require that all events be broadcast through an API. 
  
  4) We make accessibility requirements on the UI, such as checkpoint
     1.5, which requires that messages have text equivalents in the UI
     that may be used by ATs. 
  
  5) No, we are not requiring that ATs do anything. 
  
    - Ian
  
  > Charles MCN
  > 
  > On Tue, 25 Jul 2000, Ian Jacobs wrote:
  > 
  >   Hello,
  > 
  >   In the 7 July Guidelines [1], checkpoint 9.3 (Priority 3) reads:
  > 
  >      Allow the user to configure notification preferences for
  >      common types of content and viewport changes.
  >         Note: For example, allow the user to choose to be notified
  >         (or not) that a script has been executed, that
  >         a new viewport has been opened, that a pulldown menu has
  >         been opened, that a new frame has received focus, etc.
  > 
  >   1) Since this checkpoint does not specify that it is about notification
  >      through an API (which is covered by checkpoint 5.5), our document
  >      says that this checkpoint refers to notification through the
  >      user interface.
  > 
  >   2) Looking back at the history of the checkpoint
  >      (checkpoint 10.2 was introduced in the 9 July 1999 draft [3]), I
  >      believe that originally this requirement was supposed to apply to
  >      notification through an API and notification through the UI.
  >      Refer to 30 June 1999 discussion [4]. We dropped filters
  >      on the API notification at some point since applications can
  >      filter out whatever they wish.
  > 
  >   3) If notification is to be provided through the UI, then by
  >      default all events would have to be indicated to the user.
  >      How would that work in practice? We have to address that
  >      question before we discuss how filtering will work.
  > 
  >   4) If we try to identify a minimal set of events that are
  >      "common types of content and viewport changes", what
  >      would be in that set? We could use the information
  >      in the Note after the checkpoint, but that list is
  >      short and two of them are covered by other checkpoints:
  > 
  >      a) a script has been executed
  >      b) a viewport has been opened (but control over viewport opening
  >         is covered by checkpoint 4.16).
  >      c) a pulldown menu has been opened.
  >      d) a new frame has received focus (but control of focus
  >         change is covered by checkpoint 4.15).
  > 
  >      I would note that checkpoint 1.5 already requires that messages
  >      from the UA have text equivalents in the UI.
  > 
  >   5) The techniques document [2] talks about frame techniques but mostly
  >      disabling notification of changes (on an element basis, for css
  >   properties,
  >      and for changing animations. In short, we don't have many techniques
  >      explaining what events should trigger notifications, nor how that
  >      information could be communicated to the user (e.g., through the
  >      status bar).
  > 
  >   6) Who does notification through the UI benefit? For users with
  >      assistive technologies, we already require that all changes
  >      be sent through an API. What users using the UA's native
  >      UI benefit from notification of changes?
  > 
  >   I'm looking for answers to these questions to figure out what
  >   the minimal requirements for 9.3 are or whether we should delete it.
  >   I realize that notification is very important, but we should flesh
  >   this checkpoint out before we continue with it.
  > 
  >    - Ian
  > 
  >   [1] http://www.w3.org/WAI/UA/WD-UAAG10-20000707/
  >   [2]
  >   http://www.w3.org/WAI/UA/WD-UAAG10-TECHS-20000707/#tech-configure-change-notification
  >   [3] http://www.w3.org/WAI/UA/WAI-USERAGENT-19990709/
  >   [4] http://lists.w3.org/Archives/Public/w3c-wai-ua/1999AprJun/0265.html
  >   --
  >   Ian Jacobs (jacobs@w3.org)   http://www.w3.org/People/Jacobs
  >   Tel:                         +1 831 457-2842
  >   Cell:                        +1 917 450-8783
  > 
  > 
  > --
  > 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
  
  -- 
  Ian Jacobs (jacobs@w3.org)   http://www.w3.org/People/Jacobs
  Tel:                         +1 831 457-2842
  Cell:                        +1 917 450-8783
  

--
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 Wednesday, 26 July 2000 09:44:15 UTC