Notes on discussions with AOL and Adobe

Hello,

I spoke with both AOL and Adobe today and have a few notes from
those meetings. Based on our very useful discussion, I hope that
both companies will send in formal last call comments. I would
like to thank them for taking the time to chat with me and to
review the document!

Below I have jotted down some remarks that I don't think would be
part of a review by either company. 

Reference document: 9 April 2001 draft [1].

 - Ian

[1] http://www.w3.org/TR/2001/WD-UAAG10-20010409/

---------------------------------------------
Issue 1) Is focus always required?
[Proposal]
---------------------------------------------

A number of checkpoints make requirements related to the content
focus: 1.2, 5.1, 5.2, 6.5, 7.1, 9.1, 9.2, 9.3, 9.4, 9.5, 9.6,
10.2, 10.3, 10.7, 10.8, and 11.5. If the format does not include
any interactive elements, what should become of the focus
requirements? For instance, if a user agent is rendering a movie
format or an image format where there is no interactivity (in the
sense of interactive elements) by specification, what should
happen to the focus requirements?

Proposed:

Perhaps the answer is that these checkpoints don't apply because
the format doesn't allow the user agent to recognize (or control)
enabled elements.

---------------------------------------------
Issue 2) Checkpoint 9.3: Plug-ins and history
[No proposal]
--------------------------------------------- 

Suppose that a plug-in is used to render a particular image
format in a viewport "V". Who is responsible for maintaining the
history information (of focus, selection, and point of regard)
when the user takes focus from "V" and then gives it back to "V"
later? Is it the plug-in's responsibility? Or another module of
the user agent (in which case, what is the API used to send
history information back to the plug-in)?

---------------------------------------------
Issue 3) Checkpoints 3.2/4.4: All animated images?
[No proposal]
--------------------------------------------- 

It's not clear that all animated image formats make sense with
the requirements of checkpoint 3.2. For instance, those animated
images "in a box" make sense (e.g., with respect to placeholders
that are also "box-like"), but other animated SVG images may not
be "box-like".

There may be some cases where certain animations are authored in
a way that makes certain requirements not really apply. For
instance, it's possible to author an SVG animation where the
animation changes based on user input. What does "fast forward"
mean for such an image (checkpoint 4.5). In other cases, there
may be interdependencies among animations that make the
element-level control of 4.4 difficult or impossible (these are
the animations that are not "box-like")..

I hope that the Adobe review will help us get clarification on
these technical issues.

--------------
Clarifications
--------------

Here are some proposed minor clarifications:

 - The definition of user agent mentions plug-ins (as potential
   components of a conforming user agent). They should also be
   mentioned in the definition of viewport, and possibly in 
   the Note after checkpoint 9.1, since a plug-in may contribute
   a viewport to the set of (navigable) viewports.

 - There are some bugs in the definition of "explicit user
   request". The term "recognize" is used, and that is part of
   the definition, but the user agent may recognize something 
   that the *user* doesn't know about (e.g., because it's part
   of the markup that the user doesn't see). We need to clarify
   how an "explicit user request" is identified. I don't have
   a proposal today.
 
--------------
Techniques
--------------

 - Add more techniques about plug-in interfaces.
 - Checkpoint 3.6: Add the same HTML markup used in 3.5.

-- 
Ian Jacobs (ij@w3.org)   http://www.w3.org/People/Jacobs
Tel:                     +1 831 457-2842
Cell:                    +1 917 450-8783

Received on Friday, 20 April 2001 21:18:47 UTC