Chair: Jon Gunderson
Date: Thursday, April 27th
Time: 2:00 pm to 3:30 pm Eastern Standard Time, USA
Call-in: Longfellow Bridge (+1) (617) 252-1038
Chair: Jon Gunderson
Scribe: Ian Jacobs
Gregory J. Rosmaita
NOTE: This is different from the time originally sent to the group! Agenda
2. IJ: Update document based on 4/25/00 telecon resolutions
3. IJ: Contact MR on importance of arbitrary repostioning in Checkpoint 4.7 Email sent to list http://lists.w3.org/Archives/Public/w3c-wai-ua/2000AprJun/0208.html
5.DA: Send name of new organization to list that was mentioned by some from
the US Census Bureau
4.CMN: Propose a technique that explains how serialization plus navigation would suffice for Checkpoint 8.1.
6.DA: Review techniques for Guidelines 7 and 8 7.
DA: Get confirmation that the numbers for checkpoint 4.5 make sense
8.DB: Get Tim Lacy to review G+
9.GR: Look into which checkpoints would benefit from audio examples in the techniques document.
10.GR: Review techniques for Sections 3.7 and 3.8
Status: Finshed 3.7
11.GR: Send to list screen shot of JFW Window list.
IJ: Draft a preliminary executive summary/mini-FAQ for developers. (No deadline.)
12.MQ: Review techniques for Guidelines 9 and 10
13.MR: Send URI to Micrsoft's implementation of synchronized audio/video slowing down to the list
14.RS, AG: Take notification of focus and view changes to PF as possible DOM
15.WG: Read EH's proposal related to definition of content JG, MN, DP, IJ, DA, KB
1. Additional telecon on Tuesday, 2 May
2. Joint UA/WC Telecon moved to May 4th
Resolved: Adopt Eric's proposal + Ian's addendum.
Action IJ: Add definitions to the document.
/* Ian summarizes discussion of last week */
AG: The requirement is that you be able to position captions w.r.t. the video. (Needs to be clear in checkpoint.)
TL: I'd like to run this by the IE Team.
IJ: Note that the primary question is what's the accessibility requirement.
TL: I will send email to the list.
AG: You could ask Cindy King (at Galudet); she's done work in this area.
Action AG: Write to her and copy me. 3.PR#233: Checkpoint
7.6: What does "structure" mean here?
Refer to JG's email:
Refer to Al's email:
AG: I thought next week's teleconf on navigation was for fact-finding?
JG: More about markup.
AG: How can you tell the implementor what is structural? You want to discuss with WCAG WG what you know from the markup.
IJ: Some things are encoded in HTML spec.
AG: But some are encoded in layout, not in header tree. Proposal:
1) We leave the word "structure" in the checkpoint text (Allow the user to navigate according to structure.) AG: The "gotcha" I see is when the markup is not helpful to you (e.g., current practice). Skip subtree capability is assumed by WCAG (e.g., skipping navbars). That's why teleconf next week would be useful: talk to WCAG about what model in UAs may be expected by authors.
2) Explain that structure is based on the document object model (in the generic sense) but may have more semantic information when known from the markup language definition, through schemas, etc.
3) Therefore, the minimal requirement (since the goals are orientation and rapid navigation) is to allow navigation according to the semantic model known for the markup language, paying particular attention to block and grouping mechanisms, and having a subtree skip and/or hide functionality.
AG: I'm not happy with this proposal. You should state more of what the user
1) Decompose the page into principle parts
2) Be able to get to the beginning of any part (quickly).
MN: That's what the "e.g." in the current 7.6 says to me. IJ: Two
a) Where does the structure come from? - Semantic model (structure, schemas, specifications) - Repair of poorly constructed page is extra-credit.
b) What set of functionalities of navigation?
AG: I don't think that we should be espousing theories about what the UA should provide if why don't apply to actual content in the world. We need evidence that something is effective, not only readily achievable. The specs on whole are not effective in this regard (point "a" from Ian). I think you can say some minimum things from known structures (e.g., headers).
AG: I think it's easier to come up with performance requirements (e.g., lists no lnoger than five) rather than rules for extracting the navigation tree from markup.
AG: I agree with statements like "If there are headings, you should include them."
IJ: Are you saying we can't rely on the markup?
AG: No. If you just use HTML, you don't get a usable product. You have people using tables and frames to implement the same functional object. People have N ways of building a header bar. The specs don't tell you how to do this.
AG: You have evidence of user agents doing good things in structured navigation, but you have not criteria for drawing the line.
JG: Are you suggesting that we leave open the minimal requirement?
AG: Almost: include what is necessary, but indicate that it's not sufficient. More research would be necessary to know what's a minimum requirement for a markup language.
KG: I think we can make the argument about a lot of checkpoints.
DA: Then we'll continue to get the complaint that people don't know when they conform.
JG: Does the configuration checkpoint help us?
AG: It helps me. I think you have to consider 7.6, 7.7, and 8.6 (orientation) together.
AG: The user agent's view has to have structure, so that the user can form a navigation model. Yes, the markup specs and author's intent help infer structure.
IJ: In my mind, the author provides a "view" and the user agent allows the user to handle this view flexibly.
AG: The requirement for structured navigation to provide a structure that the user can move through. How this relates to what's provided in the format is part of the implementation. It's not the basic requirement. The user needs to break the problem down into parts. And to be able to navigate to them. The mission statement is missing.
IJ: I hear Al saying that "chunk" navigation is the requirement. We subsumed this under "structured navigation" since we knew structure generically (doc tree). This checkpoint subsumed a number of other checkpoints (e.g., table cell navigation), which are in techniques.
/* Charles drops out */
a) Chunk navigation is not the only piece we need to address. But it's a good example of where our problems come from.
b) In PF, we're working on XML (DTD) guidelines. So that users can break down content into component parts.
AG: We want this guideline to drive the PF dialogues. I need this to state the functional requirement for the user.
IJ: "Navigate efficiently to significant components of content."
RS: What about "contextual navigation"? AG: "Contextual navigation" is a superclass of "structured navigation".
GR: Need to provide the ability to expand and contract subtrees.
GR: Issues of navigation and others are coming to light in the effort to generalize the content guidelines.
Action AG: Write comments based on current techniques as fodder for the WCAG/UA joint teleconf on 4 May.
Action JG: Write a proposal / clarification about 7.6 to the list (to clarify "navigate" and "structure").
/* Discussion of timing issues: WWW9, then AC meeting, then Ian vacation, then Jon unavailable in June, then summer slowdown, ... */
http://cmos-eng.rehab.uiuc.edu/ua-issues/issues-linear.html#207 Refer to JG summary:
AG: I concur with summary sent by JG.
DA: Problem if I need to access the summary and all I have is a source view, that doesn't cut it. You have to have access to equivalents in the same context.
IJ: Refer to my proposal that says this: http://lists.w3.org/Archives/Public/w3c-wai-ua/2000AprJun/0223.html
DA: But it's not just alternative equivalents that need to be available in the same view. Attributes, too.