Chair: Jon
Gunderson
Date: Wednesday, October 28th
Time: 12:00 noon to1:00 pm Eastern Standard Time
Call-in: (+1) 617/258-7910
Review of of results from from face-to-face meeting
Discussion of reformatting of guidelines
Assignment of guideline and technique development
Jon Gunderson (Chair)
Ian Jacobs (Scribe)
Judy Brewer
Kathy Hewitt
Kitch Barnicle
Charles McCathieNeville
Marja-Riitta Koivunen
Paul Adelson
Scott Leubking
- Ian and Charles: Arrange a meeting to discuss an appendix specific to HTML/CSS.
The following actions involve authoring based on the upcoming 20 October draft.
1) Discussion of upcoming document restructure
Ian discusses new document structure, that will be available Friday
Charles: What is missing is a list of elements.
IJ: I agree that this should be an appendix.
Charles: I'd like a checklist w.r.t. elements/attributes/properties and their priorities that I'd like to take back to developers.
JG: Input from WG participants on issues like these is very important. Charles, would you like to assemble such a list?
Action Ian and Charles: Arrange a meeting to discuss this appendix.
2) Division of labor
JG: one problem that emerged during the FTF meeting was that grouping of information made the guidelines difficult to "check off". The new format will address this. Longer list of priority 1's, but each will be more specific. Also, Priorities will be expressed in terms of user needs. Some techniques will be required to be implemented natively. Others may be implemented by assistive technologies, with hooks where necessary.
Scott: Will the other access technolgy developers agree to this?
JG: Some of the developers at the meeting were not aware of technologies like the DOM. I would like to organize a workshop for developers to expose technologies to them.
JB: Good idea.
JG: The developers at the meeting were more enthusiastic now that they had been better exposed to what the WG and the W3C does.
Scott: Is this a philosophy that we want to pursue? That implementations by other software will lag behind.
JG: We would ideally like all UAs to be accessible out of the box ...
JB: Hang on. The issue is what the guidelines are trying to accomplish. We concluded at the meeting that many functionalities were the responsibility of user agents natively.
Scott: How well does HJ handle javascript?
IJ: I think that the WG will have to make some difficult decisions, but has the notion of "division of labor" clearly in mind.
JG: Guidelines will reflect current state of technology.
JB: In the upcoming draft, as part of an introduction.
IJ: (I already wrote one this morning).
JB: I don't agree with Scott's fear nor JG's statement about attending to this specific moment in time.
PA: One of the issues is that the document may outlast the current state of technology. Don't want the document to rot too quickly.
/* Mini discussion of DOM */
Scott: My fear is that the document will be used as a justification why something is not being supported.
PA: The statement "If the user agent doesn't support it then, " is the one I oppose. Should be "even if it supports it, it must make available the information".
3) Scripts
JG: One issue that we addressed more clearly at the face-to-face involved scripting. The WG felt that with scripting, the author was creating a new interface. The issue was not the accessibility of scripting itself, but the use of the script and that the use has to be accessible. (For example, some scripts perform invisibile tasks that don't really "concern" the user.) Since the author is creating a new interface, they should observe guidelines related to interface: keyboard accessibility, etc.
Additional suggestion that the UA provide repair strategies for inaccessible authoring (e.g., that use mouse clicks only).
KH: Let's be clear about that. The "repair strategy" is very hard to implement. Our developers don't know how to do this easily with the object model.
Charles: E.g., generate "onactivate" for an "onclick" event. "Onmouseout" and "ondoubleclick" are hairy.
IJ: I think that for the list of finite events defined by HTML, that access to behaviors attached to them should be obvious to implement (although admittedly politically difficult, e.g., due to the underlying keyboard model). This is distinct from a UA having to "know what a random script does to the document", which I agree cannot be known in the general case.
4) Getting participants involved in helping to author the document.
5) Upcoming teleconfs.
JG: Do people want to have weekly meetings?
IJ: W3C meeting in Kyoto week of 18 November
Thanksgiving the following week.
JB: I think they should be held weekly until the document goes to PR.
JB: Once Friday draft available, WG will review quickly, and assess whether another ftf will be required at the 4 November teleconference. Need to make decision <em>urgently</em> due to W3C Process concerns.
JG: A potential face-to-face could take place the second week of December.