Chair: Jon Gunderson
Date: Thursday, 11 May
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
Present:
David Poehlman
Jim Allan
Eric Hansen
Dick Brown
Mark Novak
Tim Lacy
Madeleine Rothberg
Mickey Quenzer
Regrets:
Harvey Bingham
Gregory Rosmaita
Charles McCathieNevile
Denis Anson
Rich Schwerdtfeger
Al Gilman
Hans Riesebos
Kitch Barnicle
Absent:
none
Next teleconference: May 18 at 2pm ET (regular time)
Agenda [1] [1] http://lists.w3.org/Archives/Public/w3c-wai-ua/2000AprJun/0346.html
1.IJ: Find out what HTTP gives you in the way of resource name
2.IJ: Propose a clarification text of 2.5 to make clear that the user agent is expected to associate a text equivalent with the object, text eq generated from author-supplied information.
3.IJ: Propose revision to checkpoint 3.8
7.MR: Confirm usage of "configure" in checkpoints to verify that it means "static choice" appropriately (e.g., 4.9). Open Action Items
4.IJ: Draft a preliminary executive summary/mini-FAQ for developers. (No deadline.)
5.CMN: Propose a technique that explains how serialization plus navigation would suffice for Checkpoint 8.1.
6.GR: Look into which checkpoints would benefit from audio examples in the techniques document.
1. Notice of Proposed Rulemaking: Electronic and Information Technology Accessibility Standards by the United States ARCHITECTURAL AND TRANSPORTATION BARRIERS COMPLIANCE BOARD.
2, Review and comment on proposed resolutions on the list (see agenda)
JG: I will consider these resolved unless there are objections on the list.
JG: Our charter expired at the end of the
IJ: We need time to get to Rec. Then we would produce an implementation report and a summary of our experience. In the charter after that, we would talk about our next work.
EH: Expected duration of extension?
IJ: For the moment, I've taken conservative approach: assume a last call again. However, we might get done sooner. This depends on whether the WG considers that the document is substantially different.
http://cmos-eng.rehab.uiuc.edu/ua-issues/issues-linear.html#257
See Ian's analysis of the checkpoints that need minimum requirements:
http://www.w3.org/WAI/UA/2000/05/ua-minreqs.html
EH: The grouping seemed reasonable to me.
JG: We talked about checkpoint 4.5 last week.
MR: Need to clarify that
a) Slowing all video and animations.
b) When its a multimedia presentation, for any video slowed to 80% of original
speed, need to synchronize.
Proposed clarification: - For a visual track for animation or movie: At least one setting between 40% and 60% of the original speed. - For a pre-recorded auditory track including stand-alone audio presentations: At least one setting between 75% - 80% of the original speed.
DP: For synthesized speech, you can drop to 50%.
JG: The range should be that of the synthesizer. You can query the hardware for the range.
IJ: What about software synthesis (e.g., mbrola).
MR: Pre-recorded only? What about live audio that you can slow down?
EH: Then this is almost like "Pre-recorded".
DB: This might not be possible in some cases.
Action EH: Propose definitions for auditory and visual
tracks. - For synthesized speech (4.9)
1) Get the range from the synthesizer, or
2) Otherwise, get the range from style sheets if supported
DP, TL: Don't think we should have a third option of arbitrary range.
TL: Wouldn't you lose synchronization if speech is handled externally in a multimedia presentation? How do you slow down a video presentation to match the speech rate you chose.
JG: In this case, the synthesizer would speak that the rate that the caption would be appearing.
DP: A lot of synthesizers can change rates dynamically, depending on what they're doing.
IJ: Is a user agent capable of handling any range of speeds?
JG: High rate of speech and synchronization (e.g., highlighting on the screen) might cause some problems. Otherwise, the user agent should be able to handle any range the synthesizer is capable of.
TL: How fast the ua feeds the synthesizer doesn't matter.
IJ: So what about an arbitrary range if 1 and 2 not available?
TL: I don't know that arbitrary range will fly. What if the synthesizer can't handle the arbitrary range we specify? How should that be handled?
EH: Then applicability kicks in.
IJ: What values should be supported?
JG: Normal talking rate is about 180 words per minutes. Real speech synths allow between about 120 words-per-minute to approximately 500-600 words per minute.
DP: We use about 240 words per minute daily.
IJ: Then I PROPOSE that we suggest a range of about 120 to 500 words per minute.
TL: Overruns are a non-issue
DP: We are also asking for configuration. So applicability comes into play when there is no range. I don't see configurability in multimedia players. I see some activities that use synthesized speech (e.g., chat programs). Also, "MagicSpell", that reads email, but doesn't let you change speed, but not pitch. - For a synchronized multimedia presentation: For any visual track that may be slowed from 100% to to 80% of its original speed, need to synchronize visual and auditory tracks. Below 80%, the user agent can drop out the auditory track.
MR: What about other types of equivalents (e.g., text equivalents)?
JG proposal:
http://lists.w3.org/Archives/Public/w3c-wai-ua/2000AprJun/0356.html
EH: I agree, but some of the language of JG's proposal needs fixing.
IJ: I think the minimum is:
a) For markup with a clear semantic layer: navigation of important semantic
elements according to specification,
b) Otherwise, navigation of the document object.
c) Sequential (depth-first) forward and reverse navigation is the minimum (P2).
IJ: What about dynamically changing the scope of objects?
To let you delve into lists, forms, tables and stay within that scope.
What about in a generic markup language?
How do you identify subtrees to skip over?
EH: We may need to clarify for 7.6 ("to and among"): Was "within" intended?
IJ: No. We could drop "to". It's the first step to "among".
MN: Is it a problem that navigation and the configuration of navigation have different priorities?
EH: I think it's reasonable to have different priorities.
JG: 7.7 makes 7.6 much more powerful.
MR: I think 7.7 should be both configure and control.
EH: Then are people satisfied with P3 for 7.7?
MN: I'm not comfortable with it.
IJ: Then it makes sense to clarify in 7.7 that we're talking about the objects being navigated, not the navigation abilities themselves (i.e., sequential).
EH: We have among and within / control and configure. We have a max of 4 checkpoints.
IJ: I propose keeping "control and configure" together. Who would like 7.7 (configuration/control) to be P2: JA, MN, TL, JG, DB, probably EH.
TL: If you think about navigation according to structure from farther away, a user agent that doesn't allow configuration of the navigation will exclude users. It seems that they should both be the same priority.
IJ: Note on changing priority to P3: P3 means easier, but P2 means difficult to without it.
JG: I hear people saying that without configuration, it would be difficult for people to do what they need to do.
JG: I'm concerned about the WG changing the priority at this point since that question didn't come up during any of our reviews.
TL: If 7.6 is implemented to deny access to some content, that would cause problems.
Action IJ: Propose new 7.6:
a) Clarify that the what is configurable is the set of objections, not
navigation abilities.
b) Configure and control
c) Add a note that control allows a skip-over functionality (although that
could be implemented separately as better then 7.6 minimal sequential
requirement0.
Checkpoints 7.1 and 7.3 (P1): Forward, sequential navigation.
Checkpoint 7.4 (P2): Forward and reverse sequential navigation.
Checkpoints 8.4 and 8.5 Refer to JG's proposal: http://lists.w3.org/Archives/Public/w3c-wai-ua/2000AprJun/0357.html
JG: Perhaps this is the same situation as 7.6/7.7: the functionality is not as useful if you can't configure.
MR: But the minimal requirement of 8.4 should make it useful even when you can't configure it.
EH: Same comment about "resource specification"
IJ: In any case 7.6 and 8.4 need to be harmonized. It sounds like the outline view should be those elements navigable according to 7.6. 7.7 should follow 8.5 as well.
IJ: Configurability of navigation seems more like a P2 than configuration of the outline.
IJ: The intent of the outline is to give people an overview: and to hide some (a lot of) content.
JG: E.g,. for low vision or if you have a reading disability, you want less content in front of view.
Action IJ: Propose a clarification to 8.4 to indicate that what is intended is to hide content (and why). Indicate in techniques where to get information for outline entries. Delete "lists" from example. Question: How do you compose the outline view for things that don't have a label?
JG: Get header information according to spec; otherwise compose a label to indicate type. Min Req: - 8.4 Outline view is the set of things navigable according to 7.6. - 8.5 Maps to 7.7. - The two functionalities are not bound, but the ranges are.
Action IJ: Propose synthesis of 7.7/7.7/8.4/8.5
Refer to JGs Proposal:
http://lists.w3.org/Archives/Public/w3c-wai-ua/2000AprJun/0358.html
Resolved: Adopt JG's proposal (since previously resolved).
http://cmos-eng.rehab.uiuc.edu/ua-issues/issues-linear.html#280
See JG's proposal to drop the addition of a checkpoint on synchronization
http://lists.w3.org/Archives/Public/w3c-wai-ua/2000AprJun/0337.html
IJ: Any objection to dropping the sync requirement.
Resolved: Drop the proposed synchronization checkpoint.
http://lists.w3.org/Archives/Public/w3c-wai-ua/2000AprJun/0354.html
EH: I generally endorse her proposal. People should review her proposal.