Chair: Jon Gunderson
Date: Thursday, April 20th
Time: 2:00 pm to 3:30 pm Eastern Standard Time, USA
Call-in: Longfellow Bridge (+1) (617) 252-1038
Chair: Jon Gunderson
Ian Jacons (joined at 3:00 EST)
1. A week from today (April 27th, 4:00 Eastern there will be special joint GL UA call related to markup for navigation - to develop consensus on techniques for marking up links for navigation . Hopefully results will be clear guidelines for inclusion in the techniques UA and GL (Wendy has sent proposal re: using MAP for navigation bars).
2. If UA group does not get through open issue there will be an extra telecon on the 25th Discussion
JG: This is a priority 3. Currently the doc only talks about graphical controls.
Reviewing: Ian's proposal
What would be the minimal requirement to satisfy?
JG: Jon proposed "allow user
GR: Could we change the graphic . For example, use the tool tip to replace the graphic.
AL: would like to have option to order things listed in a menu, particularly useful when accessing auditorially (like phone menus)
DB: sounds like it is going beyond rearranging graphics
JG: Original purpose was for individuals with small range of motion, could cluster controls in a area together. During review, other reasons for rearranging elements were brought up, for example rearranging menu items
HR: Do we consider the menu to be a control?
JA: Does that include min, max...we could end up rearranging whole screen?
JG: min requirement may just be changing the position or number
AL: Yes, I wording does answer the reviewers comments. The re-wording is ok.
Proposed Wording For graphical user interfaces, allow the user to configure the arrangement of user interface controls.
Resolved: Accept IJ's rewording of Tuesday 4/18/00
JG: Allow the user to change the number, position and function of the buttons on available tool bars??
DB: does that mean change the function of a button
AL: Allow the user to change the number, position and selection of the buttons on available tool bars??
DB: we don't want to require a tool bar
Allow the user to change which functions and the positions of the functions on available tool bars
Note to Ian - We want wording to the effect - allow user to choose which function/buttons show up where on the toolbar.
SHOULD MENUS BE INCLUDED
Al: Users of speech won't know about hot keys until they get to in a sequential manner.
JG: May also be an issue for individuals with the cognitive disabilities
JG: seems like order and number could be important for individuals who are blind and who cognitive disabilities
JG: We need to specify a minimum requirements so everyone doesn't interpret this differently
JG: What about - Allow the user to change the order and remove and add of items in pull down menus.
KB: I would want a restore feature if we are going to allow people to remove menu items.
JG Should we add - Also user to restore default configuration.
GR: Would restoring defaults be covered in our profile checkpoint.
JG: Allow the user to restore default the menu and toolbar settings (cross reference checkpoint 10.7)
DB: does this mean default for menus and toolbars together or just one menu
AL: point of cross reference - if there is a profile mechanism user should be able to restore to factory defaults - Users would need restore application defaults
Summary - 3 requirements
1. Allow the user to change which functions and the positions of the functions on available tool bars Note to Ian - We want wording to the effect - allow user to choose which function/buttons show up where on the toolbar.
2. Allow the user to change the order and remove and add of items in pull down menus.
3. Allow the user to restore default the menu and toolbar settings (cross reference checkpoint 10.7)
JG: At f2f - key issues
1) alt equiv is not obscured ,
2) video/image can be viewed simultaneously with the transcript
Notes from Ian UA doesn't have to allow for configuring - if information is not obscured
AL: This is p1 because of the need to support simultaneously magnification and captions. Not intent to always require arbitrary positioning, but some configuration capability that makes simul mag and captions.
AL: Wording on not obscuring not the best way to answer the commenter's questions - want to make it clear - it is not arbitrary control.
JG: Good if we can express in functional needs, that is where obscure wording is helpful
AL: User has to have the choice - content may need to obscure other content - user should have reasonable control, if captions get so big they it has to blot out video..want to leave control in hands of user - whether need to obscure or not
AL: minimum implementation under magnification - need to be able to place captions
Action AL: Al will post to the list a discussion of this topic to show that we still need to work on wording
1)this is priority 1
2) we do need to allow user to choose whether info is obscured or not
3) functional requirements - how much configuration is needed.
Review Ian's proposed new language
1.1 Ensure that every functionality available through the user interface is available through every input API implemented by the user agent. [Priority 1]
Note. This checkpoint does not require developers to implement all operating system input APIs, only to make the software accessible through those they do implement. The device-independence required by this checkpoint applies to the functionalities described by the other checkpoints in this document (e.g., installation, documentation, user agent user interface configuration, etc.). This checkpoint does not require developers to reimplement the input methods associated with the keyboard, pointing device, voice, and other device APIs For example, developers are not required to implement text input through a mouse API.
Note. Developers should use APIs available at a higher level of abstraction than device APIs, provided that, in turn, these higher level APIs make use of the standard device APIs for the operating system.
1.4 Implement the standard keyboard API of the operating system and ensure that every functionality available through the user interface is available through this API. [Priority 1] Note. This checkpoint always applies on systems with a standard keyboard API. This checkpoint is an important special case of checkpoint 1.1. Refer also to checkpoint 10.8
JG: I think we are clear that developers don't need to generate text through a mouse.
JG It is clear that if you implement voice input you should do it accessibly. Developers would probably want to make the whole thing voice accessible.
AL: May run into this problem (voice control not working) during installation. What if installation doesn't support voice input.
JG: Gregory would installation be considered a separate program.
GR: I would consider installation part and partial of product.
*** Ian joined (yeah!)
IJ: There could be a boot strap problem. If you have to install voice component it could be a problem. If OS doesn't support voice, how do you install UA with voice.
GR: Other example, screen reader installation is self voicing - installs a temporary module until actual program gets installed
JG: Let's move this to the list
IJ: any objections to incorporating.
Resolved: adopt Ian's proposal with incorporation of comments from the list.
JG: Related to configure UA to navigate structure. Current min reqs proposal for 7.6 - says 1) allow user to sequentially navigate current through rendering in the viewport human recognizable.....for example, header...... 2) if no structural defined use programmatic structure.... (see draft requirements doc)
IJ: Sounds like minimum is to use document object.
The DOM is limited to XML and HTML. Document objects can include scripts or other
Document Object - user agent's representation........ see Ian's proposed definition
IJ: not clear in generic sense what is actually in the document object... the W3C DOM is specific, document object not
IJ: Structure - What it means is document object. It has structure, so minimal navigation.
JG: For example in HTML, would have a command that would allow you to highlight each element in the document tree.
JG: this would allow users to identify an element on the screen and know what that "chunk" represents. But some people have pointed out that there are human understandable things (e.g. heading level 2 should be under level 2 etc)
IJ: For a lot of these checkpoints it is difficult to specify minimal requirements. Here is one where absolute min. req is navigation of everything in document object in a particular viewport. We expect more and when there are conventions like for HTML, we expect more. We can say min is access to every element in the tree and we can add suggestions for how to do it better.
IJ: If we express what we are thinking - structured navigation - does that help developers.
JG: we want to give people more guidance the we currently are on how to satisfy this checkpoint. The min. should be to provide the capabilities for more complex navigation in the future. This may also be device specific.
IJ: When we say navigate - what are the navigation functions, so far we have forward serial navigation. We haven't talked about going up the tree and across siblings.
Action AG: Al will take an action to post to the list about structured navigation
HR. Does this DOM tree related to rendered content
JG: just because something is in w3c DOM tree doesn't mean it is for human understanding. Just because you are navigating tree doesn't mean it is human understandable. For real structured navigation, what better navigation. Actually want to go a step beyond the DOM. More beneficial once you can configure.
JG: Current minimum is navigating the document tree.
IJ: We do want to clarify the "more" we are looking for in this checkpoint. We decided to stay more general and flexible but we have had a bunch of ideas for navigating (in techniques).
JG: not resolved until we hear from AL
IJ: can we take 2 minutes to hear from Gregory on DOm
GR: concern - we decided to make a forward looking document. Stepping back is more for ER group. There is a dependency between UA and PF to ensure that PF is monitoring DOM work and playing a greater role in ensuring DOM 2 addresses our concerns.
IJ: I think we chose to go with DOM2 because we liked what it had to offer. Dropping to DOM1 is more a strategic decision to get a document out. Few documents are XML. I encourage us to drop to DOM1 and get the UA document out the door. For CSS, I could part with it in order to get this document out and in the next round adopt CSS and events that suit us.
IJ: Still have the option to build in language that we "move up" to DOM2 some time after it is out.
1. wait for DOM 2 to come out (4-6 months)
2) go back to DOM1 - already out, we can control our timing to recommendation, could incorporate DOM2 when desired
3) add contingency language - adopt DOM2 when available (we think we know what we will get)
Harvey- we'll have to go through another 3 months after DOM2 is out
IJ: Much content does not use namespace today. We should get document out now.