See also: IRC log
<trackbot> Date: 17 September 2010
zakim this will be WAI_AUWG
can you skype to the 617 number?
+44.203.318.0479 (new) is the new UK number. Sorry, zakim needs to be updated
zakim ??P14 is ARonksley
zakim who is here?
<Jan> Conformance idea: http://lists.w3.org/Archives/Public/w3c-wai-au/2010JulSep/0089.html
<Jan> Tim's comment: http://lists.w3.org/Archives/Public/w3c-wai-au/2010JulSep/0090.html
JT: There us a circle of Tool Compliance for part A and B. There is Partial Compliance which is Part A or Part B
There is compatible Component which is Part A and Part B and optionally B4.
Jeanne: I want to insure that the conformance labels are clear so that buyers and purchasing agents can easily tell whether they are getting a system that is fully ATAG conformant, or getting a system with partial or component ATAG conformance.
the group takes a look a the examples of Authoring Tools, so see if they conform to this new conformance proposal.
HTML Editors = ATAG component
Direct editor = ATAG Component
Converting to Web = ATAG Component
Integrated Devleopment environment could be either a Component or a System if they chose to add the web libraries and accessibility checkers.
[some discussion about the definition of authoring tool and whether evaluation tools are Authoring Tools. ]
SN: I don't think eval tools are authoring tools because they don't create content, the same way that debuggers are not authoring tools.
JR: [gives an example of an authoring tool component like spellcheckers that provide a feature but don't actually write content]
Blogs, wikis =ATAG Components
CMS, LMS = Atag Component or Full ATAG System
Email client that create html emails - may be a component, but there is discussion that it is publishing so it really is a system.
SN: Compatible component needs more definition
JR: Example of a tool creating a format, for which a checker doesn't exist.
SN: So we always have to go back
to the criteria always
... People already have to meet 508 and criteria of WCAG and
the levels. ATAG added to this is adding complexity. I think
there will be problems of uptake.
GP: At least of you catagorize yourself as a system or a component, at least people know what they are looking at.
JT: If what we are trying to do is get web content then we have to look @@ missed @@
SN: I am a system, but I don't
meet all the requirements, so then that makes me a
component?
... no, you are still a system.
... there is a difference because I am still a system but I
fail some criteria
So how do we determine what is a system?
JS: Is Wordpress a component or a system?
JR: We leave it flexible so Wordpress can decide if they are a system or component.
SN: So as soon as someone knows
that they won't make it, they stop trying.
... I am very concerned about communicating it, because I'm not
sure it is communicatable.
... this breaks the normal conformance model that people are
used to. If you want people to follow this, we need to be able
to communicate it.
JR: if the system allows the introduction of accessibility problems, then it doesn't need checking.
SN: It is whether or not they met
the criteria making a claim of what it met and didn't
meet.
... I think creating other classifications is going to make it
difficult to get back accurate information.
JS: So is a blog a system? They would say they are a system - and they can use an external for check and repair. And that would make them not a system?
SN: "You have to have checking,
but if you don't, that's ok". Do we want to do that?
... what is ATAG, if we don't demand checking.
GP: Let Wordpress and Deque bundle and offer a compliant package.
Alex Li, Microsoft, has joined the call. Alex discloses that he has not been able to complete the IP rights declaration.
Jeanne confirms with Judy Brewer that because Microsoft has previously signed a patent disclosure agreement, that Alex Li can attend as a guest.
JR: MS39: Variations from WCAG on the excemption for a path of the user movements. ATAG also offers an exemption for pressure, force, angle.
AL: Use "path" as the common denominator. Force, angle and speed are all path dependent.
AL: If you have just a basic web form, if the form results in a generation of web based output, then it becomes the authoring tool. Then this is not in the control of the author, it is controlled by a back-end system.
JR: We are aware of the problem and are working on it.
JR: We are not requiring UAAG conformance, because an author is not served by an unrealistic world view.
AL: Then how will it fail?
JR: If someone creates their own
HTML parser.
... Ideas include: existing user agents, publically
available
AL: commercially available.
AL: If it is a Microsoft propriatary format, we would be ok. If it was not a microsoft format, then we do have a legal issue of liability of MS making a statement saying "information will be lost".
I do not want to cause difficulties with partners. I talked with legal, and they said it was not ok. It would cause us too much trouble.
JR: I have seen that warning with Excel saving as csv.
AL: It is very different saying
that accessibility is lost, because of the legal requirements
around accessibility.
... If we say in a very generic way "information may be lost"
that is ok, but if there is any mention of accessibility
information being lost, that is no-go from a legal
viewpoint.
JR: PDF has accessibility support, but a tool may have a Save As option to create a flat, unstructured PDF.
AL: I will talk to legal and see
if they will accept the nuance of it.
... I see a subtle difference between this and the previous
item. I will talk to the legal team.
AL: If there are multiple interfaces to control various properties, like the menu, the ribbon, a right-click menu. If I have different mechanisms, would i have to set accessibility property in every mechanism?
JS: example of image, that whereever the height, width and border are set, and the alt is set.
JT: We want an integrated approach to including the accessibility properties, if the UI is distributed, then the bounds on where the accessibility mechanisms are located is also distributed.
<gpisocky> Discussion of MS23 brings to mind another concern regarding the impact on proponents of targetted formats
GP: that a vendor could give warnings to give advantage to one format over another.
AL: Where we have anti-trust issue, that would be a problem.
JR: Because we are not closely prescriptive, we want people to know the background of what the developer is claiming.
AL: But WCAG doesn't require it.
JT: But it is important to be public.
AL: I see the inconsistency with WCAG.
JT: Then I think our next step would be to go to WCAG and ask them to make it a change.
AL: If WCAG made a normative errata publication, I would not object to it. In fact, the only WCAG compliance statements are on the web, so it is how it works in reality.
JT: We agreed to say that we will write a definition of Added Information
JT: The definitions were taken from WCAG.
AL: I will look at WCAG.
JR: ATAG goes beyond what WCAG was addressing.
Section (a) is the chrome of the widget, section (b) is the embedded widget that grabs control of focus.
JR: There is a complex relationship between the editing view and the user interface.
JT: There are behaviors and properties of the editing tool that are unique to the authoring tool perspective.
AL: So how would you exit out of (b) to (a)?
JR: There could be a keystroke that the authoring tool reserves for itself that could be used for returning control to the auhtoing tool user interface.
JT: It is visual only, and would only apply to content. It mostly means requiring a stop button.
AL: There is a big difference between a @@ and a stop button.
JT: we will reword it.
JT: If there is structure, then we want to use it for navigation.
AL: The structure may not be there for the author, but may be applied by the tool.
JR: Structured at the time of editing, not the way it will be structured in the end product.
AL: Does that apply to the comment MS12 - all structures.
JR: We don't want to specify
AL: There is so much web structure, not all of them apply to navigation.
JT: We agreed that one is
sufficient, but are considering an additional SC to require
more.
... [reads MS 18, 19, 21, 22] these are all items we agreed
with the MS comments, so there is no discussion.
AL: doesn't address real time
publishing.
... I'm also considering all real-time information - stock
information, banking application, supply chain, etc.
JT: but banking, that will only affect myself.
AL: But the bank officer will also see it. It can go all the way to regulartory authorities and other banking systems.
JR: because the input is so contrained that this may or may not introduce an accessibility problem.
AL: But we don't know where the
accessibility problems can arise. Where is the value chain end
in authoring? Does it go to the database and how the database
is related to other systems? At what point does it become an
authoring tool.
... that is something we need to address in the definition of
authoring tool.
TB: Whenever you create content, you don't know where it will end up.
JT: We have agreed to work on the condition. The read-only issue is not relevant, because of the wrapper.
AL: the word document with a graphic is created with read-only. the author cannot make the graphic accessible.
JR: That is covered in another applicability note about author permission
AL: I think it is all based on WCAG. You need to say the following normative list requires judgement.
JR: But it will vary by the tool.
AL: A contrast checker will only check contrast.
JR: does the image need long description? The tool will ask a prompt for a decision, or the tool may look at a 1x1 white image and decide not to prompt.
AL: the authoring tool developer is deciding when to ask for author decision.
JR: The UI needs to provide some support to help the author in making the decision.
AL: the developers are asking for a finite list.
JR: ANytime the author is asked for a decision, we need to provide some support in helping the author make that decision.
AL: ok
MS32 B.2.2.4 Help Authors Locate
JT: It is help in determining the bounds of where the problem may be located.
AL: It implies too much
intelligence on the part of the tool.
... people will ask "do you meet 2.2.4, you don't show the
location of this kind of error"
JR: Agreed, we need to tighten up our language.
JR: Relevant sources is just the handle.
AL: it is a "slushy" term.
AL: If the notes are normative, they have to be testable.
AL: Imagine using a web form to do an authoring? Do you need short-cut key?
JT: One of the weaknesses in WAI has been the support for users with mobility problems. We want to support authors with alternative interfaces.
AL: for web-form, it is too much. Web apps need shortcuts, but a simple form does not.
JR: Make it complexity based
JR: It could claim the find
feature of the web browser in the conformance claim.
... the user agent platform has to be identified since they
have different capabilities.
JT: A.3.6.1. applies to more than
just keystrokes. We have changed the wording of this to
"perference settings"
... it is hard enough to set the preferences to access the
tool, so we want to save the preferences that the user does not
have to set it again.
... we took out the "control and display" settings and replace
it with "preference settings".
AL: I don't dispute the validity of the success criteria, I just spotted the inconsistency. I will need to think about whether it impacts the inconsistency.
MS 44, agreed, agreed
JT: We mean features that users can use. We don't mean that hidden features need to be documented.
AL: SOme of the tools have so many features, we don't know if we ever could document all features.
JT: "All features available to the author".
AL: If the tool is big enough, something will always be undocumented.
JR: we see the issue and it is not the spirit we disagree on, just the details.
AL: Nobody does this.
JR: MS48 - provide a real life option with (c). We have not found any examples of this.
JT: There is advocacy to make it double AA because there are other systems that want to use this metadata.
This seems more of a future development than current. We agree theoretically, but it is not today.
JT: It is common practice in Dublin Core in education environments. It is current.
AL: I will have to think about it more. Having the check is fine, but associating it as meta data, I don't think it is not as practical.
JT: We can send you examples, and
speak with Bob Sinclair. It would be good to have a
conversation within Microsoft about it. AL: NPII is not close
to implementation.
... But this is a precursor to implementation.
... ATAG is not just to codify what already exists, it is also
to move the agenda of accessibility further.
AL: I see your point.
JT: We simply want a label of accessibility, not a ranking.
AL: Most of these people don't know what they are doing with accessible templates. They will check the box, and they will create a lot of misinformation. it will be wrong most of the time.
SN: We would have to create a list of criteria that a template must meet to be accessible.
JS: WCAG criteria?
AL: People won't check WCAG
JR: The spirit we are trying to get to is comparable prominence. The example of a spellchecker with underlined words is much more prominent than a checker that needs to be run from a 3rd level menu.
<Zakim> Adobe_room has Greg, Sueann, Jutta, Jeanne, Jan
<scribe> chair: Jutta
SN: iAccessible2 and DOMs are an important part of the platform accessibility architecture.
Adding Iaccessible2 to the definition of Platform Architecture should be sufficient.
<scribe> ACTION: JR to write proposal for conformance with Tim Boland [recorded in http://www.w3.org/2010/09/17-au-minutes.html#action01]
<trackbot> Created ACTION-295 - Write proposal for conformance with Tim Boland [on Jan Richards - due 2010-09-24].
<scribe> ACTION: GP to write a proposal on who can be a claimant with Jeanne Spellman [recorded in http://www.w3.org/2010/09/17-au-minutes.html#action02]
<trackbot> Created ACTION-296 - Write a proposal on who can be a claimant with Jeanne Spellman [on Greg Pisocky - due 2010-09-24].
<scribe> ACTION: JR to write proposal on Programmatically determined [recorded in http://www.w3.org/2010/09/17-au-minutes.html#action03]
<trackbot> Created ACTION-297 - Write proposal on Programmatically determined [on Jan Richards - due 2010-09-24].
<scribe> ACTION: JR to write a proposal on path, speed and pressure [recorded in http://www.w3.org/2010/09/17-au-minutes.html#action04]
<trackbot> Created ACTION-298 - Write a proposal on path, speed and pressure [on Jan Richards - due 2010-09-24].
<scribe> ACTION: SN to write proposal on A.1.2.1 non-web based accessible (comment IBM15). [recorded in http://www.w3.org/2010/09/17-au-minutes.html#action05]
<trackbot> Created ACTION-299 - Write proposal on A.1.2.1 non-web based accessible (comment IBM15). [on Sueann Nichols - due 2010-09-24].
<scribe> ACTION: JS to write a proposal on how to test whether a template is accessible. [recorded in http://www.w3.org/2010/09/17-au-minutes.html#action06]
<trackbot> Created ACTION-300 - Write a proposal on how to test whether a template is accessible. [on Jeanne Spellman - due 2010-09-24].
<scribe> ACTION: JR to write proposal Preserve Accessibility Information for B.1.2.4(a) in responseto WCAGWG25 [recorded in http://www.w3.org/2010/09/17-au-minutes.html#action07]
<trackbot> Created ACTION-301 - Write proposal Preserve Accessibility Information for B.1.2.4(a) in responseto WCAGWG25 [on Jan Richards - due 2010-09-24].
This is scribe.perl Revision: 1.135 of Date: 2009/03/02 03:52:20 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/Partial Compliance which is Part A and Part B/Partial Compliance which is Part A or Part B/ Succeeded: s/gues./guest./ Succeeded: s/NPII is not close to implementation. /AL: NPII is not close to implementation./ WARNING: Possible internal error: join/leave lines remaining: <jeanne> Alex Li, Microsoft, has joined the call. Alex discloses that he has not been able to complete the IP rights declaration. No ScribeNick specified. Guessing ScribeNick: jeanne Inferring Scribes: jeanne Present: Andrew Greg Sueann Jutta Jeanne Jan Alex_Li_(guest) Found Date: 17 Sep 2010 Guessing minutes URL: http://www.w3.org/2010/09/17-au-minutes.html People with action items: gp jr js sn WARNING: Possible internal error: join/leave lines remaining: <scribe> Alex Li, Microsoft, has joined the call. Alex discloses that he has not been able to complete the IP rights declaration. WARNING: Possible internal error: join/leave lines remaining: <scribe> Alex Li, Microsoft, has joined the call. Alex discloses that he has not been able to complete the IP rights declaration.[End of scribe.perl diagnostic output]