Accessibility Minutes 2009 04 01
From MemberWiki
Contents |
Participants
- Ann Abbott (IBM)
- Jon Ferriaolo (IBM) - OAA Director
- Jon Gunderson (University of Illinois)
- Nathan Jakubiak (ParaSoft)
- Preety Kumar (Deque)
- Tony Salcedo (ParaSoft)
- Michael Squillace (IBM)
- Rich Schwerdtfeger (IBM - chair)
- Derk Stegemann (FIT)
- David Todd (IBM)
Minutes
The use of Style
RS: David put a post on the list we should discuss
David: embedding style within HTML
David: WCAG is not clear on whether to include the style attribut or not
David: There is a debate in the WCAG on the use of the style attribute
JG: G140 talks about class attributes
David: They kind of imply not to use the style attribute
ACTION: Preety will check with the WCAG group regarding whether it is illegal to use embedded style
JG: I don't think people use the style attribute alot
David: Some IBM apps use style attribute alot
JF: Many IBM apps use CSS
RS: WCAG 2 is referring to the use of the style attribute, not using CLASS attribute
The Use of bold Tags
RS: If text is wrapped in bold tags, followed by a break, then trigger a warning for use of header, do not text decoration or color to indicate structure
RS: He wants to rewrite this, do not use BOLD unless there is the following markup, many examples
RS: There are a number of mechanisms
JG: I think we should look for heading structure than looking for problems in markup
Preety: I think what we are trying to do find places where they are overlooked a heading
Preety: We find this commonly
Preety: I am not sure of the BR
RS: We should just look for bold?
Preety: BOLD with other things and conditions, it could get complicated
Preety: Depends on how many words are in the BOLD
RS: It sounds like we need to flag something
Preety: They should not be using BOLD at all, but we are looking for possible headings
David: We should give general guidance about how to use or NOT use BOLD
RS: We need to have a section on the use of BOLD
RS: Can you update the section on the BOLD tag?
David: yes
ACTION: David update the wiki section on the use of BOLD
RS: reviewing current wiki entry, what do you mean about class?
RS: When the CLASS attribute includes a "heading" text
RS: Preety is that something that we should be looking for CLASS heading?
Preety: yes
The use of non-DOM API function functions for updating the DOM
RS: If INNNERHTML or other writes should ...
David: WCAG2 recommends using the standard DOM functions
JG: We could warn people, but it will be ignored
Preety: there are security issues
RS: we should tell people why, security aside
Preety: I am not sure structural markup is the concern
Preety: Using non-standard markup you can still create structural markup
David: I will look for why in WCAG
ACTION: David Define why the use of INNERHTML and other non-DOM calls for updating the DOM are an accessibility issue
jakubiak: http://www.w3.org/TR/WCAG-TECHS/SCR21.html
JG: You need to make sure rules make sense
Preety: I will check with the design team
David: This is a level A requirement
Layout and Data Tables
RS: Table markup of layout
RS: role="presentation" is important
JG: we have some huristics for data tables
Preety: It some times get murky when tables are being used for both layout and presentation
RS: David it will be ignored with accessibility API
RS: If you have a TH in the table, what is the significant
RS: If the role="presentation" there should not be any TH elements
RS: For AT everything disappears
Preety: We definitely need to test for layout, the difficulty is mixed use
Preety: then we tell people to split things into 2 tables
RS: If you do any of this stuff you know you have a violation, other stuff gets more difficult
RS: The next one is do we want to enumerate some of the other stuff, is this other stuff trade secrets?
RS: Should we put in the obvious ones?
Preety: Yes, a 1x1
JonGund: University of Illinois Data Table Rules
JG: Our previous table rules were related to nesting
Preety: What about SCOPE?
JG: We did not find any examples where SCHOPE made any difference, but send me some if you have examples
RS: I suggest: If we detect a TD or ROLE=COLUMNHEADER there is not an adequate number of these that we flag an error, with the author needs to determine if it is a data table or a layout table and use the proper markup for that table
JG: I agree
RS: We can help the author with using ARIA presentation
Preety: I think we need a good sample set, before we can determine
RS: I will put in a note about creating a sample set
RS: Suggest using ARIA or HTML markup to data tables and ARIA markup for presentation tables
RS: We can create test cases
Check for Summaries in data tables
RS: Generate an error when a data table is missing a summary attribute
RS: Can you use aria-describedby?
RS: This is what HTML5 group has not decided yet
RS: The summary attribute can be misused
RS: I don't know if summary will be deprecated in the HTML5 or not
RS: Is aria-describedby better than summary?
Preety: yes, AT support is the primary reason
RS: We need to have some rules, link to iCITA rules
David: Is summary compatible with IE 7?
RS: This comes down to AT support, one major AT developer uses the DOM, not accessibility APIs for IE
RS: On Safari they use the accessibility APIs operat on the DOM structure
David: If tables automatically announce when table gets focus, describedby needs extra keystrokes
JG: aria-labelledby seems more like summary
JG: Summary seems like it has a mixed usage, it is used for both short and long
Preety: the caption element is used by ATs and automatically read for data tables
David: JAWS will read aria-describedby with F1
Preety: JAWS reads both the caption and the summary attribute
Use of Captions
RS: Do we want rules for captions?
David: Caption is not mandatory, and it must be in the caption element
Preety: If a data the caption must be present
David: I had the same impression for along time, then I read the requirement
JG: I will check the best practices group
ACTION JG: Add examples to wiki for summary and possibly captions
Progress on Rules Modeling Tool
RS: Michael you are developing a tool, we now run it as a standalone, we are trying to embed it as an applet, FF3 support, make it more flexible for people to use
Mike: What format are the rules going to take?
Mike: I sent a sample to this to the group
RS: I will check with Victor at Yahoo about YSlow people
Mike: For this group we do not want to be YSlow specific
JG: I think that there is a more general rule development with the YSlow people
JG: Meeting this Friday with the YSlow people