Accessibility Minutes 2009 09 02
From MemberWiki
Participants
- Anne Abbott (IBM)
- Jon Gunderson (University of Illinois)
- Nathan Jacobiak (ParaSoft)
- Michael Squillace (IBM)
- Vio Onut (IBM)
- Rich Schwerdtfeger (IBM - chair)
- Sandy Foltz (University of Illinois)
Minutes
Mike: name spaces, Rich had action item to contact John about name spaces, there is a guideline to follow for name spaces
Mike: name spaces are in Open Ajax registry. It's a JSON object. Mitigates names space collisions There's a name space review process but that's old and they don't use the process anymore. Groups will now register their own name spaces.
Mike: We have to decide what our name space is going to be. Consist of identifier and URI - team to think about what we want to use.
Mike: "Access" name space could imply other things like security
Mike: Does the group have any ideas?
Nathan: agrees that "access" could mean many things.
Jon to Sandy: Any work on FF accessibility extension that could be used as guideline?
Sandy: No
Sandy: Are we going to sub divide the name spaces?
Mike: Yes. We want to have sub name spaces for different parts of our work.
Jon: wants name space to accommodate code
Mike: that makes sense to have name spaces to accommodate code.
Team: agrees accommodation of code is a reasonable thing to do.
Mike: Team should think about name space ideas and send to list for discussion.
Mike: No hurry to have a name space right now. However, once we do a little more work, we want to have a name space.
Vio: Regarding name spaces and code sharing: Sharing rules/code will be hard because their product is commercial.
Vio: No strong opinion on name spaces name.
Mike/Sandy: name spaces to accommodate environments such as Mozilla & IE specific code
Mike: moving back to agenda item 2.
Mike met with sub-group to discuss rules
Mike: Meeting discussed comparison and contrast of IBM HA&AC rule objects with UIUC rule objects
Mike: common attributes: rule id, message, rule logic
Mike They discussed locaization
Mike: Dojo has different approach to i18N than Mozilla
Mike: Discuss rule sets - containers for rule objects.
Mike: Rule set will determine severity of violation
Mike: Rule set will determine whether violation is under WCAG, 508, etc.
MIke: Two ways to do reporting. One way is to execute all rules, then filter report base on rule sets that are selected. Nathan was worried about performance. If a rule set is not activated it won't be executed.
Vio: Question about rule sets? Can a rule belong to different rule sets? WCAG 1.0 and WCAG 2.0
Sandy: Yes it is possible. People get upset when too many errors are displayed. So good not to repeat errors.
Mike: Agree, for example his experience is that he has a lot of possible errors related to images and they very similar.
Rich: What rules get executed. Should we have a table of rules?
Mike: A number of rules will be executed by onload but others will be executed by onfocus.
Mike: Messages must be keys to support i18N.
Mike: Rule logic contained in rule object itself or, to promote more JSON-compliant docs, have a pointing mechanism to separate modules and functions. Simple functions could easily be included, more complex might be externalized .
Vio: Rules must support customization.
Nathan: Any time heuristics are used, users may want to customize.
Mike/Sandy: Mike will work with Sandy on code merge.
Sandy: Need to have a place to document rules.
Sandy: defines rule group as a set of related rules.
Mike: Will have three categories of rules: rules, rule groups (related rules), rule sets (collections of rules)
Nathan: Prefers not to confuse the user with rule groups vs. rule sets.
Vio: agrees with Nathan
Sandy: Some rules are run on a page level and some are run on the element level. That's why you want to have rule groups.
Vio: Maybe we can accomplish what Sandy is talking about with rules tags.
Mike: His rules are grouped according to the element type.
Vio: Suggested rule groups should be created dynamically with tags.
Vio: There should be a default set of tags.
Vio: Tools may add more tags.
Nathan: Likes the idea of grouping for optimization
