Accessibility Minutes 2009 08 19
From MemberWiki
Participants
- Anne Abbott (IBM)
- Jon Gunderson (University of Illinois)
- Suanne Nichols (IBM)
- Nathan Jacobiak (ParaSoft)
- Vio Onut (IBM)
- Michael Squillace (IBM)
- Rich Schwerdtfeger (IBM - chair)
- David Todd (IBM)
- Sandy Foltz (University of Illinois)
Minutes
Rich: Code repository set up at Source Forge - refer to email for instructions
Sandy: JQuery is a lot slower than XPath
Nathan: how does Javascript fit in with XPath?
Mike: was using Webelo
Mike: what do tools vendors need?
Jon: looking at Firebug with XPath
Jon: IE6 doesn't support XPath
Nathan: still unclear where using XPath
Rich: analysis of page/walk the dom/analyze elements using JQuery or XPath to analyze the dom
Rich: how to implement JSON rule set? Is using XPath an issue?
Nathan: probably fine for both IE and FF - will have to test
Nathan: question for Sandy - using XPath, how long does it take?
Sandy: has not timed, but noticable difference from using JQuery. Simple pages are almost instantaneous, complex pages are few seconds using approx 40 rules
Rich: need to make decisions on which rules to execute when focus changes
Nathan: agrees need to limit on events
Jon: IE fires focus events before element gets focus/FF fires after
Rich: watch for activedescendent also
Vio: currently uses XPath to access DOM
Rich: trigger rules once page load settles and a subset of rules on focus change
Jon: I agree
Consensus: We will use XPath over JQuery
Rich: also focus changes on activedescendent inside a grid
Rich: activedescendent doesn't cause dom focus event change, but at OS level
Jon: add dom mutation event sync'd with kbd event
Rich: look at this issue further
Rich: what about liveregions?
Jon: group needs to lobby browsers for way to enumerate event handlers
Jon: important in order to identify interactive widgets
Jon: anyone using HTMLUnit?
Nathan: is using
Jon: takes a long time to get change incorporated
Sandy: using YSlow library
Sandy: engine originally based on YFlow with mods, manages and executes rules
Sandy: runs under Firebug
Mike: will have to write something more generic
Nathan: not sure how much they can use beyond rule set
Rich: has to be as generic as possible
Vio: would be able to use rules and need to be as generic as possible
Sandy: that's what we're doing now
Jon: long-term goal is to have report generated on server, then be able to send URL so anyone can access report
Rich: Nathan & Vio - can you determine how portable rules are for your framework, then write some of the rules?
Nathan: will start looking into that
Rich: next talk about reporting best practices
Rich: Nathan & Vio - post input to wiki
