SAZ: talks about CSUN
about EARL business case
SAZ: why should somebody implement EARL?
... updated accessibility logos
... add EARL report to logo
... some tool developers are not yet convinced
CR: conformance claim with EARL would be
... indicate what has been checked during test
which Andrew? AA or AWK?
<Zakim> wendy, you wanted to say, "concern i heard was about making more earl"
WAC: rise awareness of EARL
... chicken and egg
... like little RSS logos
CR: link to validator like Markup validator
<AWK> AWK: pondering what interests may exist that actively don't want a more comprehensive logo with EARL. If such exist we'll need to figure out how to counter
SAZ: updated test results like markup conformance claims
AWK: we cannot solve this problem
... in Firefox you have a button in the GUI for RSS
SAZ: add logo use case to use case scenarios
<AWK> AWK: it would be interesting to have an analogous item (to the firefox RSS button) for accessibility claims
<Zakim> wendy, you wanted to ask, "could you have valid earl that is not valid rdf? just use rdf validator?"
CR: accessibility check linked to logo, not validated EARL
SAZ: does RDF validator validate against RDF
... don't think so
WAC: no validation against RDF schema
<shadi> ACTION: johannes add another use case / scenario for EARL as page profile [recorded in http://www.w3.org/2005/03/22-er-minutes.html#action01]
JK: does Jena validate against RDF schema
SAZ: not by default
<wendy> validating rdf parser: http://18.104.22.168:9090/RDF/VRP/index.html
CR: use xpath for documents that don't
... fuzzy way of identifying things for changing documents
... need context, not only current element
... are there any more use cases?
CA: would xpointer be sufficient for context information?
CR: xpointer will be invalid for changing
... will post more details about fuzzy techniques
SAZ: different checkpoints have different
... pointing and persistency are interrelated
... xpointer only work on XML documents
CR: only need valid documents
WAC: what about modified DTDs?
... XHTML is not required
... publicly available schema/DTD is required in WCAG
SAZ: we cannot assume validity or even
well-formedness because this is one of the checks
... use IDs?
GB: multi-level way of locating
... fuzzy pointers based on various techniques (xpointer, fuzzy, IDs)
CR: good idea
... different strategies needed
SAZ: two dimensions: 1. locating strategy, 2. tests (context sensitive/insensitive)
CI: problem is what are we identifying?
... most content is dynamic
CR: cache image somewere? is impractical
JK: image changes, image URL and context remains the same
GB: use more properties: MD5 hash
SAZ: any time someone proposes a solution,
aonther talks about exceptions
... objects/applets, CSS background?
... visual/markup structure
... tables with captions may be moved within a documents, images may not
... for different problems we need different solutions
CA: consistency over time, consistency within context
SAZ: change overtime is very important
CA: context: does moving an element change accessibility status?
CA: store document fragment structure around
... group of links, provide skip-over
... need for structure information
... skip-over link use location of first link
CA: use only one location for each accessibility problem?
... we need to keep in mind
... information about dependencies is in the test case
WAC: use context information in xpath expression
SAZ: how to help web developers location of
problem? line/column number?
... complicated xpointer? not developer-friendly
... message oriented location? machine-readable thing?
... homework: top 3
... next meeting is right after easter, how many people can be present?
Discussion to be continued on next call
Discussion postponed to next call
<shadi> rrsagent make minutes