Evaluation and Repair Tools Working Group Teleconference

08 Apr 2009


See also: IRC log


Johannes, Shadi, Mike, CarlosV


Where we are, where we are going


MS: is this still draft?

SAZ: not yet approved by AC

MS: we are behind, were supposed to public Last Call for the EARL Schema
... also need to publish first draft of the EARL Guide
... also the Requirements document

SAZ: there are change requests for the requirements document in the f2f minutes and in tracker

MS: need to get started on test suites to catch up
... what are the plans?

CV: working now on it

SAZ: if we publish EARL Schema WD by mid-end April
... we could aim for a the big Last Call publication for beginning of June
... that means EARL 1.0 Schema + HTTP-in-RDF + Content-in-RDF + Pointers-in-RDF
... is this realistic?

JK: doing on-going changes

SAZ: forgot, EARL 1.0 Guide should be part of the "big publication"

CV: should be possible

JK: am on vacation during second half of May

SAZ: if you can get the changes done before then it would be useful
... most worried about the Guide
... back to test suites, what do we need to host them?

MS: generated tests serve as test suites?


SAZ: could we provide test results for the WCAG 2.0 Test Samples?
... would provide the according tests (for the context)
... but adds a dependency

CV: mixing two objectives
... can provide examples of valid and invalid EARL reports

MS: need to describe how the test suites should look like

CV: EARL does not require accessibility tests

SAZ: instead of inventing test criteria, my thought was to reuse the WCAG 2.0 Test Samples

EARL 1.0 Schema - prepare for publication



<cvelasco> typo in related instances earl:cantTell

SAZ: removed OWL namespace and fixed namespace for DC elements and DC terms
... added instances for OutcomeValue in Section 2.7, Appenix A, and the RDF file
... need to discuss these changes

CV: wording for Related Instances is too weak

JK: don't think it's too weak

<JohannesK> JK: change "can not" to "cannot"?

MS: maybe we get a question why we didn't do this for Test Mode

JK: last week we discussed that Test Mode doesn't need subclassing

MS: isn't that clear to me

SAZ: how about adding an editor's note in section Test Mode to ask for feedback?


JK: isn't "cannot" one word?

MS: yes, one word

RESOLUTION: Shadi to change "can not" to "cannot"

CV: don't like acronyms in upper case for instance names

SAZ: instance names were not discussed
... it is generally not considered good practice to differentiate between to entities using just the casing

[agreement to "passed", "failed", and "cantTell"]

MS: how about earl:na and earl:nt (lower case)?

<MikeS> other option sfor names is 'inapplicable' and 'untested'

RESOLUTION: Shadi to change "NA" and "NT" to "inapplicable" and "untested"

MS: everyone set?
... other issues?

SAZ: in this publication we are looking for feedback on: 1. use of foaf:Document, 2. instances for TestMode (like OutcomeValue), 3. replacing earl:Software with DOAP, 4. conformance section

<scribe> ...pending adding these 4 questions to the "Status of the Document", and making the two changes discussed above, any objection to publication?

JK: range for earl:info is Literal or XML:Literal?

MS: think Literal

JK: should be RDF namespace?
... actually RDFS

<JohannesK> <http://www.w3.org/TR/rdf-schema/#ch_literal>: rdfs:Literal

RESOLUTION: Shadi to change range of earl:info to rdfs:Literal

<JohannesK> So it should be <http://www.w3.org/2000/01/rdf-schema#Literal>

RESOLUTION: publish EARL 1.0 Schema as an updated Working Draft pending the three changes recorded above

next and future meetings

SAZ: regrets for the 15th

MS: next week discuss EARL 1.0 Guide and Requirements document
... next meeting *15 April*

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2009/04/08 13:48:10 $