ERT WG

SAZ: If no objections then we'll publish.

JK: Will it be published with HTTP notes?

SAZ: While in review period we can publish HTTP notes.

JK: Notes seem to be finished so we could publish now.

SAZ: Only an internal draft of notes is available. Need more editorial notes and fit into format.
... Nothing missing in schema.

resolution: no objections so far so we'll publish EARL 1.0 schema working draft

(review of issues)

SAZ: proposal - max cardinality of 1, properties 'assertedby' and 'result'

JK: can live with proposal

CV: not so sure
... we have no restriction on cardinality

CR: can live with proposal

SAZ: proposal - close issue
... (close issue in chapter 2.2 "foaf:Person is redundant as it is a subclass...")

(discussion on earl:TestSubject)

JK: We need to be able to record location of subject on local system - not just HTTP

<JibberJim> sure - I don't think we can go out of our way to prevent all conflicts in spec, we have to expect common sense in our users

SAZ: proposal - we adopt uri:uri into test subject
... suggest we leave as is and discuss on email
... another topic that came up from email - do we want to store source of document within report?

<JibberJim> I don't think we need a method right now to store source, it would be easy for people to add

SAZ: Perhaps add to EARL guide doc?
... Do agree with Jim (not an immediate need case)
... Any agreement that this should not be in schema but perhaps add info to guide doc?

<JibberJim> no need to extend schema

CR: no need to extend schema

SAZ: some agreement that we should add to EARL guide
... proposal - Consider maximum cardinality restrictions for earl:validity, and earl:confidence

(agreement)

SAZ: proposal - Consider dropping confidence level

(discussion but no agreement)

<shadi> possible directions: 1. keep it as is; 2. drop it; 3. define it as a numerical value (percentage?)

<JibberJim> percentages don't make it work - chaal's 70% is going to be different to my 70%

<JibberJim> oh yes live with 1

<shadi> CV: vote for 1

<shadi> JL: vote for 2, can live with 1

thinks we should drop it. Confidence is more closely related to the test.

<JibberJim> er other way around?

<carlosV> no shadi, CV: vote for 2

<shadi> JL: vote for 1, can live with 2

CR: vote for 2 but can live with 1

JK: votes for 2 but can live with 1 - we need clear use cases if kept.

SAZ: we need to think more about min/max cardinalities for properties