W3C

Evaluation and Repair Tools Working Group Teleconference

08 Feb 2012

See also: IRC log

Attendees

Present
Shadi, Don, Philip, Samuel
Regrets
Christophe, CarlosV
Chair
Shadi
Scribe
Philip

Contents


New ERT WG Participant

<shadi> Don Raikes

<shadi> http://lists.w3.org/Archives/Public/public-wai-evaltf/2012Feb/0013.html

Causality and EARL

<shadi> http://lists.w3.org/Archives/Public/public-earl10-comments/2011Dec/0000.html

Shadi: any thoughts?

<shadi> earl:cause

Shadi: Proposal is, to have a property called "cause", describing the cause of an error

<shadi> http://www.w3.org/TR/EARL10/#TestResult

<shadi> earl:info

<shadi> earl:outcome

<shadi> earl:pointer

Shadi: "cause" could be a kind of an "info"

<shadi> my:cause is sub-property of earl:info

Shadi: ok to propose this?

PhilipA: Can a sub-property extend the range?
... Or must it use a literal as in "earl:info"?

<shadi> http://www.w3.org/TR/rdf-schema/#ch_properties

<shadi> http://www.w3.org/TR/rdf-schema/#ch_subclassof

Shadi: I think it would keep the range and cannot extend it, but not quite sure. Will look into it.

SamuelM: I'm think it mixes application logic (cause of error) with the report

Shadi: Do we need to have more structure?

SamuelM: from the perspective of webid it could be useful, agreed

Shadi: Are there other use cases than for for webid?
... first question / ACTION: check if sub-property inherits range or if it can change the range
... maybe another approach is to drop the range for "earl:info"

<shadi> ACTION: shadi to check if domain and range of a property are transposed to a sub-property [recorded in http://www.w3.org/2012/02/08-er-minutes.html#action01]

<trackbot> Created ACTION-136 - Check if domain and range of a property are transposed to a sub-property [on Shadi Abou-Zahra - due 2012-02-15].

Shadi: second question: think about other use cases for this

<shadi> Topics: Close open issues

<shadi> http://www.w3.org/WAI/ER/EARL10/issues#overall

<shadi> [[Drop rdfs:Domain and rdfs:Range for OWL constraints to increase reusability of the terms]]

Shadi: 1.) drop rdfs:Domain and rdfs:Range
... 2.) use OWL constraints
... I did ask a semantic web colleague about this
... conflicting comments from the semantic web people
... there isn't any reason to change it
... my proposal to close this item without dropping rdf:Domain and rdfs:Range and withoud using OWL constraints
... any objections?

<shadi> RESOLUTION: close issue "Drop rdfs:Domain and rdfs:Range for OWL constraints to increase reusability of the terms"

<shadi> RESOLUTION: close issue "Use OWL constraints to express conformance requirements"

<shadi> [[Use of prefix "dc" vs "dct"]]

SamuelM: we have two different namespaces, one pointing to dc terms and to dc elements
... in case of dc terms we use the prefix "dct", in case of dc elements we use "dc"
... but we will align to the community

<shadi> http://dublincore.org/schemas/rdfs/#NAMESPACE

<samuelm> http://dublincore.org/documents/dcmi-terms/

SamuelM: but we need to be consistent in the EARL specification

<shadi> http://dublincore.org/documents/dcmi-namespace/

Shadi: "dc" seems to be more widely used as "dct"

<shadi> "dc" usually more than "dct" or "dc-terms"

SamuelM: but "dc" refers to the older specification

<shadi> "dcmi-terms"

<shadi> [[suggestion "dct" or "dc-terms"]]

SamuelM: suggestion would be to use "dct" or "dc-terms"
... in no case use "dc", because it points to the older specification

<shadi> RESOLUTION: use prefix "dc" with namespace http://dublincore.org/documents/dcmi-terms/

<shadi> RESOLUTION: use prefix "dct" with namespace http://dublincore.org/documents/dcmi-terms/

<shadi> http://www.w3.org/WAI/ER/EARL10/guide-issues#serialization

Shadi: EARL can represent TAP and xUnit
... TAP seems to be more line-based with sequentially output
... mabe put in a paragraph in the EARL guide
... clarifying the relationship from EARL to TAP an xUnit

<shadi> RESOLUTION: no overlapping use cases even though all three specifications allow the representation of test results; could mention that EARL results can be serialized as TAP and/or xUnit, and vice-versa, in the corresponding section

Upcoming Meetings

<shadi> 22 February 2012

<shadi> 7 March 2012

Summary of Action Items

[NEW] ACTION: shadi to check if domain and range of a property are transposed to a sub-property [recorded in http://www.w3.org/2012/02/08-er-minutes.html#action01]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.136 (CVS log)
$Date: 2014-02-24 22:45:36 $