Provenance Working Group Teleconference

20 Sep 2012


See also: IRC log


pgroth, Luc, TomDN, MacTed, jun, [IPcaller], +1.315.330.aaaa, tlebo, christine, dgarijo
James_Cheney, Curt_Tilmes, Ivan_Herman
Paul Groth
Tom De Nies


<trackbot> Date: 20 September 2012

<pgroth> Scribe: Tom De Nies



<pgroth> http://www.w3.org/2011/prov/meeting/2012-09-13

pgroth: approval of last week's minutes

<pgroth> Minutes of the September 13, 2012 Telecon


<Dong> +1

<jun> +1

<pgroth> accepted: Minutes of the September 13, 2012 Telecon

Charter Extension

pgroth: we need scribes, please sign up in advance
... We are close to the end of the charter, but our extension request was approved

<pgroth> http://www.w3.org/2011/01/prov-wg-charter

pgroth: initially, we wanted it up to April/May, but we got an extension to september 2013
... However, we're still aiming to deliver everything according to the schedule agreed at F2F3

<pgroth> +q

pgroth: We will make an updated timeline some time next week


pgroth: Last week, we had an action to make an FAQ on PROV

<pgroth> http://www.w3.org/2001/sw/wiki/PROV

pgroth: Among others, the use of dc:hasPart for sub-activities is in there
... the WG is invited to populate this FAQ

<pgroth> +q

pgroth: This is at the Semantic Web activity wiki, which means that we can keep updating this, even after the WG is finished


pgroth: There were a couple of issues that we didn't resolve last week, perhaps Stephan can go over them

<pgroth> http://www.w3.org/2011/prov/track/issues/493

pgroth: Where are we in issue 493?

zednik: Right now the schema reflects the DM completely
... We didn't want to move away from the DM, so we haven't made any changes in the schema
... We use RDF types, to respresent type information, as it is the most natural/intuitive way
... We are still discussing.

pgroth: So each serialization can have its own type. Do we all agree to this?

zednik: Not sure, the telecon on monday was too small to tell if everyone agrees

Luc: We don't want to be too restrictive for these types
... to avoid overconstraining things.
... In the ontology, we use RDF types, which is fine for the RDF mapping, but I'm not sure what we should use in XML
... It would be up to the translator to determine this.

pgroth: Seems like a good idea, except that it would be problematic to convert PROV-XML to RDF, if the types don't agree

<Luc> in xml, we have prov:type and not xsd:type

<Luc> in xml, we have prov:type and not xsi:type

zednik: We don't have a complex type for agents that are also entities
... e.g. "this is an entity with type agent"

pgroth: So the resolution would be to leave things as they are: loose like in the DM.
... and leave typing up to the implementer

Luc: elaborates on "in xml, we have prov:type and not xsi:type"
... which is an xml attribute

zednik: so you can only have one?
... That is not the case in the DM or ontology

<pgroth> proposed: leave prov-dm to allow for loose types and each serialisation should define their own type system

<pgroth> proposed: leave prov-dm to allow for loose types and each serialisation should adopt their own type representation

Luc: it would be better to say "should adopt their own type representation"


<MacTed> +1

<zednik> +1

<Dong> +1

<tlebo> +1

<pgroth> accepted: leave prov-dm to allow for loose types and each serialisation should adopt their own type representation

pgroth: that resolves issue 493

<pgroth> http://www.w3.org/2011/prov/track/issues/551

issue 551?

zednik: Curt made some changes to the schema that resolved this

<Luc> I think it was a good solution to introduce this documentElement

zednik: It basically flattened out the schema

<zednik> div: http://dvcs.w3.org/hg/prov/rev/b88f0f02344f

zednik: so you don't have a separate section for the relations

pgroth: Were there any objections?

zednik: no

Luc: There was some refactoring of the schema that took place. All the attributes were placed in a single group "commonattributes"

zednik: We havent been able to talk about it in the group, so we should probably raise an issue about it

<pgroth> +q

sorry i missed that last

pgroth: I think we did have a resolution about how close you have to get to the DM in a serialization
... If we favour something that's "natural" for RDF, we should do the same for XML

zednik: Since you can have attributes on almost anything, we grouped it as such
... I think it's doable to revert, although it unsimplifies the schema.

<Luc> it's already done http://dvcs.w3.org/hg/prov/rev/7c238aba1d07

Luc: the group commonattributes has been removed from the schema

pgroth: So how far are we away from a draft document?

zednik: So how ready/stable is the schema?
... Well, Curt and I have been going through the issues. Most are resolved now, so it is pretty stable.

<Luc> there is still the type of identifiers to address

Zednik: So we are getting close to being able to start on the draft of the Note

pgroth: OK, we'll check back next week

Luc: Will there be a call on monday?

zednik: Curt can't make it, but I can, and will send out an email to ask if it's possible.

Resolving Public Issues

<Luc> http://www.w3.org/2011/prov/wiki/ResponsesToPublicComments

pgroth: We need to clarify who is responsible for response to public comments.
... My tendency would be to have me or Luc do it.
... Any other suggestions?

tlebo: It would be best if we work at the response on the weekly telecons, and then consistently have the chairs formulate the response to the commenter

<christine> +1

pgroth: Any objections to this?

<tlebo> note: "formulate" -> "provide", naturally, the group would be formulating the response in our usual way.

pgroth: So it's agreed. Paul will handle the responses

Luc: There were no objections to the resolutions to the following issues:
... (Lists the issues)


<pgroth> accepted: the suggested resolutions in http://www.w3.org/2011/prov/wiki/ResponsesToPublicComments to ISSUE-532, ISSUE-525, ISSUE-507, ISSUE-504, ISSUE-503, ISSUE-447 were accepted as responses by the working group. there were no objections to the resolutions on the mailing group only support

Luc: I'm happy to take feedback on the proposed responses now, but I will also follow the usual protocol, and ask for feedback on the mailinglist
... If there's no objections by next Tuesday, those will be our responses.


<dgarijo> thanks for the links, I'll have a look at the responses.

Luc: We have addressed about 10 of the 35 issues about the DM. I'm planning to draft responses next week

Prov-o issues

pgroth: Last week, on Monday, we tried to clean up some issues in PROV-O, and defined some actions
... We will walk through them here

<pgroth> http://www.w3.org/2011/prov/track/issues/491

<Luc> stian?

pgroth: Stian was going to handle ACTION-107, I will follow up on this with him

<tlebo> the trig example.

issue 479?

<dgarijo> http://www.w3.org/2011/prov/track/issues/479

<MacTed> issue-479?

<trackbot> ISSUE-479 -- cite TriG for examples -- open

<trackbot> http://www.w3.org/2011/prov/track/issues/479

scribe: Satya was going to take ACTION-108. No changes have been made yet, will follow up.

(tnx macted, forgot the hyphen)

pgroth: Next one is ACTION-109 on issue-349

Jun: I took over David's actions
... and ACTION-109 is done.


<trackbot> ISSUE-349 -- examples for each term in cross-reference section -- open

<trackbot> http://www.w3.org/2011/prov/track/issues/349

pgroth: There were two more actions for this issue: ACTION-110 and ACTION-111 for Satya and Stian

<dgarijo> we reviewed the examples to check that they were updated with the latest DM

tlebo: I've seen no changes to the examples yet.
... also. ACTION-112 was on the same issue

<pgroth> daniel?

pgroth: last one was on Daniel, ACTION-113 about issue 446

<dgarijo> I have lost connection


<trackbot> ISSUE-446 -- prov:involvee not documented in PROV-O -- open

<trackbot> http://www.w3.org/2011/prov/track/issues/446

<tlebo> daniels' action 113 is "prov:involvee not documented in PROV-O"

<dgarijo> I have restarted the discussions

<dgarijo> with Kai, Simon and Michael

<dgarijo> we plan to address his comments next week

pgroth: Tim, do you need help on closing any other issues?

tlebo: Yes, could use help with issue 461


<trackbot> ISSUE-461 -- provo cross reference inadequate in printed form -- raised

<trackbot> http://www.w3.org/2011/prov/track/issues/461

scribe: Would be nice if someone could take over this one.

pgroth: Someone needs to talk to Graham to identify what he needs

jun: I can do that

<dgarijo> +q

pgroth: It would be good to know all the remaining issues on PROV-O


tlebo: 349 is closed
... Not sure about 446, 479 and 491

daniel: What happened with the issues about the figures?

tlebo: There's no formal issue, but it is being looked into
... now waiting on review of the changes made in the draft

<pgroth> http://www.w3.org/2011/prov/track/issues/552

pgroth: 552 was an external comment about subclasses

<Luc> issue-552?

<trackbot> ISSUE-552 -- Check subclass definitions in prov-o -- raised

<trackbot> http://www.w3.org/2011/prov/track/issues/552

pgroth: Is this a DM issue or PROV-O issue?

<Luc> issue-523?

<trackbot> ISSUE-523 -- Data Model Section 5.3.5 -- open

<trackbot> http://www.w3.org/2011/prov/track/issues/523

<pgroth> http://lists.w3.org/Archives/Public/public-prov-comments/2012Sep/0000.html

tlebo: seems like a problem with the interpretation of the DM, so it is a PROV-O issue

<christine> Thank you Paul, Luc and everyone. Apologies, I have to leave to chair a call starting on the hour.

Luc: I think there's a general problem about inheritance that we may want to revisit in the DM
... , illustrated by issue 523, about influence
... In the constraints, we say that IF wasDerivedFrom(b,a) THEN wasInfluencedBy(b,a)
... So we may want to opt for inheritance, which is what PROV-O does
... and it works well for most ontologies I think. However, I am not sure if we'd also want it in PROV-XML
... Would be nice to have feedback here.

<tlebo> +1 to what luc said

pgroth: Would that only apply to influence? or all of the DM?

Luc: I'm focusing on influence, currently

<tlebo> does "influence" == "relation" ?

pgroth: My only issue with that is: whether or not we would change the UML in correspondence with that
... It may not be normative, but developers do use it for their implementations

<jun> [have to go now. will catch up on the minutes]

Luc: any subtype of influence can be used in any position of influence.
... Perhaps "interface" would be better than inheritance
... to indicate that all subtypes are all "influence"

<Luc> issue-552?

<trackbot> ISSUE-552 -- Check subclass definitions in prov-o -- raised

<trackbot> http://www.w3.org/2011/prov/track/issues/552

pgroth: For now, this issue should be considered together with issue-523

Luc: There's another issue in the DM, dealing with inheritance
... So we should tackle these as a set.
... Stephan, working with the XML schema, have you considered extension of types?

zednik: It has come up, but we haven't looked into it. It should be done after the schema is stabilized.

<dgarijo> Sure

pgroth: Given the time, we will save the topic on notes for next week.

ok, tnx!


<pgroth> rrsagnet, set log public

<pgroth> trackbot, end telcon

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.136 (CVS log)
$Date: 2012/09/20 16:08:18 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.136  of Date: 2011/05/12 12:01:43  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/Thursday/Tuesday/
No ScribeNick specified.  Guessing ScribeNick: TomDN
Found Scribe: Tom De Nies
Default Present: pgroth, Luc, TomDN, MacTed, jun, [IPcaller], +1.315.330.aaaa, tlebo, christine, dgarijo
Present: pgroth Luc TomDN MacTed jun [IPcaller] +1.315.330.aaaa tlebo christine dgarijo
Regrets: James_Cheney Curt_Tilmes Ivan_Herman
Agenda: http://www.w3.org/2011/prov/wiki/Meetings:Telecon2012.09.20
Found Date: 20 Sep 2012
Guessing minutes URL: http://www.w3.org/2012/09/20-prov-minutes.html
People with action items: 

[End of scribe.perl diagnostic output]