Provenance Working Group Teleconference -- 29 Mar 2012

<trackbot> Date: 29 March 2012

<pgroth> Scribe: Daniel Garijo

<dgarijo> scribe: dgarijo

yes

<pgroth> Minutes last telco: http://www.w3.org/2011/prov/meeting/2012-03-22

<Curt> +1

<Tom_De_Nies> +1

<tlebo> +1

<jcheney> +1

<Paolo> +1

<pgroth> Approved: Minutes of the March 22 2012 Telecon

pgroth: open actions
... there is still 1 open, on me
... I think the action is done (new version of the paq)
... next topic: reminder about the scribes
... welcome to Tom De Nies

tom: colleague of SamCoppens
... work in enrichment of news and provenance assessment of news. Very happy to be here

paul: questions?

PROV-DM

pgroth: stauts update?

<Luc> http://www.w3.org/2011/prov/wiki/ProvDMWorkingDraft5

<MacTed> it would be VERY helpful if links to relevant docs can always be included in (added to) the Agenda

luc: We've been working quite hard on the model. The documents are ready today (deadline is monday). In the wikipage there is a to do list and most of the items are covered.

<Paolo> "align collection productions [PM]" done, actually

luc: invalidation will not be covered. It will be written and merged separately.
... for provn we've done checks for everything except collections

<Paolo> @Luc collections productions done this morning

<sandro> sorry, previous meeting ran late.

luc: constraints part has not been started.
... will try to have it for monday, although we are not sure.

pgroth: questions?

<Paolo> :collection (paolo) -- fix the constraints to allow for multiple contained() assertions" --> please see my mail this morning re: this

tlebo: curious about the "finished today but still something to do for monday"

<Paolo> (have to go, sorry in between trains)

luc: part 2 is the one we are going to start tomorrow.

<tlebo> (DM and N are finished today, CONSTRAINTS is for Monday.)

luc: we have identified a number of questions for reviewers:
... can the doc be released as a fwd?
... blockers?
... reviewers with pending issues please confirm if they can be closed
... typos gramatical issues can be submitted in issues 331, 332 and 333

paul: reviewers for provd and provn can start reviewing now?

luc: they could. But there are a few tweaks we'll have to do once we reach the constraints on part to. Up to the reviewers.

<tlebo> Monday is fine :-)

pgroth: reviewers were: Tim, khalid, curt, jun and...

luc: if you are in agreement the we can send a notification on monday

<GK> I'm not seeing the document URIs at the page posted previuously

<Curt> I can't really do it until next week anyway

pgroth: ok, so wait till monday to begin the review.

<tlebo> http://www.w3.org/2011/prov/wiki/Meetings:Telecon2012.03.29#PROV-O

PROV-o

tlebo: different parts of the team working on different parts of the doc.
... some questions to the reviewers are on the link posted in the irc
... will be ready for review by monday

<Luc> where is the prov-o.html document?

tlebo: the organization scheme has changed

@Luc: in aquarius :(

<Luc> thanks

tlebo: is it enough to understand everything, are the cross references useful?
... please start to be familiar with the ontology.
... should we add more examples.

pgroth: would you move this to the w3c site (what people are supposed to review).

tlebo: yes, it will be a save as in mercurial

pgroth: could anybody else volunteer to review?

<SamCoppens> I will review it also

<GK> (I may, but reluctant to over-commit right now)

luc: macTed?

macTed: I'll try to have a look

<Luc> @paul, could you remind us the timing for review

paul: so sam an macTed as new reviewers.

<pgroth> april 9th

pgroth: monday is the release, April 9th is the due date for the reviews.
... will try to reach consensus on 12 or 19th depending on the feedback. On monday we'll send an email to remember everyone

<Luc> they will be emailed on Monday

GK: Can I ask for confirmation of the URIs of the docs to be reviewed.

pgroth: will send them on monday, when they are ready to be released.

Prov-Primer

pgroth: simon couldn't be here, but sent an update: Comments by Stian and Paolo are addrssed.
... he will have the rest by monday (examples)
... he might be postponing collections.
... questions for the reviewers: The primer doesn't talk that mucho about qualified involvement. What is the opinion of the reviewers about that? Can that be left for a later draft?
... blocking issues?

luc: how do we talk about roles withouth qI?

<Luc> yes

<stainPhone> time can be mentioned wirh location and custom attributes

<stainPhone> +1

paul: that might be the question he's asking to the reviewers.

scribe: so sounds reasonable to delay the discussion about collections in the primer.

luc: agrees.

pgroth: I will send an email to Simon.

PAQ

<stainPhone> and I am Stephan are doing the prov-o mapping of collections.

<pgroth> sure

GK: I made a past to the doc simplifying sections.
... still to discuss some of the changes with paul.

<pgroth> updated version is here: http://dvcs.w3.org/hg/prov/raw-file/2b3bd4e1beb1/paq/working/prov-aq.html

GK: there are still issues to resolve in the doc.
... unify the namespace. Added an appendix. Issues like 70, 211 have been addressed. 233, 76 too.
... should we be changing the prov:entity uri with sime other thing?

paul: we need to follow up some of these issues.
... how to use sparql and linked data.

<MacTed> mute would be appreciated on clacky keyboard

paul: with prov.
... people wanted to retrieve the prov of entities and activities

@graham, sorry, I couldn't hear :(

<Paolo> akim, ??P0 is me

paul: what do the gorup thinks about that?

<stainPhone> provElement or similar?

MacTed: distinction between entity and activity is artificial.
... we're building ourselves problems.

luc: paul mentioned activities and entities. There are a lot of terms with ids (nearly all of them). I don't think it is artificial

MacTed: however it is a redefined type. Entitiy has a common usage in the world, and we are using it in a total different way. The struggle to keep that in a different way is the thing that I don't like

<Luc> but this would be teh case for any word we use, wouldn't it?

<Curt> the concept represented by our term entity is inherently complex

pgroth: for the paq we need a term that represents everything with an identity in the model. Maybe provenance element, or ideantifiable.

<stainPhone> "about" ? ;)

<GK> Ah...

MacTed: it continues to be confusing for us all.

<Zakim> GK, you wanted to ask if any alternative terms have been suggested?

GK: has an alternative name been suggested?

pgroth: maybe we should go back to the paq itself
... we will release it next week for people to look at it

<Luc> @GK: I agree that one can always propose better names, but for now, we don't have any other suggestion

<GK> @Luc hence my Q :)

luc: what is your timetable. Synchronized release?

pgroth: I don't think so.

<Luc> we could release it shortly afterwads

pgroth: it is almost there, but almost everyone is commmitted to reviewing something. And also, it is somehow separate.

<GK> Not so much "separate" as "orthogonal" ?

pgroth: next week we'll have something when people have time to review

GK: I may be travelling next week. Limited internet connection access.

Namespace Unification

<pgroth> http://www.w3.org/ns/prov#

pgroth: last week we talked about the ns unification. The use of the hash
... is this use violated in rdf/xml serializations?
... sandro contacted the xml working group
... they encouraged us to have a common namespace
... the one proposed is fine with xml
... it doesn't really matter what you have.

<GK> I'm happy with that resolution.

<pgroth> Propose that we use http://www.w3.org/ns/prov# as the common namespace for all prov specs

<GK> +1

<Curt> +1

<tlebo> +1

<jcheney> +1

<sandro> +1

<zednik> +1

<SamCoppens> +1

<StephenCresswell> +1

<jun> +1

<MacTed> +0

<Tom_De_Nies> +1

<pgroth> Accepted: use http://www.w3.org/ns/prov# as the common namespace for all prov specs

<pgroth> Close ISSUE-224 http://www.w3.org/2011/prov/track/issues/224

pgroth: we can close issue 224
... provide an html index to link all the docs toghether.
... are there volunteers for this?
... I'll go ahead and start doing it after monday
... I'll need some help

luc: regarding collections: 3 relations are being proposed.

Collections

luc: do we want to have the same form (in the past) as the o ther properties?

it would be nice to have some input

<GK> What's the anticip[ated context of use?

paolo: any time you want to track the prov of the collection.

<jun> do we have a link to the proposal?

<GK> SO, if it's for tracking provenance of collections as well as members, I think using the same form is appropriate.

paolo: for workflows is more important (to know if an element of the collection belonged to an execution, etc)

paul: I want to come back about the name in the PAQ for entity URI
... any suggestions would be helpful

luc: why not resource?

- Provenance element? provenance resource?

gk: It was because we wanted to have a way of talking about the views.

paul: the scruffyness of the dm allows us to refer to resource.

gk: I'm trying to see what are the implications about this. I'm not sure
... I would have to check the rest of the document. Maybe works.

luc: It is a reasonable suggestions

pgroth: the distinction between entity/process is more a dm problem.
... you could raise an issue, MacTed, and propose other names.

Luc: do you have suggestions for other names?

<Curt> we went through other terms people didn't like.... They don't like entity either, but it was left standing after other terms were eliminated.

Luc: that was the best one we came up with. We used to have Bob (and we don't want to go back there).

<pgroth> +1 curt

Luc: the only way to move forward is through new proposals

MacTed: I don't think we are the first people to look at these questions.

pgroth: the conclusion is: If you have a better name for consideration, feel free to propose it.

<Zakim> GK, you wanted to say I think I can see a way to revising PAQ 1.2 to remove "entity" and just talk about "resource" without losing the essence of the material

GK: we can eliminate the term entity without loosing much in the doc.

pgroth: good bye everyone
... and good luck for monday

<pgroth> daniel I'll do the minutes

ok, thanks!

bye!

<Tom_De_Nies> bye!

<pgroth> trackbot, end telcon

- DRAFT -

Provenance Working Group Teleconference

29 Mar 2012

Attendees

Contents

PROV-DM

PROV-o

Prov-Primer

PAQ

Namespace Unification

Collections

Summary of Action Items

Scribe.perl diagnostic output