Provenance Working Group Teleconference -- 02 Jun 2011

<trackbot> Date: 02 June 2011

<Luc> Scribe: Paolo Missier

trackbot, start telcon

<Luc> paolo, it's all set up

<trackbot> Meeting: Provenance Working Group Teleconference

<trackbot> Date: 02 June 2011

do I need to do anything at all now?

<Luc> nothing, we just need to wait for start of call

sweet

<Luc> Hi Stephen, welcome!

<tlebo> what is the phone listing command?

<dgarijo> http://www.w3.org/2011/prov/wiki/Meetings:Telecon2011.06.02

<frew> +1 minutes

Luc: accept minutes from previous confcall?

<dgarijo> +1

<DavidSchaengold> +1

<tlebo> +1

<dcorsar> +1

<kai> +1

<Yogesh> +

<iker> +1

<Edoardo> +1

<jorn> +1

<GK> abstain (was present but not in audio)

<zednik> +1

<JimMyers> +1

<smiles> +1

Luc: minutes accepted

review of actions

Luc: invited experts -- not all experts on board yet
... calling for Sandro but he's not responding
... apologies for delay

<frew> someone keyboard is drowning out the speakers

F2F1

<Luc> http://www.w3.org/2011/prov/wiki/F2F1

<khalidbelhajjame> +1

Luc: please signal whether you can attend
... meeting objectives are set, docs will be produced and posted to the wiki

<dgarijo> I'll attend online to the f2f

Luc: also indicate whether you will attend online

<smiles> http://www.w3.org/2011/prov/wiki/ProvenanceTaskForces

Luc: invited people to sign up to Task forces, some have not yet done so
... Model task force: Jun, Satya, Khalid, Paolo have started adding their definitions on the wiki
... others please contribute
... provenance access and query TF: Yogesh, Simon Miles have agreed to be coordinators
... Connection TF: Eric, Stephen, Kai coordinate
... Implementation TF: still looking for confirmed coordinators

<tlebo> are the coordinators listed someplace other than http://www.w3.org/2011/prov/wiki/ProvenanceTaskForces ?

<kai> Its Eric Stephan, not Eric and Stephen

Graham: what does it mean to be a TF member wrt membership of group at large?

Luc: TF membership involves active contributions + autyhor/review docs
... roles and activities within a TF may vary, people can choose. This is to understand who the coordinators can expect to interact with

<Zakim> kai, you wanted to desperately ask for more contributers to the connection TF.

Kai: need more contributors to the connection TF

Luc: TF3/4 -- possible model is: template to be produced by coordinators, contributors to fill in the template
... means that for these TF workload is expected to be very distributed
... coordinators expected to propose a doc structure in the short term in view of the F2F. Outlines to be discussed in next week's telecon

Yogesh: will work with Simon to get something ready for next week

Luc: natural deadline is F2F meeting date, however one week review time would be good. This means end of June effective deadline
... actions will be created on each coordinator for doc outlines to be created

<jorn> already italized coords of TF3

tlebo: are coordinators listed on the TF page?

Luc: not yet, will do

<GK> @tlebo TF wiki page has space for coordinators: http://www.w3.org/2011/prov/wiki/ProvenanceTaskForces

model task force

Luc: need provenance about the definitions that are added to wiki! :-)
... at SW coordination teleconf: debate on Web architecture takes majority of time and resources W3C-wide. We need to have time bounds
... Luc and Paul identified few key points on which consensus is critically needed
... following 5 proposals identified in the agenda
... discussions on provenance model and provenance in the Web architecture are best kept separate at this time

<JimMyers> +1 - is the mapping to web arch part of the access task force? or still model?

Luc: soliciting comments on this

GK: concerned that we may end up with different views that may be hard to reconcile at a later time

Luc: possibly so, but at least we will have made progress on both

<satya> I tend to agree with GK

GK: sees common thread emerging

<JimMyers> +q

jcheney: we many not need to resolve all divergences in the group, let's keep working with provisional definitions, try to be cohesive on each of the two threads separately (?)

<GK> Agree with @jcheney's thrust - don't get hung up on perfect definitions, say something and make progress, review later

Luc: separation of model/arch to continue only up to F2F, at which point we will reassess
... use of term "resource" not helpful in the context of the model (?)

<JimMyers> -q

Luc: first define concepts, worry about mapping of model onto Web arch later

smiles: given this separation: def for resource is just "what is the subject of provenance"?

Luc: term "resource" may not be adequate for the model on its own

<GK> Listening to this discussion: I would move to accept the proposal for now, but review in 2 weeks.

<stain> GK, yup, sounds like the resource discussion is on again.. :)

<dgarijo> +1 to what satya said

satya: use journalism example to ground a concrete def. for resource, and then expand from there. Model and arch view may be reconciled in the context of the example

Luc: in practice issues have emerged recently precisely in the context of the example

<zednik> +1 for separation of concept model from mapping to web architecture (access)

satya: start with "resource" has anything we want to describe the provenance of (check?)

Luc: not yet clear what we mean by "provenance of a resource". leads to "mutable thing" vs "immutable thing"

<Zakim> GK, you wanted to move to accept the proposal for now, but review in 2 weeks.

satya: use "what should be a resource" in the context of the journalism example

GK: propose to accept proposal 1 with option to review in case a divergence is evident

<zednik> +1 for renaming resource

jorn: term "resource" seems overloaded. should also rename "resource" as part of this proposal

<Zakim> jorn, you wanted to propose to rename "provenance resource" so it isn't confused with web resource all the time?

Luc: agree. need a good term to refer to "the thing that doesn't change"

<smiles> +1

<dgarijo> +1

Luc: propose to accept proposal 1 and review it in 2 weeks

<dcorsar> +1

<GK> +1

<Edoardo> +1

<dgarijo> +1

<jun> +1

<stain> +1

<kai> +1

<Yogesh> +1

<Luc> propose: to define provenance-related concepts independently of the web architecture in a first instance, and review it in two weeks

<khalidbelhajjame> +1

<zednik> +1

<iker> +1

<jcheney> +1

<JimMyers> +1

<jorn> +1

<tlebo> +1

<Yogesh> +1

<tlebo> w3.org is great from Troy, NY

<Yogesh> I have issues fro Los Anlges too

<jun> no problem here

satya: agrees with proposal 1

correct: satya agrees with Luc's proposal above

subtopic: proposal 2: the subject of provenance

frew: if the model TF agreed with the OPM definitions at this time, would the TF be done?

GK: not having been involved in OPM or other prior initiatives, position is not to simply adopt one of those models

<paulo> who is on the call?

Luc: not all is good in OPM. So even coming from there, do not think it should be adopted as is. Community will want to evolve the model anyways

<satya> agree

<smiles> yes

Luc: proposal 2 from agenda: "the subject of provenance may be anything, whether physical, digital, or otherwise"

<GK> waiting for proposal in IRC...

<jcheney> +1

<GK> +1

<zednik> +1

<khalidbelhajjame> +1

<stain> +1

<smiles> +1

<Luc> proposed: the subject of provenance may be anything, whether physical, digital, or otherwise

<dgarijo> +1

<Edoardo> +1

<dcorsar> +1

<satya> +1

<kai> +1

<JimMyers> agent was a special case (like PML:source) to capture the idea of a resource that could participate in processes (along the lines of my emails and wiki entries) - agent just couldn't be an artifact if they are completely immutable

<jun> +1

<tlebo> +1

<jorn> +1

<JimMyers> +1

<stain> +1

<YolandaGil> I wonder what category is "otherwise"

<JimMyers> conceptual, logical

<GK> I wouldn't prohibit imaginary, conceptual at this time

YolandaGil: is the subject of provenance anything that we can refer to?

<tlebo> anything to which one may want to refer.

<zednik> mutable?

<GK> Yes, point taken about "can refer to" - maybe the TF can tighten up the definition?

<paulo> in PML, we use the identifiedThing cncept (something that we can refer to)

YolandaGil: correct as "physical, digital, conceptual, or otherwise"?

<Luc> proposed: the subject of provenance may be anything, whether physical, digital, conceptual or otherwise

<JimMyers> +1

<GK> +1

<satya> +!

<Yogesh> +1

<kai> +1

<stain> +1

<Edoardo> +1

<tlebo> +100

<dcorsar> +1

<smiles> +1

<frew> +1

<dgarijo> +1

<jorn> +1

<khalidbelhajjame> +1

<jcheney> +1

<zednik> +1

<YolandaGil> +1

<jun> +1

Luc: accepted

<paulo> if it is anything, can it be a thing that we cannot refer to?

-q

<satya> +q

Luc: mutability seems to get in the way. Provenance of immutable things is a low hanging fruit. A few people made proposals

<JimMyers> +q

Luc: leading to proposal 3: "in a first instance, to define the necessary concepts that allow us to express the provenance of a thing that does not change"

GK: fine to focus on immutable resources initially. but not make immutability an a priori requirement

<satya> +1 for GK's point

satya: what do we mean by immutable things?

<GK> @satya Good question: it's kind of why I don't want to exclude the mutable.

satya: use journalism example and understand what is required regardless of mutable/immutable

<GK> @satya, agree, focus on what's required

JimMyers: mutability leads to a number of special cases

<Zakim> jorn, you wanted to say we can't hinder people from issuing provenance about things which are mutable (web is a distributed system)

<JimMyers> -q

<tlebo> is there anything that is universally immutable? Roles seems to be a good approach.

<satya> good point @jorn (good point on owl:sameAs)

jorn: if we restrict certain things to be immutable, that may be an artificial constraint that may not work for whoever uses the model

<JimMyers> I don't know how to explain except in the context of my proposed 'solution' - mutability is a role of a resource w.r.t. a process - if that's a good model, I don't see how we could discuss immutability first and then change the definition of resource in some way to address mutability

<tlebo> must go. apologies.

still muted...

<GK> I think everyone is basically agreeing... focus on the case of immutable resource example, but don't assume immutability unless we really have to

<khalidbelhajjame> +q

smiles: immutability may not be the issue

<GK> @smiles like your phrasing "insofar as it's immutable we can talk about its provenance"

khalidbelhajjame: if we tackle mutability at a later time, that may lead to revisiting many other definitions

<JimMyers> if the question is whether we should have a way other than resources to describe changes in state - +1 - there's a role for mutable resources but we don't need a mechanism to define state changes of mutable resources separt from defining immutable resources that encapsulate that state (but are just resources)

paolo: isn't that the case that things that do not change only have a provenance if they have changed in the past? I am confused

<frew> "WORM" resource?

<satya> I think we need more discussion - over mailing list?

<GK> Agree in principle with wjhat we discussed

Luc: is there a consensus?

<JimMyers> -1 - I'd like to discuss things together...

<jcheney> what's the formal proposal now?

<satya> Proposal: we do not make assumption about mutability/immutability of object

<zednik> +q statement about mutability

<kai> I think we have too many mutable resources out there so I would try to deal with them from the beginning.

satya: a few things not clear, but we can go with mutability/immutability in the context of the running example

<GK> @satya broadly agree with "Proposal: we do not make assumption about mutability/immutability of object" but would add "unless the use-case requires us to"

JimMyers: the distinction is significant in the context of (relative to) processes. possibly this pov gives us a way forward in the discussion

<Luc> would it help if we said state of a thing

<Luc> instead of a thing that does not change

<dgarijo> even the example has "mutable things", so it will be difficult to leave them out of the discussion

<satya> @GK agree, if required for use case

<scribe> ACTION: JimMyers, satya to formulate proposals that we can vote on next week [recorded in http://www.w3.org/2011/06/02-prov-minutes.html#action01]

<trackbot> Sorry, couldn't find user - JimMyers,

<JimMyers> I put a 'definition' of resource on the wiki page just before the call - that's my proposal for a model

<JimMyers> Perhaps I could try to apply that to the use case to make it clearer...

<jorn> subject ?

<satya> entity?

<zednik> entity

Luc: term "resource" not useful here as too loaded as architectural term

<kai> +1 for entity

<GK> "Subject of provenance" (Luc's phrase from an earlier proposal)?

+1 for > "Subject of provenance" (SoP)

<dgarijo> +1 for Subject of Provenance

paulo: connection b/w mut/immut and phyisical/digital

(missed the rest)

paulo: other topic to discuss is how to refer to things, either mutable or immutable

Luc: true, but not current topic

<stain> mutability is very related to identifiable - depending on how you identify it might be mutable or immutable

<kai> Maybe it would be doable to restrict provenance to immutable subjects and provide ways to see mutable subjects as immutable, e.g. by adding a version or a timestamp.

Luc: reminder - provXG summary presentation by Yolanda tomorrow

<Edoardo> Bye

<dcorsar> bye

<stain> I am wondering if some kind of "observation" is needed

<Luc> paolo, I will just to the magic incantation to have the irc log in the wiki for you to edit

ok go ahead

<kai> @stain yes, thats what I mean. Will keep it in mind.

<Luc> trackbot, end telcon

<trackbot> Sorry, Luc, I don't understand 'trackbot, end telcon '. Please refer to http://www.w3.org/2005/06/tracker/irc for help

trackbot, end meeting

- DRAFT -

Provenance Working Group Teleconference

02 Jun 2011

Attendees

Contents

review of actions

F2F1

model task force

Summary of Action Items

Scribe.perl diagnostic output