Permissions and Obligations Expression Working Group Teleconference

<victor> Hi!

<benws2> Can someone please give me the webex link.

<renato> https://www.w3.org/2016/poe/wiki/Meetings:Telecon20161003

<scribe> scribe: phila

<scribe> scribeNick: phila

<renato> Chair: Ben

Approving TPAC minutes

They're linked from the agenda

-> https://www.w3.org/2016/09/22-poe-minutes.html Minutes 22 Sept

<renato> https://www.w3.org/2016/poe/wiki/Meetings

-> https://www.w3.org/2016/09/23-poe-minutes.html Minutes 23 Sept

<benws2> No

phila: Will try and make minutes appear in main list

benws2: Any objections to the minutes?

renato: I'd like to go through them here if we can

<simonstey> +q

renato: I just thought we'd go through the topics...

benws2: Might others go through it offline and bring any issues up?
... I don't want to spend the whole of this call going over stuff we've done already [scribe paraphrase]

simonstey: An admin comment - I pointed out last week that if you look at the summary of the resolutions, for example resolutions 5-8 which are just 'not accepted'. The minutes don't necessarily tell you what was resolved
... There's a resolution called POE.R.V.03 ACCEPTED is clearer

<renato> https://www.w3.org/2016/09/22-poe-minutes.html#ResolutionSummary

-> https://www.w3.org/2016/09/22-poe-minutes.html#ResolutionSummary Resolution summary

<renato> https://www.w3.org/2016/09/23-poe-minutes.html#ResolutionSummary

benws2: I see. Who would do that work?

phila: Who was scribing?
... It will have to be me but future scribes please take note.

benws2: So we can run through the summary of resolutions which isn't all that long.

renato: We had 17 topic areas on day 1 and we only have 10 resolutions

<simonstey> +q

benws2: OK, I do think people should go through it offline rather than take up time here.
... And almost everyone on this call was there.
... So when can we approve the minutes, renato?
... It's an approval that the minutes are a fair reflection
... and don't contain any misrepresentations
... So unless I hear an objection?

[None]

simonstey: A minor comment - if we encounter... the question is, if we have made resolutions, but if they're not scribed fully, do we have to re-do that in another call?
... Renato said that we have a large number of topics discussed at TPAC, b ut we only have 10 officially noted in the minutes, so if the resolutions aren't recorded in the minutes, do we need to go back and re-do those?

benws2: I think we care. I suggest we update the reqs doc with the resolutions from TPAC. Where we find gaps, see if there is an implicit resolution in the minutes, then maybe we need to go back.
... Any objections?

PROPOSED: Update the UCR according to the minutes. Highlight any gaps in those minutes and bring matters arising to future telco

<simonstey> +1

renato: The UCR has no reqs in it at the moment. Are we putting all the Reqs in the UCR and then note the ones that were rejected?

simonstey: At this stage, I'd note it in the wiki. And only put the agreed ones in the GH version

benws2: I agree. Reflect it in the wiki, not the doc itself

+1 from me on that

PROPOSED: Update the UCR wiki according to the minutes. Highlight any gaps in those minutes and bring matters arising to future telco

<simonstey> +1

<Brian_Ulicny> +1

<benws2> +1

<victor> +1

<michaelS> +1

<ivan> +1

+1

<sabrina> +1

<smyles> +0

RESOLUTION: Update the UCR wiki according to the minutes. Highlight any gaps in those minutes and bring matters arising to future telco
... Minutes of TPAC face to face meeting accepted

Use Cases

benws2: We got two new UCs from the book industry at TPAC
... They should be added. Who gets that job?

renato: I had the impression that those UCs were being drafted and we're going to get a new version soon.

benws2: How long do we have? It took a while to get their originals

<ivan> +1 to Ben

benws2: I don't really want to hold us up for another couple of months?

renato: We can give them a deadline as well as discuss overall deadline for UCs?

benws2: You're point person with that group

renato: I can send them an e-mail and give them a week to respond.

Data Quality policy Use case

<renato> https://www.w3.org/2016/poe/wiki/Use_Cases#POE.UC.26_Data_Quality_Policy

<michaelS> https://www.w3.org/2016/poe/wiki/Use_Cases#POE.UC.26_Data_Quality_Policy

<victor> http://pro.europeana.eu/person/antoine-isaac

phila: Gives pocket background to Antoine

benws2: It looks to me as if he has given an accurate reflection of his use case using the existing ODRL spec

<simonstey> +q

renato: I cane to that conclusion. In his Duty, there's no action. I think he missed that the assignee up time {... ??}
... In ODRL 1.1 we had a group doing a profile for services, looking at things like uptime

benws2: There's another issue here that we're deadling with. It's the diff between a dcat:Dataset and a Distribution.
... He's targeting the permission at the Dataset and the duty at the Distribution

phila: Rambles about the diff between dataset and distribution
... So ODRL has a problem with target and duty being different things

renato: What he has is fine from an ODRL POV
... Duties can be obligations on other resources, not necessarily the original target of the policy.

[Discussion on detail of assignees, policies]

benws2: So it looks as if existing ODRL can do it.
... Should we go back to Antoine?

<victor> A qualification of assets with FRBR (https://en.wikipedia.org/wiki/Functional_Requirements_for_Bibliographic_Records) would ease Ben's case. Once these entities are connected (dataset, distribution), logical relations can be established in a simple manner.

simonstey: Just reading the policy, I was wondering whether a duty must have an action?

benws2: Yes.

simonstey: It looks as if there is an action missing.
... He says "The expression of constraints in ODRL seems quite unfit with expressing general constraints on values in RDF graphs"
... I would argue that this is not the goal of ODRL, that's what SHACL is for

<renato> ODRL ne SHACL ;-)

simonstey: So in this case, ODRL can be used, but I'm sure there are situations where this constraint mechanism of ODRL is not expressivbe enough to handle all use cases
... Maybe we need to incorporate a means to refer to a SHACL shape from ODRL.

benws2: Between us, can we get an example of a UC from Anotine that can't be so easily satisfied, where the use of SHACL would satisfy it.

smyles: I wonder if Antoine's comment... is he saying there' nothing in the ODRL policy that tells you the relationship between the dataset and the distribution
... So I wonder if we need a means to express the relationship. And only AI can answer that.
... I'd like to caution - not everyone uses or wants to use RDF. Let's not require RDF, adding in lots of RDF processing

ivan: I would be careful with bringing SHACL here. I tried to re-read Antoine's work. I think he uses the word constraint differently from ODRL.
... He has difficulties in attaching rights expression-like things to an RDF graph.

<simonstey> The mission of the RDF Data Shapes Working Group is to produce a language for defining structural constraints on RDF graphs. In the same way that SPARQL made it possible to query RDF data, the product of the RDF Data Shapes WG will enable the definition of graph topologies for interface specification, code development, and data verification.

<simonstey> https://www.w3.org/2014/data-shapes/charter

benws2: That would seem to advise that we ask for a use case that is not so readily satisfied by ODRL

phila: I am happy with pointing to 'a validation mechanism' but I would caution against specifying SHACL. Too RDF-centric and creates dependency we can avoid

benws2: Discusses who will write to Antoine

<scribe> ACTION: phila to write to Antoine as heads up for Renato's mail [recorded in http://www.w3.org/2016/10/03-poe-minutes.html#action01]

<trackbot> Created ACTION-32 - Write to antoine as heads up for renato's mail [on Phil Archer - due 2016-10-10].

<scribe> ACTION: renato to write to Antoine asking for example that ODRL can't handle [recorded in http://www.w3.org/2016/10/03-poe-minutes.html#action02]

<trackbot> Created ACTION-33 - Write to antoine asking for example that odrl can't handle [on Renato Iannella - due 2016-10-10].

Educational Use Only use case

[Scribe missed some discussion]

benws2: Meaning of 'Educational use' varies according to country but the concept exists everywhere
... So do we try and define it?
... We have terms like read and derive, but 'Educational Use Only' seems to be embedded in local jurisdictional meaning. Should we support those?

renato: If there's no support for it, we don't do it.

phila: Does this speak to the enumerated list we talked about in Lisbon?

<renato> https://w3c.github.io/poe/vocab/#term-policyUsage

renato: It's a value for an actiuon.

benws2: So it sounds as if we need a more popular appeal for 'Educational Use Only' if it's to go into the vocab.

smyles: When we were working on ODRL we identified Educational Use as a need to indicate for some of the rights work that we do.
... We might say of a video, not for these purposes, but Ed Use is OK.

benws2: So you have a specific meaning of Educational Use?

smyles: Nope. It;'s just written that way.

<victor> s/thatn/than

smyles: We often have restrictions for ... this sporting video can be used in a news programme but not a sports programme.
... That sounds similar to Ed Use
... You're not allowed to use this for parody.

benws2: We have similar things/.
... But I wouldn't expect ODRL to manage and define.
... This is domain specific.

smyles: So ODRL currently has a mechanism (purpose) but it doesn't provide the values.

benws2: That's my intuition.

Topics: Issues and Actions

action-17?

<trackbot> action-17 -- Simon Steyskal to Update ucr editor's draft as discussed -- due 2016-07-11 -- PENDINGREVIEW

<trackbot> https://www.w3.org/2016/poe/track/actions/17

simonstey: That's an old action.
... This was for our first PWD

close action-17

<trackbot> Closed action-17.

<scribe> ACTION: simon to update UCR according to today's discussion [recorded in http://www.w3.org/2016/10/03-poe-minutes.html#action03]

<trackbot> Created ACTION-34 - Update ucr according to today's discussion [on Simon Steyskal - due 2016-10-10].

renato: One of these actions came from TPAC

action-30?

<trackbot> action-30 -- Stuart Myles to Can we only have a json-ld serialisation? will it impact righstml? -- due 2016-09-30 -- OPEN

<trackbot> https://www.w3.org/2016/poe/track/actions/30

renato: That's on Stuart who wasn't there
... The question is, if we had JSON-LD only, would that impact on Rights ML (i.e. no JSON-only)

smyles: I'll come back to the WG on that.
... If we have that and eliminate 'pure JSON' do we also eliminate pure XML?

ivan: I would say you shouldn't become an expert... I can show you a spec that uses JSON-LD but you can read it as a JSON-only

<ivan> https://www.w3.org/TR/annotation-model/

ivan: I think... it does refer to JSON-LD in the intro...
... All the examples there are all JSON. It's just the @context file that makes it JSON-LD which you can safely forget.
... My ideal would be that ODRL can be expressed in a similar way.

smyles: This is a spec I wanted to look at anyway :-)

simonstey: I see it as anyone can raise an issue, but the WG can decide not to open it or not.

-> https://www.w3.org/2016/poe/track/issues/raised Raised Issues

renato: Issue 7 and 9 are now merged.
... We're only going to include requirements in the UCR that we're going to fulfil/address. The wiki will retain rejected ones.

close issue-8

<trackbot> Closed issue-8.

[Discussion on possible future changes to call timing.]