– DRAFT –
DXWG plenary

I can scribe

Initial discussion: we will have fewer meetings during the months between now and September, because it seems all europe is on vacation.

We will move to every two weeks.

PROPOSED: accept last week's minutes

<kcoyle> https://‌www.w3.org/‌2018/‌08/‌09-dxwgdcat-minutes

<alejandra> https://‌www.w3.org/‌2018/‌08/‌07-dxwg-minutes

<alejandra> 0 (was absent)

<roba> 0

0 (was absent)

Resolved: : accept last week's minutes

<azaroth> Belated +1

open action items: most are DCAT, one for Jaro isn't due til September. Skipping the one about creating an outline. Nick is not here for his one about creating a draft.

roba: Nick is taking on most of the work for the first draft, as Rob doesn't have time at present. Rob will support him in this effort, believes NIck is still able to do the work.
… One item to consider is the naming of profiles, "the thing and elements in it"

kcoyle: Antoine recommended doing a data model. We should make that decision before going through all of that.
… next item is in the profile guidance group. I suggested some deadlines but haven't heard from anyone. By August 21, to have the documentation that we've asked for presented to the group, so the group can say whether they feel they understand what's being proposed as an ontology. Until we have that, we can't really figure out where it fits in with our requirements.
… Rob, Nick, and Simon probably aren't on summer vacation. By September 11, the group could make a decision.

roba: yes, OGC meeting in Stuttgart is on the 12th, so we need a decision for that as well.
… I will drive that.

I do have a bit of info overnight that some funding for some future evolution of the work is on the way.

kcoyle: it would be good to include a bit about that in the documentation, where did the motivation come from.

roba: the documentation itself should make that clear. The definition is quite clear. Without a mechanism, it's just whistling in the wind.

kcoyle: we need to develop enough clarity for everyone that we can gain consensus.
… For those who haven't been involved in the process you've been involved in, putting in the motivation is important, how does it fit into workflows, etc.

roba: there's already been a lot of comment about individual aspects. We've helped clarify the model. We need to separate out how flat the implementation should be. I think people have already been engaged. Lars and Rueben are fairly agnostic. Discussions have settled down to consensus. We do have to give the group the opportunity to reflect on why the thing exists and is felt necessary, but in terms of bringing people on board and unders

tanding it, we need to be more clear about what specifically needs clarification.

kcoyle: I think you are talking about the model rather than the ontology.

kcoyle: are there comments?

<roba> ontology is the model..

<roba> ontology is just formalisation of a model

<roba> we know if its words interpretation is too loose to be useful

<roba> i can do a UML version - but the LODE documentation provides equivalent diagramming of the model

<roba> roba

roba: as long as we are transparent about the expectations, I don't think we need to be prescriptive about it.
… any model needs to be expressed in some formalism. The formalization is all the ontology is. It's just keeping the model in a machine-readable way. I guess you can have ontologies that are implication-focused.

kcoyle: it can be implemented in multiple forms. We haven't gotten to the point of describing a profile in any particular implementation.

roba: I've done a lot of work in the space of profiles for UML and schemas from that. for my day job, I generate ontologies out of such models. At some level, they look quite similar, but an owl data model is just a W3C way of generating a data model.

kcoyle: what we see so far is that the profile guidance document is not going to be at the point of code. It's going to be very high profile. The question is the profile description. We are not generating any model for profiles themselves. How do we do these two things that are at very different levels?

alejandra: Is Antoine's data model UML? Otherwise, I agree with Rob, that there is no problem in choosing OWL, given that we are a W3C group.

kcoyle: I don't belive anyone was talking about a formalization.
… what I'm talking about is what we have to decide, what the issues are that mean we have to decide those things.

roba: I propose kcoyle takes an action to get something in writing from Antoine.

kcoyle: this is what we'll talk about in late August/early September

roba: we have a report that there is a concern…

kcoyle: it's in github

roba: I read Antoine's comments very differently. I'm comfortable that where he's coming from is not inconsistent. We can deal with that, that's fine.

kcoyle: by August 21, we need the documentation we requested relating to profile description.

roba: we've accepted a lot of rewordings of requirements, deduplications, so there's a lot of work needed to boil them down. I think we can connect to them by identifier, knowing that they will change. If we refer to 3 requirements that say the same thing with different wording, I guess we just let them all be in there.

kcoyle: yes

roba: there already was discussion around the ontology which made the decision that…I think following the process we should go back to this original subgroup's discussion.

kcoyle: can you share a link to that?

roba: I'll do that. I think that's where some of the context has been lost.

kcoyle: this is a plenary decision now.

roba: this is treating it as a flavor of DCAT.

Action: roba to make sure that the documentation for the DCAT decision is highlighted

<trackbot> Created ACTION-177 - Make sure that the documentation for the dcat decision is highlighted [on Rob Atkinson - due 2018-08-21].

kcoyle: anything else we need to cover?

kcoyle: I will suggest to Peter that we not meet next week. We'll look at the calendar, but I suspect we won't get anyone from the EU.

Action: kcoyle to get in touch with Peter to determine future schedule for plenary meetings.

<trackbot> Created ACTION-178 - Get in touch with peter to determine future schedule for plenary meetings. [on Karen Coyle - due 2018-08-21].

<alejandra> thanks, and bye!

addendum: discussion around 1:23 with Annette about the value of content negotiation for profiles. annette_g expressed concern about the variability of content in negotiated versions. azaroth pointed out that content can techncially differ.

annette_g pointed out that variation is common for URIs for non-content items.

… but not for content items (documents)