– DRAFT –
DXWG Plenary

<ncar> hi all

Agenda

<PWinstanley> https://‌www.w3.org/‌2017/‌dxwg/‌wiki/‌Meetings:Telecon2018.10.30

PWinstanley: let's start with some admin. looking at minutes from previous meetings.

Admin

<PWinstanley> https://‌www.w3.org/‌2018/‌10/‌16-dxwg-minutes

<LarsG> +0 (wasn't there)

<PWinstanley> proposed: accept minutes of 16th October 2018

+1

<alejandra> +1

<ncar> +1

<PWinstanley> +1

<Jaroslav_Pullmann> +1

<roba> +1

Resolved: accept minutes of 16th October 2018

<PWinstanley> https://‌www.w3.org/‌2018/‌10/‌25-dxwg-minutes

<PWinstanley> proposed: accept minutes of 25th October 2018

+0 (wasn't there)

<ncar> +0 (was actually partly there on the phone)

but I very much appreciate them!

<Jaroslav_Pullmann> +1

<alejandra> +1

<LarsG> +1

Resolved: accept minutes of 25th October 2018

<roba> +1

<PWinstanley> +1

<PWinstanley> https://‌www.w3.org/‌2018/‌10/‌26-dxwg-minutes

<PWinstanley> proposed: accept minutes of 26th October 2018

<alejandra> +1

<PWinstanley> +1

+0

<ncar> +0

<Jaroslav_Pullmann> +1

<LarsG> +1

Resolved: accept minutes of 26th October 2018

<roba> +1

open action items

<PWinstanley> https://‌www.w3.org/‌2017/‌dxwg/‌track/‌actions/‌open

PWinstanley: 237 - Igot info about timelines. There were some tools that Phillipe pointed me to, and the steps for the complete W3C procedure.

PWinstanley: Rob has multiple items, 243 is finished, 241? that was done.

close: action-237

close action-237

<trackbot> Closed action-237.

close action-243

<trackbot> Closed action-243.

close action-241

<trackbot> Closed action-241.

Face-to-face wrapup

PWinstanley: Karen made good notes about the main event.

<AndreaPerego> http://‌lists.w3.org/‌Archives/‌Public/‌public-dxwg-wg/‌2018Oct/‌0630.html

<AndreaPerego> I think is this one.

<alejandra> https://‌lists.w3.org/‌Archives/‌Public/‌public-dxwg-wg/‌2018Oct/‌0630.html

PWinstanley: first question is how was it for you? did everyone involved with the F2F feel that it went okay?

<AndreaPerego> +1 from me

<Jaroslav_Pullmann> +1

<PWinstanley> +1

<LarsG> +1

<roba> +1 from me

<alejandra> +1

PWinstanley: was there anything we should have done but didn't?

<AndreaPerego> I think we covered pretty much what was planned.

alejandra: Karen's point about alignments, I think we agreed we would leave them out, but her email suggested the schema.org alignments should go in.

PWinstanley: there was certainly an action for myself and danbri to look at the alignment.

<AndreaPerego> +1 to alejandra. This is what I recall

PWinstanley: we discussed that there would be a variety of options, like crosswalks etc. Did we want them to be static or not? Do we link to a frozen-in-time version, or a github.

alejandra: I think there is value in having something for people to use. My point was that the email suggested we would have it in the document.

PWinstanley: I think it was going to be somewhere in the document, maybe just a link, but that remains to be determined. One option would be an opportunity to update if we pointed to a github folder of a set of crosswalks.

<AndreaPerego> I think that's exactly the point. Once the WG will be closed, we won't be able to update the alignments.

Jaroslav_Pullmann: the obvious benefit was to synchronize on the requirements for profiles.
… this was a benefit for the UCR document
… Are there some actions to be taken?

PWinstanley: I'm not sure we can deal with that in the DCAT. we're seeing the benefit of even one well considered response. We hope others will respond with meaty problems that need to be thrashed out.

Jaroslav_Pullmann: I've very curious about the details, to see the weaknesses of the current design.

PWinstanley: yes, Clemens sounds like somebody with whom one can have a bit of a dialogue.

AndreaPerego: Clemens when he was talking about errors, that was for the geospatial community.
… As far as I know, once a group is closed, the wiki is frozen and so is the repository, so putting things there would not work for long-term maintenance. Who would take care of maintaining it? Schema.org is changing multiple times a year. It would be a challenge to keep up.

<alejandra> +1 to AndreaPerego's concerns - these alignments need to be maintained and it is not something we can commit to do

AndreaPerego: There is discussion in the repository to add the web API. I'm a bit concerned about how we can deal with that. I have the impression that the best place where these alignments should be maintained is schema.org itself.

PWinstanley: It makes a lot of sense if the DCAT proposal is stable and schema.org is always changing that they be responsible for the maintenance. It also makes sense, given that Google says they will be focusing on DCAT as well as schema.org for data discovery, it looks as it it's something they will want to be maintaining.

PWinstanley: I think there's a lot to be said for what you say. Phillipe was saying that there were a range of options, some more flexible than others. But sort of wrapping up everything and freezing it makes sense.

DCAT 2PWD and review

<PWinstanley> https://‌www.w3.org/‌TR/‌vocab-dcat-2/ the publication

PWinstanley: DCAT 2PWD was published Oct.16, we had text in English and Spanish for the announcements. Remember the spreadsheet with contacts.
… this is the text

<PWinstanley> https://‌www.w3.org/‌2017/‌dxwg/‌wiki/‌DCAT_revision_-_public_working_drafts#Text_for_Request_for_Comments text for announcements

PWinstanley: have people started writing about this? emailing and talking about it at meetings?

<AndreaPerego> Done from my side.

PWinstanley: I've parasitized quite a bit of the text and turned it into a LinkedIn chunk.

<Jaroslav_Pullmann> yes, all listed Fraunhofer colleagues were contacted

<AndreaPerego> BTW, we got another comment on the mailing list: http://‌lists.w3.org/‌Archives/‌Public/‌public-dxwg-comments/‌2018Oct/‌0003.html

I've also contacted some businesses directly. Any other things we could do?

PWinstanley: the DCAT group can pick up that comment from Luca Trani.

<AndreaPerego> At the f2f I mentioned that we should have also a tweet from the W3C account, but I don't know if this was done.

PWinstanley: also on the agenda is abstracts in other languages.

ncar: I haven't followed the DCAT closely enough, but if I get the text I can do a Polish version.

<AndreaPerego> ncar, this is call for comments: http://‌lists.w3.org/‌Archives/‌Public/‌public-dxwg-comments/‌2018Oct/‌0001.html

<alejandra> have you got a link ncar ?

ncar: we have an example in which distribution services are really working.

<ncar> example of a DataDistributionService: http://‌linked.data.gov.au/‌dataset/‌gnaf?_format=text/‌turtle

PWinstanley: The sooner we start on implementations the better.
… things like what Nick is describing can be very helpful as well as raising awareness.

<alejandra> https://‌www.w3.org/‌2017/‌dxwg/‌wiki/‌DCAT_revision_-_public_working_drafts

<ncar> Thanks, got it

alejandra: this is the wiki with the English and Spanish version, maybe the other languages can be added there.

PWinstanley: this is extremely helpful.

Profile negotiation

LarsG: we had extensive discussion during the F2F. We accepted 3 more requirements and decided how to proceed. The plan is that next week we will vote on having the current editor's draft as a first public working draft.

LarsG: The conneg working group's request is to have the current editor's draft reviewed by the whole group and discussed next week.
… If we can't agree, we will publish a new version.

PWinstanley: is there anything contentious?

LarsG: yes

PWinstanley: where do you want to focus?

<LarsG> https://‌w3c.github.io/‌dxwg/‌conneg-by-ap/

<AndreaPerego> Just an idea. Contentious parts can be possibly complemented with a note and/or a GH issue. This would help readers understand what is still under discussion, and focus their comments.

LarsG: motive, introductory text, sections 5 and 6. In 5 we define an abstract model. In 6, we define three ways of implementing the model.

PWinstanley: Please review as early as possible and make github issues as early as possible in advance of the meeting.
… the sooner he gets and idea of which way the wind is blowing, the better placed they will be to redraft.
… what we're doing here is publishing a straw man, testing the ideas. Getting criticism is something we should be looking for. Let's try to engage in debate through the issues as soon as possible.

LarsG: It's Rob and me both working on this.

PWinstanley: You're an abbreviated form ;)

<roba> * and blame :-)

PWinstanley: what we've seen through all the work in the last month, we should have a big cheer for how much we've accomplished.

LarsG: I created the number 500 at the F2F

Profile guidance

ncar: there is an action item on me to review the document.

roba: A lot of the work has been around introductory text, intro to how the pieces fit together. I think Clemens's comment is a useful thing to revisit. It reinforces some use cases that may not have been far enough at the front of people's minds.

<AndreaPerego> +1 to the roba's point about Clemens's comment.

roba: some things can be pushed out now that we have a decision. We need to get the stubs to a position that people are comfortable with.
… if you use one particular term from another vocabulary...at what point is something really profiled?

PWinstanley: we may not necessarily be in a position to publish a recommendation, we might push in a rough direction. The jury is still out as to how the doc looks.
… We were talking about the profile description ontology as being a note. We were talking with W3C team, asking how do we know if it's a note or a rec? The answer was to send it through as a FPWD and they'll let us know.

roba: do you want to update on the profile ontology?

Profile ontology

roba: it's always been in a form that could go as a recommendation if we want that. In terms of stability, there've been quite a few comments coming in. Mostly it's improvements of naming, not new terms.

PWinstanley: At the F2F, it was the quality of presentation that we wanted you and Nick to take a look at.

roba: Nick has been picking up a lot of the work, so I'll let him comment on that. There are a couple open debates around the issue of entailment. Who is responsible for drawing the connections? What mechanisms are available to flatten the hierarchies out? I think that's an issue for the profile guidance, not the ontology.

PWinstanley: with Riccardo's input, we've been talking about transitive properties.

<AndreaPerego> I think that transitiveness of profile entailment is a possible feature, if intended by who designs the profiles, but it is usually not the case.

<PWinstanley> https://‌www.w3.org/‌TR/‌skos-primer/#sectransitivebroader

roba: It confirms the model being proposed. The broader issue is an architectural concern about scope. Whether the mechanisms have been provided to do the flattening. Where that happens, we don't specify. It may impact the conneg guidance. We can't ignore that the hierarchies exist in the wild. We have to support it.

ncar: The discussions we've had around the presentation has been about diagramming. We haven't had strong feelings around radically different diagramming. What Peter mentioned is useful. There may be diagrams we can add.

PWinstanley: people would naturally have a concern about the explosion of linkage when you get into things like Dublin Core that get used everywhere.

ncar: there is a question of how we handle notions of transitivity. We could reproduce a diagram or link it. How much discussion of what would actually happen in the real world do we include?

PWinstanley: the real world is important. SKOS vocabularies are usually well contained and not connected to external resources as they could be.

ncar: the DCAT AP profiles are outlined, but we don't outline the properties of transitivity.

PWinstanley: we can provide advice as to what is practical.

ncar: I think that's outside of the scope for FPWD

PWinstanley: we're at the top of the hour. We'll have to put the rest of the agenda at the top of the next meeting.

PWinstanley: we need to think about how to get involvement of others, and we have a week to get familiar with the profiles guidance doc.

PWinstanley: for the profiles ontology, the path forward is to have the team assess it.

ncar: we might be talking doing that next week.

PWinstanley: I think it's important to pay attention to the audience, be sure you give it the best shot to get a good reception. If there are obvious improvements we can make, don't wait.

ncar: nothing off the top of our head
… we think it's ready

PWinstanley: it will next need to be put in front of the plenary.

<Jaroslav_Pullmann> thanks & bye!

<AndreaPerego> Thanks, and bye!