Dataset Exchange Working Group Teleconference

09 January 2018

Meeting Minutes

<nicholascar> great!

scribe nick:PWinstanley


<kcoyle> https://‌www.w3.org/‌2017/‌12/‌12-dxwg-minutes

<alejandra> could we update the table at https://‌www.w3.org/‌2017/‌dxwg/‌wiki/‌Meetings with all the scribes, please?

Approval of last meeting's minutes

kcoyle: I will try to update the scribes list

Resolved: approve minutes of previous meeting (Dec 12, 2017)

promote UCR

roba: I note that we have a new member on the list - Nick Car

<kcoyle> https://‌docs.google.com/‌spreadsheets/‌d/‌1_f5CAZv7rgUjJH5YXkmD6w5xS6GX5X2AmEp_LMptAfQ/‌edit#gid=0

kcoyle: we need to think about doing a second round of emails to promote/stimulate responses to the UCR publication

DaveBrowning: what is the green colour?

kcoyle: Caroline did that , but not sure what it means. The column 'sent' should be the key
… if your name is on the spreadsheet then send it

nicholascar: should we add new contacts?

kcoyle: yes please

alejandra: people are supposed to reply by 20 Jan, we reply to comments. What's next?

kcoyle: comments feed into the next version of the draft
… we don't have a set date for the next draft

alejandra: but if there are no comments?

kcoyle: we ourselves might want to update the draft too ...

open action items

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

kcoyle: jaroslav is still intending to cover action #50
… andrea - Action 59?

AndreaPerego: I am still on the case, hope to get completion in the next couple of weeks

roba: Mine can be marked as done

kcoyle: Rob, please can you say what is going on in OGC (open geospatial consortium) that's relevant

roba: OGC specialises in spatio-temporal matters. I have a role as OGC staff member supporting knowledge management, and we are interested in describing relationships between our specifications, their versions and their profiles. I'm building a linked data environment to describe this
… Because it is early days I have freedom in choices and have decided to link it to DXWG activity stream

kcoyle: on our home page we have related docs/vocabs. If there is anything related please link it there

roba: I plan to share a model for describing profiles.

kcoyle: I am doing a similar thing for Dublin Core

<AndreaPerego> Just to note that OGC has also a group working on DCAT for geo metadata - already included in the list of stakeholders that have been contacted for the UCR doc.

Makx: Do you want to look at the EC ADMS schema - a DCAT-based schema for profiles and schemas.

kcoyle: please add to the home page

<SimonCox> @Makx - what is the scope of ADMS?

<AndreaPerego> ^^ https://‌www.w3.org/‌TR/‌vocab-adms/

<kcoyle> https://‌www.w3.org/‌2017/‌dxwg/‌wiki/‌Main_Page#Background_About_Profiles

<SimonCox> (I have looked at it but not fully clear what set of problems it targets)

roba: while I've been working on requirements I'm happy to slot in with wider activity - the namespace must relate to a body that OGC has liaison with already, but it can be brought through this group to do that
… Gathering prior art would be a good starting point

kcoyle: does this become a use case?

roba: I think it might be covered by existing generic UC, but we might need a more specific one.

kcoyle: add it as a github issue to allow discussion before we firm up the action on this


kcoyle: Peter has volunteered to take over from Caroline when she steps out
… we will work Peter into the schedule

<LarsG> PWinstanley++

<AndreaPerego> +1!

<annette_g> +100

meetings for DCAT & Profiles

<antoine> hurray!

kcoyle: we determined that DCAT meetings will be on Wednesdays, 20:00 UTC

alejandra: After we made this decision Andrea updated his availability and it doesn't include this time. We will start with the decided time and review depending on attendance

AndreaPerego: I changed the timeslot when the poll was extended. The Wednesday 20:00 will be difficult for me
… If I'm the only one affected then just go ahead without me

kcoyle: we need to know who is going to call the meeting. Use of webex is through Dave Raggett - but if you are using other things then that's fine. Use the IRC and Zakim to facilitate getting a record of the meeting
… There needs to be a public record

SimonCox: Do you want me to do it?

kcoyle: Yes please

<SimonCox> I will manage DCAT meetings then

<DaveBrowning> +1 to SimonCox

<alejandra> +1

kcoyle: Please post on email the schedule and connection

<AndreaPerego> What about the IRC channel to be used? #dxwgdcat?

<alejandra> so, use the same channel as now

kcoyle: we can use #dxwg
… the same channel as now
… using RRSAgent we will have to experiment with naming of minutes

SimonCox: I will contact Dave Raggett and sort arrangements

kcoyle: we can also ask about naming conventions and options for automated management of minutes
… Next is the profiles group

AndreaPerego: I propose a separate IRC channel - just in case there are parallel sessions
… Dave R can advise

<Zakim> RubenVerborgh, you wanted to ask whether Profiles group includes content negotiation

RubenVerborgh: the last email was about splitting profiles into guidance and content negotiation - what does the group cover

kcoyle: separate groups for each deliverable - so how do we want to split
… Did you see the note after antoine posted to suggest that the content negotiation could be independent and be covered before the profiles
… RubenVerborgh and LarsG can proceed without waiting for others
… The profiles should be renamed to profile guidance

RubenVerborgh: There should be an invitation email

kcoyle: most of those interested in profiles were interested in guidance.

<roba> I intend to implement profile negotiation.. so will join that group.

RubenVerborgh: I will send an email around

alejandra: I want to return to the DCAT group - are we planning to start this week?

antoine: Organisation of profile discussion - it can meet in the same space as the content negotiation so as to cross-fertilise

<AndreaPerego> +1 to antoine

annette_g: We might want to start out together and split into subgroups later

kcoyle: that is what we were trying to do on the content page. It was clear that there were different parallel paths that were not meeting

annette_g: the first joint meeting should discuss the broad definition and then determine the commonality that allows the content negotiation people to push ahead

<roba> +1

<alejandra> +1

kcoyle: What do people think? Does the initial joint meeting followed by splitting up later suit?

roba: I support this.
… Also at least 3 of us are modelling profiles, so this needs to be done in plenary so that we have an agreed formalism as a starting point

antoine: I was suggesting for cross-fertilisation, as I don't think we can get an agreement in just one meeting

<annette_g> +1 to antoine

kcoyle: Rob mentioned a plenary. Would it make sense to have this discussed in a general meeting?

<roba> regular agenda item :-)

<annette_g> :D

kcoyle: I am willing for us to do that if that is what the group wants

roba: I think we should ask the question
… are we in a position to parallelise activities, or not?

kcoyle: I fear we will have to come up with more than one term - one for ConNeg, and another for guidance

<roba> -1 for different term - +1 to formalise model so we can see if it is the same

antoine: does the charter allow us to split definitions in the final specs?

kcoyle: the charter uses 'application profile' for both, but we could decide on a macro-level and a more specialised version. I've talked to Phil Archer who said we can refine the terms as we need

AndreaPerego: looking at the profiles use cases, we are using the word more than two ways -
… We need to align with the terminology we use in the UCR otherwise it will be challenging for readers

kcoyle: we need a way of distinguishing them

roba: at this stage we cannot make a call if they are different or not. If there are strong reasons for general and special views then that's OK, but we should not prejudge and we should do the work. Life would be simpler if they were the same thing (though different aspects)

<Zakim> SimonCox, you wanted to clarify how will we track and tag requirements?

roba: The ISO definition seems to be broad enough to handle all

SimonCox: I am concerned that we are getting into material that should be covered in subgroup.

<antoine> +1 with roba : in general I'm rather optimistic that we can get an agreement. I just think it will be super hard to do it quickly, and without first draft specs under our eyes.

SimonCox: I am concerned too about process - I started work on DCAT but had problems finding all documentation. In the last meeting we agreed to focus on github, yet we still have UCR in an HTML doc. We need to sort requirements to deliverables and this is challenging with the HTML doc esp as we agree to use github
… The subgroups will benefit from a clarification of process. I have started by suggesting to group leads that we convert the UCR into github

kcoyle: we agreed that we work on the documents in github, but not all our work will take place this way (e.g. google spreadsheet)

SimonCox: we are creating work by using too many tools and not using github as it can be used

roba: I dont disagree with Simon, but it would be good to chat to Jaroslav to see if the conversion can be automated. Doing it by hand would be too lengthy. We need to explore potential for scripting the parsing
… I will go with whatever we decide

kcoyle: roba will you check with Jaroslav?

<SimonCox> It is easy to generate a list of github issues tagged as 'requirement' for a single deliverable - e.g. https://‌github.com/‌w3c/‌dxwg/‌issues?q=label%3Arequirement+label%3Adcat

roba: if we decide to go with github then yes

<SimonCox> (currently empty)

kcoyle: we don't have much time - so a straw poll now?

<kcoyle> strawpoll: use github for discussing/deciding on requirements

<SimonCox> +1

<nicholascar> +

<alejandra> +1

<Makx> +1

<nicholascar> +1

<roba> +1

<antoine> +1


<LarsG> +1

<annette_g> +1

<DaveBrowning> +1

Action: roba contact Jaroslav re: scripting github issues from UCR

<trackbot> Created ACTION-72 - Contact jaroslav re: scripting github issues from ucr [on Rob Atkinson - due 2018-01-16].

<nicholascar> We use GitHub Issues in other WGs like RDA's

kcoyle: I will put the discussion on how we are defining profiles into the next meeting agenda

<LarsG> thanks, bye

<annette_g> thanks folks!

<alejandra> thanks and bye!

kcoyle: If you can formulate the question that the group has to decide on then please let everyone know

<AndreaPerego> Thanks, and bye!

<RubenVerborgh> bye!

<nicholascar> bye

Summary of Action Items

  1. roba contact Jaroslav re: scripting github issues from UCR

Summary of Resolutions

  1. approve minutes of previous meeting (Dec 12, 2017)
Minutes formatted by Bert Bos's scribe.perl version 2.37 (2017/11/06 19:13:35), a reimplementation of David Booth's scribe.perl. See CVS log.


Succeeded: s/Dubllin/Dublin/

Succeeded: s/#dxwg-dcat/#dxwgdcat/

Succeeded: s/using the word two ways/using the word more than two ways/