W3C

Data Exchange Working Group teleconference

07 August 2017

Meeting Minutes

present?

<Caroline> https://‌www.w3.org/‌2017/‌07/‌31-dxwg-minutes

<SimonCox> I'm very awake - I just played a game of Ultimate!

<annette_g> +1

<Caroline> +1

Caroline: let’s approave last week’s minutes. Any objections?

+1

<roba> +1

<SimonCox> +1

Resolved: Approved minutes from 31 July

<Caroline> dsr: About the Github repository

<Caroline> ... people can send Dave their emails and he will add them as dxwg members there

<Caroline> ... actually send their github usernames

<Caroline> ... about the Webex we discovered that the system could be used for other people and there was an abuse of the system

<Caroline> ... because of that we are not suppose to share the links

<Caroline> ... I sent an email to everyone witht the Webex link

<Caroline> ... those who have members access maybe access it and the invited experts must go throgh the link I sent by email

use cases group report

We were discussing whether to use github or the wiki for the use cases

alejandra: I am okay with either option, but it would be harder to update both places. I suggest that we use github issues for tracking contributions.

My question is when we need to discuss something in the group versus just with the document editors?

kcoyle: what we’ve been doing so far, is to discuss new use cases as a group and as a group to decide whether it is in scope, in which case it is added by the editors to the document

roba: using github issues makes sense for new use cases and pull requests for later modifications to the github document

Caroline: I agree that using github issues is the way to go. I don’t think everyone needs to be preocupied about the respec but can use the github, as the editors may find that easiest to do based upon the feedback they get

kcoyle: Is there a way to duplicate the wiki template for people to use in the github issues?

roba: you can point to the wiki for the templae

<alejandra> It is possible to create an ISSUE_TEMPLATE: https://‌github.com/‌blog/‌2111-issue-and-pull-request-templates

<alejandra> +1 to dsr, that is described in the document I linked to

<SimonCox> [I just transferred the GitHub Etiquette proposal to https://‌www.w3.org/‌2017/‌dxwg/‌wiki/‌GitHub_etiquette]

dsr: we could use a markdown template given that markdown is allowed in github issues

roba: given the need to transcribe the use cases, I don’t think this is too big an issue

Caroline: we need to clearly document the process for people to follow

kcoyle: I volunteer to write up the github process

<kcoyle> PROPOSED: use github issues to add new use cases to the list

<alejandra> +1

<roba> +1

<Caroline> +1

<annette_g> +1

<riccardoAlbertoni> +1

+1

<SimonCox> +1

<kcoyle> +1

<DaveBrowning> +1

<Ixchel> +1

Resolved: use github issues to add new use cases to the list

dsr: suggests writing the process as a markdown document in the repository, as github renders this automatically as rich text

<riccardoAlbertoni> +1 to dsr

Ixchel: provides a quick status report and will be in touch with roba about her progress

What process are we using here for decisions, is it a simple majority vote?

Caroline: we’re using votes to gauge whether we have a rough concensus

We first need to agree on adding the use case to the one’s we’ve accepted, and we then need to agree on the requirements that emerge from the accepted use cases

kcoyle: the voting process is W3C specific, where we use +1 for yes, 0 for don’t care and -1 for disagree. If someone disagrees we need to discuss the issue as a group, and if it can’t be resolve the chairs need to decide on how to progress

roba: I just want to reitterate that there will be a fair bit of work to identify the requirements and link them back to the use cases. We will need to work carefully on the wording of the requirements.

This will involve some de-duplification

Caroline: thanks for pointing that out

SimonCox: Each W3C group has a certain amount of latitude to determin the details of its process, but the important point is to be transparent

<Caroline> dsr: W3C focus on consensus

<Caroline> ... we try to understand the objections and resolve them

<Caroline> ... if it is not possible to resolve them we can document them as part of the report as a group

<Caroline> ... ex. the 1st public WG draft can note it

dsr: W3C works on a rough consensus basis, and there is a provision to record minority opinions in our reports when these can’t be resolved within our discussions

<SimonCox> It ain't summer holidays here!

alejandra: several people are away currently so let’s keep the discussion in this call for now

SimonCox: as one of the nominated editors, I have a certain level of reluctance until the discussion of the use cases and requirements is a little further along. The spreadsheet has been helpful

If we find some holes, filling them will require further discussion

alejandra: I will be away as well towards the end of August, so we may find it better to defer starting separate meetings until next month.

I have started a google doc, and can give people write access if they send me an email to that effect

<alejandra> https://‌docs.google.com/‌spreadsheets/‌d/‌12t6a9seTmCi47p6k8HH1Mz4k2CedhZSZ9sbUpaRrQO8/‌edit#gid=0

dsr: the url for the document

alejandra: public has the ability to read and add comments, but not to update

Caroline: any other comments ?

[no]

Open actions

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

<Caroline> https://‌www.w3.org/‌2017/‌dxwg/‌track/‌actions/‌24

We can close 24 now given Dave’s report earlier

<alejandra> for the DCAT application profiles document (https://‌docs.google.com/‌spreadsheets/‌d/‌12t6a9seTmCi47p6k8HH1Mz4k2CedhZSZ9sbUpaRrQO8/‌edit?userstoinvite=dr.shorthair@gmail.com&ts=59887a50&actionButton=1#gid=0), do send me request for access if you wish

In respect to action 29, Dave needs further details before he can ask Ivan

kcoyle: we may not need that the real work will be taking the existing use cases and consolidating them.

kcoyle: we can therefore close action 29

<Ixchel> what about those pending review

<Caroline> https://‌www.w3.org/‌2017/‌dxwg/‌wiki/‌Use_Case_Working_Space#ID3

We’re missing the people for the other open actions on this call.

Ixchel: I noticed from the minutes that some of the use cases are still open for pending review. Who has the responsibility for closing actions the related actions - the chairs or the editors?

kcoyle: actions 4, 6 and 7 I can close, but there are several others relating to specific use cases.

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

Caroline: we didn’t decide that, and the idea is to discuss them during the meeting. I think the editors can decide when to close specific actions on use cases pending review, Karen is that okay with you?

kcoyle: I would like to have some time to review the comments. Let’s talk about this amongst ourselves and aim for a resolution next Monday

RubenVerborgh: in respect to use case 3, this is about that a single response can correspond to multiple profiles

Caroline: can you explain further?

RubenVerborgh: a document could conform to both specific and generic profiles

Use case 3 says we should have a means to indicate this

<alejandra> https://‌www.w3.org/‌2017/‌dxwg/‌wiki/‌Use_Case_Working_Space#Responses_can_conform_to_multiple.2C_modular_profiles_.5BID3.5D

kcoyle: I think we need to be careful to avoid mixing up serialisation and content, as it does seem to right now

<Zakim> RubenVerborgh, you wanted to comment on serialization and contetn

RubenVerborgh: I agree, that was just an example

annette_g: what is the problem we’re trying to solve here?

I am not sure that this is something that users will want to ask for

<Zakim> RubenVerborgh, you wanted to explain problem we're solving

RubenVerborgh: one possible example is where a request is in one profile and the response also conforms to a more specific profile, we want to ensure interoperability

annette_g: so this is a case with a nested profile?

RubenVerborgh: it could be, but doesn’t need to be hierarchical and could be a combination instead

annette_g: I want to push back on hierarchal profiles

RubenVerborgh: being able to name the profiles could avoid a combinatorial explosion

roba gives some examples of parallel, polymorphism and deep nesting …

<RubenVerborgh> Just want to emphasize here that UC3 is independent of nesting, i.e., agreeing on UC3 does not mean agreeing/disagreeing on nesting.

We need to be able to declare multiple profiles for each catalogue

annette_g: I was thinking of the rationale for using content negotiation

roba: content negotiation should be pretty easy to handle

alejandra: I am in agreement with the use case, but we could clarify it

I like the idea for a server to indicate conformance to a profile. Do we want to make this specific in terms of Linked Data shape rules?

roba: I am reasonably comfortable that the issues have been addressed, but we have some more editing to do

annette_g: we need to separate the firm requirements based upon what’s needed, we will end up with a better end result

i.e. to drive the discussion from the use cases

<Caroline> https://‌www.w3.org/‌2017/‌dxwg/‌wiki/‌Use_Case_Working_Space#ID3

<Caroline> PROPOSED: accept the use case ID3 https://‌www.w3.org/‌2017/‌dxwg/‌wiki/‌Use_Case_Working_Space#ID3

roba: we haven’t had any push back on the use case, so I would recommend we accept this use case

<SimonCox> +1

<Caroline> +1

<roba> +1

<RubenVerborgh> +1

+1

<kcoyle> +1

<alejandra> +1

<riccardoAlbertoni> +1

<DaveBrowning> +1

<annette_g> +0

<kcoyle> and maybe note particular care needed on requirements editing

Resolved: accept the use case ID3 https://‌www.w3.org/‌2017/‌dxwg/‌wiki/‌Use_Case_Working_Space#ID3

Caroline: we will continue with the remaining use cases on the next call

<roba> bye

<alejandra> thanks and bye!

<riccardoAlbertoni> bye thanks

Summary of Resolutions

  1. Approved minutes from 31 July
  2. use github issues to add new use cases to the list
  3. accept the use case ID3 https://‌www.w3.org/‌2017/‌dxwg/‌wiki/‌Use_Case_Working_Space#ID3
Minutes formatted by Bert Bos's scribe.perl version 2.25 (2017/05/19 13:30:32), a reimplementation of David Booth's scribe.perl. See CVS log.

Diagnostics

Succeeded: s/Resolve/Resolved

Succeeded: s/to generate pull requests/to be preocupied about the respec but can use the github

Succeeded: s/group report/use cases group report/

Succeeded: s/uses/users/