DXWG Telecon2018.06.26

26 June 2018

Meeting Minutes


PROPOSED: approve minutes of June 19

<PWinstanley> https://‌www.w3.org/‌2018/‌06/‌19-dxwg-minutes


<annette_g> +1

<LarsG> +1

<PWinstanley> +1

<Jaroslav_Pullmann> +1

<roba> +1

Resolved: approve minutes of June 19

open actions

<PWinstanley> https://‌www.w3.org/‌2017/‌dxwg/‌track/‌actions/‌94

PWinstanley: kcoyle - # 110

kcoyle: no result yet - can't find a better wording

PWinstanley: dsr - add annette to march 13

PWinstanley: LarsG can close 131 133

antoine: riccardo has complete 126; can be closed

kcoyle: action items closed

PWinstanley: note deadlines - publishing moratoria of w3c; deadlines are listed in the agenda. July4, July 25, plus around TPAC

PWinstanley: any estimates on next drafts? will discuss in subgroup reports

sub-group reports

PWinstanley: DCAT: was a small meeting so didn't do open issues
… talked about approach toward soliciting feedback from fpwd
… concerned about lack of response
… also heard from Stijn on positioning of DCAT vs other similar ones
… need to have more info in the draft about the motivation for DCAT vs the others
… based on broad interoperability
… we need to be able to show that people are looking at what we are doing
… need to keep up advertising, contacting folks for feedback
… otherwise unable to show that recommendation has been viewed by community

antoine: there is feedback that we have received and it needs to be answered

PWinstanley: I'll make sure that gets onto agenda for DCAT

Jaroslav_Pullmann: Do we have a comparison of DCAT vs the other related standards?

PWinstanley: We have a listing on home wiki page; there was something that makx reported to in an EC report
… but don't know of a comparison

Jaroslav_Pullmann: suggest that people create such a page

<AndreaPerego> About relationship with ISO 19115 (and SDMX), see the discussion in https://‌www.w3.org/‌2018/‌06/‌26-dxwg-minutes

PWinstanley: ISO 19115

<AndreaPerego> Sorry, the discussion is here: https://‌www.w3.org/‌2018/‌06/‌21-dxwgdcat-minutes

<PWinstanley> A report was published recently with an analysis of research metadata standards in relation to DCAT-AP, including CERIF. https://‌joinup.ec.europa.eu/‌document/‌research-data-analysis.

roba: 19115 is xml-based; but hard to profile xml schemas; results are very convoluted

PWinstanley: profile group

LarsG: there is some conflation between profile guidance and negotiation; hard to separate them

PWinstanley: do you think we should join the groups together?

LarsG: the same people were on each call

<Jaroslav_Pullmann> .. there is already a section on standards related to DCAT: https://‌www.w3.org/‌2017/‌dxwg/‌wiki/‌Main_Page#Related_non-W3C_vocabularies

AndreaPerego: answering Jaroslav_Pullmann about positioning of DCAT vs others; there is discussion in the minutes of the last call
… AndreaPerego has an action to draft something on this for the specification
… that's the last DCAT subgroup call

antoine: +1 for merging groups

<PWinstanley> kcoyle: afaik the profile guidance group hasn't met, does't have a chair. There is overlap in discussion, but ....

<PWinstanley> there hasn't been because we are going through the requirements

kcoyle & Roba - profile guidance hasn't met

<antoine> roba's recollection seems correct to me.

<PWinstanley> ... I have no problem with merging the group as long as we don't lose the ability to cover the topics

LarsG: discussed requirements from Google doc; proposed some wording changes
… those are on the agenda for this meeting
… how do we get re-worded requirements into the UCR?
… once accepted there should be an action on someone to generate a pull request on UCR document. Is that the right process?

roba: putting together UCR was a huge job to dedup etc. - UCR group should provide format support not content support
… don't want to repeat the past

<Zakim> antoine, you wanted to suggest to wait and see

LarsG: my action was to bring it to this group; makes sense to not have UCR group do all of the editorial work

antoine: I would suggest that we not worry about this until we finish the requirements

PWinstanley: be pragmatic and practical?

antoine: there may be other changes so it should wait until the end

Jaroslav_Pullmann: editors could take what comes out of discussions; they should pick up the results of the discussion here

PWinstanley: is this consensus?

<antoine> I'm fine with it - if editors take their time and wait until we're finished with the current effort :-)

PWinstanley: can we combine the two groups?

<PWinstanley> kcoyle: I see some dangers of putting them together - there are different functionality (technical in negotiattion, but best practice in the profiles)

<PWinstanley> ... so although we discuss what a profile 'is' they are different topics

roba: tend to agree but think we can wait and see; discussion in conneg group weighs in on nature of profiles
… and the ability to describe profiles; but otherwise wait until requirements are done
… have to re-org anyway

<PWinstanley> ... interested parties can join in as appropriate

<PWinstanley> kcoyle: If there is an agenda posted before the meetings then people can join in at the right points

PWinstanley: ID42 https://‌github.com/‌w3c/‌dxwg/‌issues/‌255

<PWinstanley> Requirement: There needs to be a property in the profile where the rules for the descriptive content can be provided. This would apply to the entire profile. [ID42] (5.42) (github)

<antoine> +1



<PWinstanley> +1

<LarsG> +1

<annette_g> +1

<AndreaPerego> +1

<Jaroslav_Pullmann> +1

roba: this is a duplication

PWinstanley: we can allow duplicate

antoine: this is #19, and it has been re-written

<roba> +1

Resolved: in scope There needs to be a property in the profile where the rules for the descriptive content can be provided. This would apply to the entire profile.

PWinstanley: ID2 - Requirement: Clients should be able to determine which profiles are supported by a server, and with which content types.

PWinstanley: also: Requirement: There should be a way for a client to look up additional information about a profile. for ID2
… reworded as "There should be a way to look up additional information about a profile - this may be machine readable for run-time mediation or used to develop or configure a client".
… rewording done by conneg group in their meeting

<ncar> +1

<PWinstanley> +1

PROPOSED: accept both ID2 requirements

<roba> +1

<AndreaPerego> +1 (although it would be better to specify what we mean with "additional information")

<LarsG> +1

<Jaroslav_Pullmann> +1

<antoine> +1

+1 (but agree with AndreaPerego - that could be an action)

<PWinstanley> +1

<annette_g> +1 under the condition that we aren't specifying it be through conneg

ncar: we purposely left "additional information" open because it would have been huge

<annette_g> :)

<annette_g> I mean the first one

PWinstanley: yes, to annette, not necessarily conneg

Resolved: ACCEPT both ID 2 requirements

annette_g: which item are we talking about - my concern was about the first one - that it not be limited to conneg
… needs to be supported not just within the header

LarsG: there are ways to resolve this requirement but conneg is not the only way

roba: it's a requirement for content negotiation; that info should be accessible in other ways

ncar: we've discussed this a lot; we want to be sure that if there are different ways to do things they are aligned

<annette_g> yes

<PWinstanley> Profiles must support discoverability via search engines (UC 5.40) #222 (Github discussion) ID40 (5.40)

antoine: we found this was a dcat requirement

<antoine> discussion was at https://‌github.com/‌w3c/‌dxwg/‌issues/‌222

roba: this is about DCAT

PWinstanley: Fix our agenda and bring this back next week - something copied wrong

<PWinstanley> Requirement: Profiles can be modular, with a given response made up of more than one module. Servers can indicate that a response conforms to multiple, modular profiles. ID3 (5.3) [conneg] [profile]

PROPOSED: accept Requirement: Profiles can be modular, with a given response made up of more than one module. Servers can indicate that a response conforms to multiple, modular profiles. ID3 (5.3

annette_g: discussion on github; this is about a single profile is pulling together elements from other profiles; if so, I support it

<AndreaPerego> Is this feasible with conneg?

antoine: annette_g's reading is partly right; it is even more general - conformance may not be through a mediating profile
… it could be just a gathering of data that conforms to several profiles directly

AndreaPerego: unclear how this can be implemented with conneg. is this going too far?
… is this feasible?

PWinstanley: could the group discuss feasibility while still considering the request valid

antoine: suggested: Some data can conform to multiple profiles at the same time

annette_g: in the simplest form a profile can take definitions from other schemas; and can be presented as flat

<Zakim> LarsG, you wanted to ask if antoine's proposal really is a requirement

LarsG: antoine's proposed wording isn't a requirement, just a fact of life

roba: we came up with a different wording of that. let's take this offline and look for the re-wording

PWinstanley: this brings up the question of inheritance;
… given the time we have to take this one offline

<Zakim> LarsG, you wanted to discuss cneg subgroup

<Jaroslav_Pullmann> bye!

<AndreaPerego> Thanks, and bye!

<ncar> bye

Summary of Resolutions

  1. approve minutes of June 19
  2. in scope There needs to be a property in the profile where the rules for the descriptive content can be provided. This would apply to the entire profile.
  3. ACCEPT both ID 2 requirements
Minutes formatted by Bert Bos's scribe.perl version 2.41 (2018/03/23 13:13:49), a reimplementation of David Booth's scribe.perl. See CVS log.


Succeeded: s/decided/found/