DXWG DCAT subgroup teleconference 30 January 2019 21:00 UTC

need to clarify documentation that accrualPeriodicity gives the maximum temporal resolution available for a Dataset, but Distributions might have different spacings

Action: DaveBrowning to respond to Pop

<trackbot> Sorry, but no Tracker is associated with this channel.

Hello AndreaPerego can you report on any actions https://‌www.w3.org/‌2017/‌dxwg/‌track/‌actions/‌open

... do we still have to respond to Ine?

<alejandra> https://‌www.w3.org/‌2017/‌dxwg/‌track/‌actions/‌278

<alejandra> https://‌lists.w3.org/‌Archives/‌Public/‌public-dxwg-wg/‌2019Jan/‌0363.html

AndreaPerego: importnat comments on introduction of Services to DCAT
… some happy, some see it as disruptive to their current way of doing things
… worried that they suddenly become non-conformant

<alejandra> The big note about the slight change of semantics of Distributions is in the distribution section: https://‌w3c.github.io/‌dxwg/‌dcat/#Class:Distribution

<Zakim> alejandra, you wanted to ask about services implementations and background compatibility re: distributions

<PWinstanley> +1 to alejandra - we cannot make the previous inferences with the current model

alejandra: it can be argued that DCAT-rev is not backward compatible
… the revision disentangles an unhelpful confusion in scope of dcat:Distribution
… need to find more support for the change, to demonstrate implementations of the new classes
… e.g. Amsterdam City, others in Europe - we should contact them to encourage implementation
… also to clarify which UC they are responding to

DaveBrowning: make warning note more obvious, maybe repeat in dcat:DataService definition
… suggest that people use profiles to say (I'm using "DCAT 2014" interpretation of dcat:Distribution)

<PWinstanley> but many people start with keyword searches and in the new model that is ambiguous

AndreaPerego: 2 aspects of changes: 1. Services are first-class catalogable entities
… it is a choice of catalog implementer whether they include services or not
… CKAN does not distinguish datasets and services
… should we work with CKAN devs to add 'Service'
… 2. link from Service --> Distribution is not well understood
… people worried that they can't continue to tie these together in the same way
… 'this is not a file, this is a service, so you have to do more than just click on the link'
… note that the instructions in DCAT-rev uses SHOULD not MUST on this topic so legacy is protected that way too

alejandra: yes - this last point is key - it is strictly backward compatible, but maybe need to add some guidance on how to differentiate the two styles of use
… how to recognise new representation? It uses the new classes and properties!

AndreaPerego: what is happening now:
… 1. GeoNetwork - can already recognise if a URL points to a service or a file
… 1a. CKAN Geospatial - uses format information to detect a WMS
… but these are guesses that depend on the platform (geospatial)
… would be improved by making it explicit! i.e. dcat:DataService and sub-classes
… the missing piece to distinguish services (WMS, SPARQL endpoint) from distros (files) is exactly what is addressed in the DCAT-rev proposal for services
… should we contact CKAN people? maybe not, they are already responding to what they see in DCAT APs
… need to focus on the community (customers) not the tool providers

DaveBrowning: good angle on text to be added
… 2014 users need guess-work to succeed. -rev is enabling catalog managers to be more clear, thus better for their customers
… aligning with standard is irrelevant unless customers are actually demanding it
… actually we are not hurting you here, it is a better solution if you can adopt it, but as-you-were if you are not in a position to adopt it now

<AndreaPerego> s/serv ice/service/

DaveBrowning: maybe this is part of Distributions spring?

<alejandra> it doesn't hurt that Andrea starts looking at this and then we discuss the new text in the sprint

AndreaPerego: people don't read everything - the note on scope of dcat:Distribution just needs to be moved to a more prominent place, perhaps in the introduction

<alejandra> actually, it would be helpful

<DaveBrowning> +1 to alejandra

Action: AndreaPerego to propose changes to add a note re legacy use of dcat:Distribution for services more prominent higher in rec

<trackbot> Sorry, but no Tracker is associated with this channel.

<AndreaPerego> +1

alejandra: 1. license and rights information everywhere

<alejandra> https://‌github.com/‌w3c/‌dxwg/‌pull/‌717

2. too much inheritted material is repeated in discussion of catalog class

<alejandra> https://‌github.com/‌w3c/‌dxwg/‌issues/‌704

DaveBrowning: some of the text is tweaked for context ...

alejandra: however, needs reader to do more work, and interpret semantics of property in context

sprint planning

DaveBrowning: 1st sprint next Tuesday(/Wednesday) on Distributions

<DaveBrowning> distribution related issues are https://‌github.com/‌w3c/‌dxwg/‌issues?q=is%3Aopen+is%3Aissue+label%3Adcat%3ADistribution

all sprint participants should review this list of issues before next week

alejandra: OK for us to skip plenary for sprint, but do we also have regular wednesday(/thursday) meeting?

<AndreaPerego> +1

<PWinstanley> +1 - keep regular

<alejandra> +1

+1 to keep meeting

DaveBrowning: regular meeting is for progressing checking, spring is for focusing on a feature

<alejandra> SimonCox: should I set up a webex for the sprint meeting?

Action: SimonCox to set up webex for sprint

<trackbot> Sorry, but no Tracker is associated with this channel.

<alejandra> +1

Action: DaveBrowning to send out announcement of sprint meeting

<trackbot> Sorry, but no Tracker is associated with this channel.

PWinstanley: maybe use a google doc during the meeting which we can collaboratively edit
… paste into the real place immediately following

+1 to collaborative editing in real time, github for asynchronous

... style editing

<AndreaPerego> Issue with references: https://‌github.com/‌w3c/‌dxwg/‌issues/‌702

<trackbot> Sorry, but no Tracker is associated with this channel.

AndreaPerego: references are being sorted into the wrong sections (normative vs non-normative)
… October change to ReSpec - looks like a bug?
… respec is pretty brutal on section tagging
… we need to add ? or ! to all refs
… and class='normative' vs 'non-normative' on sections

e.g. ref Linked Data Patterns is listed in Normative - shouldn't be!

DaveBrowning: has marked some issues 'due for closing' please all review so we can do some cleanup!
… editors prerogative to close issues, they can always be re-opened

– DRAFT –
DXWG DCAT subgroup teleconference 30 January 2019 21:00 UTC

30 January 2019

Meeting minutes

approve agenda

approve minutes of last meeting

actions

public comments

sprint planning

Summary of action items

Summary of resolutions

Diagnostics