– DRAFT –
Profile Guidance doc St Valentines Day Sprint

<kcoyle> I'll send you the link on skype

<ncar> hi

<kcoyle> https://‌w3c.github.io/‌dxwg/‌profiles/

<kcoyle> https://‌docs.google.com/‌document/‌d/‌1Y4jP4SGZMnt63EpjTX11-hW6-3mxlaq1i-Lbiw4tx1M/‌edit#

<kcoyle> https://‌www.w3.org/‌2017/‌dxwg/‌wiki/‌Meetings:ProfGui-Telecon2019.02.14

<kcoyle> https://‌www.w3.org/‌2019/‌01/‌23-profgui-minutes.html

kcoyle: we didn't create an action for the resolution for the last resolution
… about creating a milestone

kcoyle: has anyone changed this milestone?

Action: kcoyle to update milestone as per last week's minutes

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

<alejandra> https://‌github.com/‌w3c/‌dxwg/‌milestone/‌10

Resolved: approve minutes of 1/23

Proposed: that we approve minutes

<PWinstanley> +1

+1

<PWinstanley> antoine: we are approving minutes from 3 wk ago? there were minutes from the one roba scribed.

<PWinstanley> ... this other was a sprint, but minutes need approval too

https://‌www.w3.org/‌2019/‌02/‌06-profgui-minutes.html

<roba> +1

Resolved: approve minutes from 6 Feb

<PWinstanley> antoine: section1 - minutes of the last call, action on me to incorporate

<PWinstanley> ... edits into html master

<PWinstanley> roba: I looked at it and it was OK

<kcoyle> https://‌w3c.github.io/‌dxwg/‌profiles/

<PWinstanley> kcoyle: so section 1 shows the results of the last discussion, and there have subsequently been more changes, but we press on

kcoyle: we should rather dive in section 2 now, we can come back to the details of the new suggestions for section 1

<PWinstanley> ... and return to the more recent changes in due course

[kcoyle goes through the intro of section 2]

<PWinstanley> antoine: asking roba what he means in his comments

<PWinstanley> roba: we are not taking responsibility for every use of the word profile in the outside world

kcoyle: "outside of this document"?

roba: maybe "outside of this context"

PWinstanley: "in other contexts" - we want to claim our context

Discussion on keeping MUST/SHOULD/ETC etc

We agree to keep them

roba: this section is about constraints

kcoyle: this requirements is not about constraints

(requirement #275)

<roba> +1 that the language used actually has no meaning without context..

PWinstanley: it's formal, but we should try to have a introduction

<roba> if we are not talking about "named graphs" then we need to define all the connecting words in terms of the agreed definition of a profile - not introduce new terminology

<roba> "collection of properties" needs to have a meaning w.r.t. to the definition - and its not obvious or defensible as is

Action: PWinstanley will rewrite the beginning of 2.1

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

<roba> remove "status of Standard" - we are based on the very loose dct:Standard which has no requirement for a formally defined status ..

kcoyle: discussing the second paragraph with "Distribution:

<PWinstanley> antoine: I think I was the one writing it - the paragraph was to be a guide to this document - there would be hyperlinks to different sections

https://‌www.w3.org/‌TR/‌skos-primer/#secsimple

kcoyle: ok we move it back up

kcoyle: section 2.2 (multiple) base specifications

roba: specification is strightforward
… this paragraph is quite RDF: the re-use of elements

kcoyle: it's also XML

roba: ok but it's also more than re-using vocabularies

<roba> +1

alejandra__: yes having first a general wording is good and then we can go into an example like the RDF and XML one

+1

Action: alejandra will re-word the intro for 2.2 and issue 268

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

alejandra__: have we kept "base specification"?

roba: yes we've removed it as a class in PROF (as it was more a role) but we've kept the wording.

kcoyle: 2.3 profiles or profiles

roba: we've introduced "level" but there might not be anything as a "level of profile"
… we could remove the first line

<roba> what is being left open?

antoine: issue 270 is really about saying that it;'s ok to have several levels
… the other paragraph is about a more general issue

<roba> yes!

[discussion on mutliple base specifications and hierarchical profiling. We agree and need a diagram]

<alejandra__> +1 to DAG

<PWinstanley> antoine: I was volunteering for diagram duty

<alejandra__> +1 to add the diagram in the document

<PWinstanley> +1 to non-technical

Action: Antoine to create diagram(s) for the different possiblilty for profiles, on 2.2 and 2.3

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

<roba> https://‌github.com/‌w3c/‌dxwg/‌issues/‌641

<PWinstanley> I need my friends to be able to understand it

kcoyle: profile inheritance (2.4)
… it connects to ODRL

roba: can we say that profile should indicate how formal semantics are inherited?

kcoyle: it comes from external comments.

<PWinstanley> antoine: at this stage let's put it as an open issue and return to it

<ncar> I was first!

<ncar> * no, other relevant comment about procedure of Issues

<alejandra_> +1 to roba comment's about inheritance

ncar: the discussion on inheritance relates to a PR for PROF, on examples of inheritance
… if it's accepted people will see

<PWinstanley> antoine: reacting to ncar , having been involved in the discussion, the example deserves some improvement, but there should be a proper issue for the requirement [github issue is missing]. I agree with roba , and if he can open an issue for the explanation of inheritance it would be great

Action: roba to create a requirement issue for inheritance

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

Action: Antoine to relate other issues from external comments on inheritance to Rob's requirement, after he has created it

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

kcoyle: Data publication according to different profiles

<roba> dcat

kcoyle: my take is that it's conneg

roba: it's a requirement for DCAT expressivity
… provide a mechanism to express it

https://‌github.com/‌w3c/‌dxwg/‌issues/‌608

<PWinstanley> antoine: pasting a link to the requirement that indicates that we decided to create it as a general fork of something else, in the profile category. It illustrates the need for expressivity - a mechanism in conneg shows that there can be many satisfying profiles and one can be chosen. We want this to be a general feature

<PWinstanley> ... and I would prefer that not all examples come from DCAT

roba: the documents are not disjoint, and the guidance can be the place where things are put together
… so DCAT can be one mechanism, but there can be others.

kcoyle: we don't say how data relate to profiles

ncar: we had feedback on this from external comments
… there are examples that we have to align with

alejandra: I agree it makes sense to put it in the guidance
… it's about linking profiles to data

kcoyle: 2.5 and 2.6 are very similar
… it seems we need to make a single section about data and profiles

roba: I can see the difference but not in the wording
… difference between conformance of the data and the access to the data

<PWinstanley> antoine: I'm fine if roba starts it. I'll check that what he writes fits with #274 ... schema.org and DCAT compatibilty

<PWinstanley> ... data according to two profiles in one distribution

kcoyle: there are notes that can be useful
… and then a part on validation. We can see where it belongs, in the end.

<PWinstanley> antoine: should we try to go to the suggestions for section 1

<PWinstanley> ...?

<PWinstanley> antoine: it reads well, but do you want it as an editorial note? Both sentences? or as a regular paragraph?

<PWinstanley> antoine: no pull requests now please.

<PWinstanley> ... I agree with roba original comment on a base specification and a profile.

<roba> * what are we looking at Peter? the edits to section 1?

[discussion on merging 2.2 and 2.3. we agree to wait until next time - and the diagrams]

PWinstanley: [introduces his proposal for 2.1]

PWinstanley: the Baltimore classification https://‌en.wikipedia.org/‌wiki/‌Baltimore_classification could be used as a guide
… it tells the rules to compose and match DNA

kcoyle: feel free to make suggestions

PWinstanley: it's about patterns/strategies

<roba> +1 for these edits..