W3C home > Mailing lists > Public > public-dxwg-wg@w3.org > April 2018

Re: scope of profile (negotiation) group

From: Rob Atkinson <rob@metalinkage.com.au>
Date: Fri, 27 Apr 2018 16:15:37 +1000
Message-ID: <CACfF9LzszpT2rpBjwySKMP0tEQ+=b7h0-T6JZzS6hsecQQhTJA@mail.gmail.com>
Cc: Dataset Exchange Working Group <public-dxwg-wg@w3.org>
as modelled dct:conformsTo works fine as the predicate to link DCAT to a
profile..

I had a long chat with Simon about where this fits in, and he will come to
the F2F with a sense of its role in the wider challenge of dataset
description. Hopefully you can resolve the best "home" for this -
personally I don't mind and am willing to be the editor for a separate
deliverable.

Its distinct piece of work, but one that will IMHO help the "guidance for
DCAT profiles" deliverable - as it will enable formalism of the key
statements about profile hierarchies and different resource types we
already see in the DCAT profiling practices.

Rob





On 27 April 2018 at 14:56, Karen Coyle <kcoyle@kcoyle.net> wrote:

>
>
> On 4/26/18 12:36 AM, Svensson, Lars wrote:
> > On Wednesday, April 25, 2018 5:51 PM, Karen Coyle [mailto:
> kcoyle@kcoyle.net] wrote:
> >
> >> I'm reading the 'profile description' offered by Rob and Nick as
> >> 'metadata about the profile'. It gives what I would consider to be
> >> administrative and descriptive information about the profile, but is not
> >> itself an actionable profile.
> >
> > What exactly is an "actionable profile"?
>
> I believe it is the same as what you call a "schema". Something written
> in code that can be processed by programs.
>
> >
> >> The profile description links a DCAT
> >> expression to a profile, but is not either one. Is that a correct way to
> >> view it?
> >
> > I'm not sure I agree here. I thought what Rob and Nicholas have worked
> out links a "Profile" to a "Standard" it's a profileOf, and then it links
> the "Profile" to an "ImplementationResourceDescription" that describes
> things like XML Schemas or ShEx Documents that implement the "Profile". But
> maybe I got things mixed up again...
> >
>
> In my mind, what you say here is what the profile description *is* - it
> describes any standard that the profile is a profile of, it then
> identifies (and links) to any of the expressions of the profile. What
> I'm not sure of is whether there's a DCAT property that links from the
> DCAT description to the profile description - that is, if DCAT and
> profiles are linked through the profile description. I would really like
> to see a macro diagram that shows what (if anything) links this all
> together.
>
> >> If so, it's kind of a fourth deliverable, and to my mind could be
> >> associated either with DCAT or with the guidelines, as we prefer.
> >
> > I think we should be careful not to commit ourselves to too many
> deliverables...
>
> I agree. Which is why I'm asking where this profile description fits
> into the the deliverables that we already have.
>
> kc
>
> >
> > Best,
> >
> > Lars
> >
> >> On 4/25/18 2:21 AM, Antoine Isaac wrote:
> >>> Sorry Rob my inbox had messed up with your mail :-/
> >>> Good that we agree!
> >>>
> >>> Would it be possible to have the people working on profile description
> >>> as a subset of DCAT be also (or instead) attached to the profile
> >>> sub-group? That would make things more natural, i.e. the 'method and
> >>> technology' would be discussed in general not in the specific of DCAT I
> >>> know that you are not confused when you work on it (and I think that I
> >>> am not confused, anymore) but getting things a bit more formal and
> clean
> >>> may help a bit. If just by giving motivation for the profile
> >>> guidance/description work to progress. I.e it would exist with a
> >>> 'client' (i.e. DCAT) waiting for its progress.
> >>>
> >>> I'm also saying this because I've tried to join the profile
> >>> (negotiation) sub-group for the first time today and there was only
> Lars
> >>> and I. If DXWG creates a sub-group on profile guidance, or re-use and
> >>> extend the profile negotiation calls to discuss guidance/description,
> we
> >>> may need all the people working on related matters to be also formally
> >>> attached to that group, in order to get a critical size.
> >>>
> >>> Cheers,
> >>>
> >>> Antoine
> >>>
> >>> On 20/04/18 01:06, Rob Atkinson wrote:
> >>>> Thanks Antoine.
> >>>>
> >>>> I agree with you - its a separate sub-group who should in turn empower
> >>>> the (yet-to-form) guidance sub-group to explain how to simply handle
> >>>> profile creation and description in a Web friendly mechanism.
> >>>>
> >>>>   Note that the people working on profile description are more a
> >>>> subset of the DCAT group - but of course everyone is encouraged to
> >>>> engage because it seems we are all touched by the need to describe
> >>>> profiles :-)
> >>>>
> >>>> Rob
> >>>>
> >>>>
> >>>> On 20 April 2018 at 08:19, Antoine Isaac <aisaac@few.vu.nl
> >>>> <mailto:aisaac@few.vu.nl>> wrote:
> >>>>
> >>>>     Hi,
> >>>>
> >>>>     I agree that the vocabulary should be a part of the guidance on
> >>>> profiles, and that profile negotiation or dcat revision are not
> >>>> heavily impacted by the description issue.
> >>>>
> >>>>     Or at least they should not be heavily impacted. In fact this is
> >>>> perhaps where we could solve the issue that Karen noted ("profile" is
> >>>> intertwined both with DCAT and with content negotiation): we should
> >>>> make sure that the DCAT and content negotiation refuse to go into the
> >>>> details of guidance/description of profiles and just point to another
> >>>> area. For example the DCAT draft should try not to include the
> >>>> descriptions of profiles at
> >>>>     https://github.com/w3c/dxwg/tree/gh-pages/profiledesc/examples
> >>>> <https://github.com/w3c/dxwg/tree/gh-pages/profiledesc/examples> - at
> >>>> least not until the work is stabilized in another DXWG.
> >>>>
> >>>>     I guess the easiest way to do is to give a home in the group for
> >>>> that work - and for the one that Karen has just started on
> requirements.
> >>>>     Ideally it would be a separate, new sub-group, to make the
> >>>> difference clear.
> >>>>     However if the people working on guidance/description are very
> >>>> much the ones involved in the profile negotiation subgroup, it may be
> >>>> simpler to formally extend the scope of the negotiation group, so that
> >>>> it also includes profile/guidance as a second stream of work.
> >>>>
> >>>>     Cheers,
> >>>>
> >>>>     Antoine
> >>>>
> >>>>     On 19/04/18 00:12, Rob Atkinson wrote:
> >>>>
> >>>>         My own view is that a "profile description vocabulary" is a
> >>>> necessary part of guidance on profiles, a deliverable we have not yet
> >>>> started - it fills a gap in expression of the requirements.
> >>>>
> >>>>         I see that options 1&2 are the same in this context (because a
> >>>> profile is a resource with a URI) - and possibly with some additional
> >>>> best practice guidelines the proposed vocabulary could meet all the
> >>>> requirements in 3.
> >>>>
> >>>>         We have a definition - a model to formalise and explain, and
> >>>> worked examples to test should help us understand it better.
> >>>>
> >>>>         I dont think either profile negotiation or dcat revision are
> >>>> heavily impacted by the description issue - its "fine-grained
> >>>> semantics" - but that support for whatever forms of short identifiers
> >>>> needed for negotiation should be taken on as a requirement for the
> >>>> profile description language.
> >>>>
> >>>>         Rob
> >>>>
> >>>>         On 19 April 2018 at 02:06, Karen Coyle <kcoyle@kcoyle.net
> >>>> <mailto:kcoyle@kcoyle.net> <mailto:kcoyle@kcoyle.net
> >>>> <mailto:kcoyle@kcoyle.net>>> wrote:
> >>>>
> >>>>              Antoine, thanks, this is indeed what I hope we will have
> >>>> resolved by the
> >>>>              end of the f2f, but it could be very helpful to begin the
> >>>> discussion in
> >>>>              email and/or github.
> >>>>
> >>>>              I think what is tripping us up at the moment is that the
> >>>> concept of
> >>>>              "profile" is intertwined both with DCAT and with content
> >>>> negotiation,
> >>>>              but we do not yet have a clear definition of what we mean
> >>>> by profile. It
> >>>>              may be best to get clear on that before we talk about
> >>>> profiles in the
> >>>>              two contexts.
> >>>>
> >>>>              We have a base definition [1] which reads:
> >>>>
> >>>>              "A profile is a named set of constraints on one or more
> >>>> identified base
> >>>>              specifications, including the identification of any
> >>>> implementing
> >>>>              subclasses of datatypes, semantic interpretations,
> >>>> vocabularies, options
> >>>>              and parameters of those base specifications necessary to
> >>>> accomplish a
> >>>>              particular function."
> >>>>
> >>>>              This is a good start but we'll need to get into more
> >>>> detail before we
> >>>>              can resolve the larger issue that you bring up, and which
> >>>> I think is
> >>>>              about how we scope the concept of "profile". Here's a
> >>>> short list of what
> >>>>              I see as possible full definitions:
> >>>>
> >>>>              1. A profile is anything that meets the above definition
> >>>> and has a URL
> >>>>              (this is essentially Lars' proposal [2])
> >>>>              2. A profile is anything that meets the above definition
> >>>> and has a
> >>>>              (optional?) profile description (Nick & Rob's proposal
> [3])
> >>>>              3. A profile is anything that meets the above definition
> >>>> and all of the
> >>>>              approved requirements [4] [5]
> >>>>
> >>>>              I'll soon post something about the profile requirements
> >>>> which may help
> >>>>              us discuss this all further.
> >>>>
> >>>>              kc
> >>>>
> >>>>
> >>>>              [1] https://w3c.github.io/dxwg/profiles/
> >>>> <https://w3c.github.io/dxwg/profiles/>
> >>>> <https://w3c.github.io/dxwg/profiles/
> >>>> <https://w3c.github.io/dxwg/profiles/>>
> >>>>              [2] https://github.com/w3c/dxwg/issues/196
> >>>> <https://github.com/w3c/dxwg/issues/196>
> >>>> <https://github.com/w3c/dxwg/issues/196
> >>>> <https://github.com/w3c/dxwg/issues/196>>
> >>>>              [3] https://github.com/w3c/dxwg/
> tree/gh-pages/profiledesc
> >>>> <https://github.com/w3c/dxwg/tree/gh-pages/profiledesc>
> >>>> <https://github.com/w3c/dxwg/tree/gh-pages/profiledesc
> >>>> <https://github.com/w3c/dxwg/tree/gh-pages/profiledesc>>
> >>>>              [4] https://github.com/w3c/dxwg/issues/72
> >>>> <https://github.com/w3c/dxwg/issues/72>
> >>>> <https://github.com/w3c/dxwg/issues/72
> >>>> <https://github.com/w3c/dxwg/issues/72>>
> >>>>              [5] https://github.com/w3c/dxwg/issues/75
> >>>> <https://github.com/w3c/dxwg/issues/75>
> >>>> <https://github.com/w3c/dxwg/issues/75
> >>>> <https://github.com/w3c/dxwg/issues/75>>
> >>>>
> >>>>              On 4/18/18 7:42 AM, Antoine Isaac wrote:
> >>>>               > Hi everyone (esp Karen, Peter, Lars, Rob and Ruben)
> >>>>               >
> >>>>               > I'm considering trying to be more involved in the
> >>>> profile work, but I am
> >>>>               > not sure where I can fit in - and what are the
> >>>> responsibilities and scopes.
> >>>>               >
> >>>>               > It starts from the discussion we had yesterday on
> PR198:
> >>>>               > https://github.com/w3c/dxwg/pull/198
> >>>> <https://github.com/w3c/dxwg/pull/198>
> >>>> <https://github.com/w3c/dxwg/pull/198
> >>>> <https://github.com/w3c/dxwg/pull/198>>
> >>>>               > Apparently there is now a wiki page that says who
> >>>> would approve/merge it:
> >>>>               >
> >>>>
> >> https://www.w3.org/2017/dxwg/wiki/GitHub_etiquette#
> Contributing_to_the_normative_
> >> deliverables
> >>>>
> >> <https://www.w3.org/2017/dxwg/wiki/GitHub_etiquette#
> Contributing_to_the_normativ
> >> e_deliverables>
> >>>>
> >> <https://www.w3.org/2017/dxwg/wiki/GitHub_etiquette#
> Contributing_to_the_normativ
> >> e_deliverables
> >>>>
> >> <https://www.w3.org/2017/dxwg/wiki/GitHub_etiquette#
> Contributing_to_the_normativ
> >> e_deliverables>>
> >>>>
> >>>>               >
> >>>>               > There Lars, Rob and Ruben are indeed assigned to the
> >>>> object of PR198
> >>>>               >
> >>>> https://github.com/w3c/dxwg/blob/gh-pages/profiledesc/
> profiledesc.html
> >>>> <https://github.com/w3c/dxwg/blob/gh-pages/profiledesc/
> profiledesc.html>
> >>>> <https://github.com/w3c/dxwg/blob/gh-pages/profiledesc/
> profiledesc.html
> >> <https://github.com/w3c/dxwg/blob/gh-pages/profiledesc/profiledesc.html
> >>.
> >>>>
> >>>>               > But this ontology by Rob and Nick is not really about
> >>>> content
> >>>>               > negotiation - it's more about describing what is
> >>>> negotiated.
> >>>>               >
> >>>>               > On the other hand, the wiki page does not list Lars,
> >>>> Rob and Ruben as
> >>>>               > responsible of a document that shows them as editors:
> >>>>               > https://w3c.github.io/dxwg/profiles/
> >>>> <https://w3c.github.io/dxwg/profiles/>
> >>>> <https://w3c.github.io/dxwg/profiles/
> >>>> <https://w3c.github.io/dxwg/profiles/>>
> >>>>               > Actually I'm not sure what is the scope of this
> >>>> document: the title
> >>>>               > seems to hint that there is more than negotiation into
> >>>> it, while the
> >>>>               > content is still quite focused on negotiation, as
> >>>> Karen remarked in this
> >>>>               > issue:
> >>>>               > https://github.com/w3c/dxwg/issues/196
> >>>> <https://github.com/w3c/dxwg/issues/196>
> >>>> <https://github.com/w3c/dxwg/issues/196
> >>>> <https://github.com/w3c/dxwg/issues/196>>
> >>>>               >
> >>>>               > As noted in issue 196, I've tried to look through all
> >>>> our past minutes
> >>>>               > about organizing this work, and it's still not clear
> >>>> whether we want to
> >>>>               > have one deliverable on both negotiation and guidance,
> >>>> or two
> >>>>               > deliverables, and whether we should progress on both
> >>>> at the same time.
> >>>>               > And whether Lars, Rob and Ruben need help for what
> >>>> they are (perhaps
> >>>>               > informally) tasked to do!
> >>>>               >
> >>>>               > Hopefully the F2F (or perhaps even an earlier call?)
> >>>> will shed some
> >>>>               > light on all this.
> >>>>               >
> >>>>               > Cheers,
> >>>>               >
> >>>>               > Antoine
> >>>>               >
> >>>>               >
> >>>>
> >>>>              --     Karen Coyle
> >>>>         kcoyle@kcoyle.net <mailto:kcoyle@kcoyle.net>
> >>>> <mailto:kcoyle@kcoyle.net <mailto:kcoyle@kcoyle.net>>
> http://kcoyle.net
> >>>>              m: 1-510-435-8234 (Signal)
> >>>>              skype: kcoylenet/+1-510-984-3600
> >>>>
> >>>>
> >>>>
> >>>>
> >>>
> >>>
> >>
> >> --
> >> Karen Coyle
> >> kcoyle@kcoyle.net http://kcoyle.net
> >> m: 1-510-435-8234 (Signal)
> >> skype: kcoylenet/+1-510-984-3600
> >
>
> --
> Karen Coyle
> kcoyle@kcoyle.net http://kcoyle.net
> m: 1-510-435-8234 (Signal)
> skype: kcoylenet/+1-510-984-3600
>
>
Received on Friday, 27 April 2018 06:16:22 UTC

This archive was generated by hypermail 2.3.1 : Monday, 29 April 2019 13:44:58 UTC