Re: Profile definitions / DCAT update

1.       *DCAT vocabulary*   = "DCAT core".
I used the latter term to emphasise that DCAT vocabulary should only have
generic core terms

Agree also we dont need to define profiles, but IMHO its not useful to
leave it as vague wording about "write a document" - we need to provide an
example of a specific coded profile - even if we dont publish that as a
profile ourselves. Effective guidance means a tested mechanism that meets
requirements for profile specification, and has pedagogical examples.

Also I think there remains deep confusion about the nature of a profile -
it should describe one of many possible "interfaces" not a single concrete
implementation - it not analagous to a schema  - i.e. we could "suggest" a
profile that only specifies use of Prov-O to describe data provenance. We
don't need to boil the ocean by describing every aspect of every possible
use within a monolithic single profile, not withstanding the tendency of
people to create massive human readable documents and then clone and hack
these to specialise every single aspect in a single artefact. Thats just
getting sidetracked by the limitations of the documentation mechanisms (PDF
documents are inherently monolithic). We dont need to adhere to the
mistakes of the past, however well intentioned.






On Fri, 5 Jan 2018 at 12:49 <Simon.Cox@csiro.au> wrote:

> I tend to agree.
> Our job is to supply the tools.
> The communities will define their profiles.
> These might be done as community activities under W3C auspices, but most
> will likely not.
>
> Simon
>
> -----Original Message-----
> From: Annette Greiner [mailto:amgreiner@lbl.gov]
> Sent: Friday, 5 January, 2018 12:00
> To: public-dxwg-wg@w3.org
> Subject: Re: Profile definitions / DCAT update
>
> I don't think we are supposed to be proposing specific profiles. My
> understanding is that we are providing guidance as to how others can
> propose their own. In any case, making a single profile for all scientific
> data would be about as easy as (in homage to Phil) boiling the ocean. But I
> do want to ensure that profiles are something that scientific communities
> can use (and define for themselves).
>
> -Annette
>
>
> On 1/4/18 4:31 PM, Rob Atkinson wrote:
> >  i.e. should we propose a DCAT profile for scientific data (and then
> > delegate it to either a keen subgroup to add as a deliverable, or
> > describe the potential scope in the guidleline deliverable and then
> > park the issue)?
> >
> >
>
> --
> Annette Greiner
> NERSC Data and Analytics Services
> Lawrence Berkeley National Laboratory
>
>
>