– DRAFT –
(MEETING TITLE)
22 August 2018
<ncar> Agenda for this meeting is https://www.w3.org/2017/dxwg/wiki/Meetings:ProfGui-Telecon2018.08.21
ncar: discuss scope of the group
… how will group procede
… how to deal with requirements? list in the document?
… address them in the document
alejandra: question about scope - we are not supposed to do profiles; odd to produce document without some examples
… need to provide some idea of how it would be implemented
ncar: I do profiling in my work
https://www.w3.org/2017/dxwg/charter
kcoyle: charter says creating profiles is out of scope
roba: there are ranges of communities of practice for creating profiles
… guidance should be how to do it better; so we can look at existing examples
… geo-dcat, etc.
from the charter: "A definition of what is meant by an application profile and an explanation of one or more methods for publishing and sharing them."
... an analysis of what is happening in profile land and how we can support that process
... defining profile itself is idiosyncratic around community practices
... general process is not a DCAT profile
... look for best practices and requirements
... but not generic profiling language, but support discoverability, conneg, etc. and tying profile to resources
... some requirements will not have a best practice
ncar: have done profiles of ISO standards; we followed actions that we knew from before, but we had no guidance
alejandra: we already have a list of DCAT profiles
… we should reference existing practice
ncar: have profiles for iso standards as used in Australia
<alejandra> alejandra: I was proposing to collect a list of existing profiles, which nicar has already been compiling
ncar: rob could gather igc profiles
<alejandra> where is the outline?
<roba> and in the use cases : DCAT and europeana at least
<alejandra> https://w3c.github.io/dxwg/profiles/
kcoyle: need a wide range of profiles; e.g. from many communities
<alejandra> +1 to put them directly in the document
https://www.w3.org/2017/dxwg/wiki/ProfileRoundup
roba: if you list these then you have to show how they relate to each other
… need to describe them formally
<ncar> Profiledesc profiles examples: https://github.com/w3c/dxwg/tree/gh-pages/profiledesc/examples
https://github.com/w3c/dxwg/issues/242 for a variety of outline
roba: could be included in document
ncar: has pull request , kcoyle doesn't agree
ncar: listing of profiles; have a mechanism to describe profiles; and populated already
Action: ncar list other profiles in profileDesc space
roba: there are no perfect profiling languages, but still need to say what profiles exist
… what the profiles describe may come out of the requirements
… but not saying dataqube
… we have a metamodel for profiles; but we have to catalog the profiles first
<ncar> asck ncar
kcoyle: we have requirements that are about the functionality of profiles
ncar: in general, when you do profiles, this is what you can do (this would be the document)
… then you describe the environment
roba: some misconceptions; profdesc should not be in the guidance document? but we need to highlight best practices
… we will refer to shacl and shex; we can't give a recommendation to use those and ignore the rest of our requirement
… we can publish it saying this requirement will need to be met with an ontology
… nature of guidance doc is to tease out best practices
… do we leave it at that, or do we publish an ontology?
alejandra: what I'm hearing is that the charter gives us something very generic; guidance could just be a description, or it could go deeper
… or it could speak of ontologies
<alejandra> https://www.w3.org/2017/dxwg/wiki/ProfileRoundup
https://www.w3.org/2017/dxwg/wiki/ProfileRoundup
Action: ncar do examples for the plenary of options for profileDesc