Meeting minutes
<RiccardoAlbertoni> RSAgent, draft minutes
<RiccardoAlbertoni> RSAgent, draft minutes
<RiccardoAlbertoni> https://meet.google.com/
<RiccardoAlbertoni> PROPOSED: approve last meeting minutes https://www.w3.org/2020/09/16-dxwgdcat-minutes
<AndreaPerego> +1
proposed: accept minutes of last meeting
<RiccardoAlbertoni> +1
+1
Resolution: accept minutes of last meeting
<RiccardoAlbertoni> https://www.w3.org/2017/dxwg/wiki/Meetings:Telecon2020.09.30
Is anyone interested in doing anything over the TPAC meeting events?
RiccardoAlbertoni: Time is scarce for me. But I can contribute
AndreaPerego: Would we like to join in with other groups?
AndreaPerego: I'm not sure if there are any relevant groups.
RiccardoAlbertoni: I'm proposing not to participate, but to try to get a publication and a presentation (DCMI community)
PWinstanley: I think there are some W3C talks scheduled for November - will need to investigate more
Versioning PR
<RiccardoAlbertoni> https://github.com/w3c/dxwg/pull/1257
<AndreaPerego> Preview here: https://raw.githack.com/w3c/dxwg/dcat-versioning/dcat/#dataset-versions
RiccardoAlbertoni: it deals with many of the issues that we have had re: versioning, and I think it should be part of the doc asap. There are still issues (extra examples; specific relationships for DCAT depending on the type of DCAT artifact; and point relating to qualified and unqualified relationships_
AndreaPerego: I drafted based on the FRBR discussion
… After I had drafted KCoyle provided additional points, and so this isn't the final version, and there might be a dramatic revision of the current draft
… the preview I build on different types of version (3 types) from the library field: replacement; serial publication ; translation / arrangement / reduction, etc
… So, since we are not taking a position about what a version is, these are the types of changes that might happen to a given resource.
… This isn't only related to versioning, it might relate to status also (situation in the production workflow)
<RiccardoAlbertoni> attempt of listing relations made by @makxdekkers in the following post https://github.com/w3c/dxwg/issues/868#issuecomment-616628077
AndreaPerego: In the library field the point is less about version and more about relationships - the ones in FRBR - relate to different levels of classes (which are all disjoint, btw)
… These might help us model how a resource might change over time. The implicit proposal was that we should try to look at the problem from another perspective and rather than seeing everything as a 'version' try to look at other classes of relationship
… We need to make explicit our position about what we consider a 'version' to be, and at the moment this is perhaps confusing. We can recognise the domain and publication specific nature of versioning, and we can then isolate things like serial publication and treat them in their own section
… etc
… I am still working on this
… we need to ensure that dataset series is dealt with
… Our options are either to update the spec or else use the PR to get more information from others
RiccardoAlbertoni: I support the second option. we have discussed for several weeks and the draft covers the bulk of the discussion - it mentions almost all the specific issues. the side discussion on github should be included, and it should reference the issues made by Karen. The draft is likely to change, but merging the PR shows that we are advancing on some issues, while others need more considered thought.
… The points made by Makx about data series also need to be taken into account
… We have decided to stay agnostic on community-specific issues (as in the PR) but there is a legitimate interest in some specialisation and we can point these out in specific issues
AndreaPerego: do we avoid calling the dataset series 'versions'? Probably so
AndreaPerego: Another consideration is that there is a risk in both these discussions that we get caught in a theoretical perspective and not from the POV of addressing real user need
AndreaPerego: the notion of dataset series is used in different ways in different fields (e.g. in geospatial, satellite imagery is a series - a sequential list). Equally we might be developing trees of related datasets
AndreaPerego: What is the situation that existing data catalogues model, and how do we ensure that this can be covered in DCAT?
RiccardoAlbertoni: there are not common practices across communities - so in DCAT we need a common denominator approach that people can specialise for their own domains
… So let's merge the PR and see what the reaction is. The RDA proposal on versioning contains some use cases that are within our scope. These can stimulate discussion of our proposals
proposed: to merge the PR
<AndreaPerego> +1
+1
<RiccardoAlbertoni> +1
Resolution: to merge the PR
AndreaPerego: We need to change the title and also to track changes -
<RiccardoAlbertoni> https://github.com/w3c/dxwg/pull/1259
RiccardoAlbertoni: we can do the title change in the PR #1259. A new section can be used for tracking changes
RiccardoAlbertoni: Can we look forward to a sprint on a group of issues