Meeting minutes
<RiccardoAlbertoni> rsagent, draft minutes
<RiccardoAlbertoni> r
<RiccardoAlbertoni> https://
Repository Migration & Ownership
Simon: have experience from OWL Time ontology, key is moving directories with full history
Simon: new repo is not a submodule now
Simon: successfully pushed the dcat directory, but did not test the rest of the functionality
Simon: Contained history since 2017
Simon: Still missing - issues transfer (automatable), pull requests cannot be moved
RiccardoAlbertoni: Also to be moved: implementation report, wiki, DCAT releases
… for DCAT3, proper release functionality not used, but should be used in future
Simon: Implementation report should definitely be moved. Unsure of wiki
… is it repository per deliverable? Is the implementation report a deliverable, or a dependency?
RiccardoAlbertoni: The implementation report was not a separate deliverable. It should be in the same repo with DCAT
… let's discuss who takes ownership of the repo migration
… It makes sense that Simon gets admin permission and replicates what has been done elsewhere, RiccardoAlbertoni will take up missing items
Simon: Then the moved DCAT resources should be removed from the original repo?
RiccardoAlbertoni: This should be discussed at the WG plenary meeting
… created a couple of issues regarding RDF representation management and things that would help in future maintenance of the repo
Simon: Will proceed with the move of the DCAT directory to the new repo
ACTION: to simon Will proceed with the move of the DCAT directory to the new repo
ACTION: Simon to move the directory
Maintenance Scope vs. Versioning Policy
RiccardoAlbertoni: Tried to use issue labels for Classes of change
… Classes 1 and 2 can be published in the current REC upon voting in the WG
<RiccardoAlbertoni> /www.w3.org/policies/process/#revising-rec///www.w3.org/policies/process/#revising-rec
actions?
<RiccardoAlbertoni> https://
RiccardoAlbertoni: Different kinds of recommendation revisions
… in the W3C process
… we need to decide whether we are going to deliver a new REC or not
… for minor changes, we maintain the list of current editors, for new features and substantial changes (class 4 and 3) we may remove editors of the previous version
Simon: only periferally involved in DCATv3
… not really up to speed with the scope and scale of the work to be done with DCAT
… was this presented at the WG plenary?
RiccardoAlbertoni: We have many issues, most of them reformulation, explanation, guidance improvement
… already requests for new properties though
… e.g. the new dcat:datasetSeries property, which is probably easy
… for that, we do not need to collect an actual use, it may be enough to just say it makes sense to have it
… Jakub had some ideas about machine readability of DCAT
… might be substantial change or feature
Jakub described what is missing in the DCAT REC from the machine-processability point of view - machine-readable representation of the DCAT REC as a whole, perhaps using PROF vocaulary, and machine readable representation of profiles of terms from other vocabularies in the context of DCAT
RiccardoAlbertoni: Jakub should open an issue on this and explain
… every new feature in REC requires independent implementations
ACTION: Jakub open issue on what could be done for improving machine readablity of DCAT
RiccardoAlbertoni: we are not 100% sure we will deliver DCAT4
… it is a difficult process
Simon: we also considered a living standard, but that is hard to do for W3C
RiccardoAlbertoni: Next actions are to finish the repo transfer, so that we can start working
… class1 and class2 issue resolutions could be merged, class3 and class4 should not be merged
… we need to decide about providing a new DCAT release or not before that
RiccardoAlbertoni: This is the first time we are not starting with a new deliverable, which is more straightforward than having the options now
actions?
RDF Automated Maintenance (w3c/dxwg#1639 )
RiccardoAlbertoni: experimenting with gh actions for automatic generation of RDF representations in other formats. Goal is to keep the diff to minimum.
… goal is to change only RDF Turtle files, generate the others
… let's take it step by step, first work on smaller issues
… integrate Echidna later
Simon: There is experience in the SSN ontology, just got it working
RiccardoAlbertoni: Maybe we could first integrate with Echidna? Every merged PR going into the working draft could be dangerous
Simon: merged about 30 PRs in the SSN ontology, feedback gets back quickly. But there was a policy decision early in the process that merging in the main branch requires 2 positive reviews
… when the reviews took a long time, it was frustrating.
… the review process reduced the number of small merges.
RiccardoAlbertoni: alternatively, we could have a side-branch for quick publication, and keep the main branch as an editor draft
… so only once more changes are reviewed, they get to the main branch, in a more controlled way
… we already have the 2-review requirement in place
Simon: Is the main branch gh-pages?
RiccardoAlbertoni: now yes, gh-pages