Meeting minutes
proposed: accept https://www.w3.org/2020/09/22-dxwg-minutes
+1
<plh> 0
<ana> +1
<AndreaPerego> +1
<riccardoAlbertoni> 0 (i was not there)
Resolution: accept https://www.w3.org/2020/09/22-dxwg-minutes
agenda
<AndreaPerego> https://www.w3.org/2017/dxwg/wiki/Meetings:Telecon2020.10.06
AOB
PWinstanley: thinking about the future of the group - what do we see as the longer-term future, given the work we are doing and the changes that the Process 2020 brings?
plh: the slides were presented 6/12 ago
… Process 2020 was answering questions about the ways to show the current version of a standard that is in active development
… where there are no requirements to demonstrate implementation. The only illustration being WG consensus.
… At the other end of the spectrum there are existing recommendations that need small changes/corrections without having to wait a year or more to effect change
… P2020 allows this. It requires the WG to work out what they want and then to pick the type/s of mode/s that work best for them
… There are not the same constraints on timing.
… Standards can be fixed in time, with periods of a specific version being in effect.
… Although P2020 is active, it is not active (yet) for all WG - this administrative step will take place in November.
… In the new world the royalty-free commitment will be secured from the point of CR Snapshot and PWD1, rather than from the publication of the Rec
… Management of DCAT is up to the group to decide
plh: About DCAT, in case it is used by some gov agency, maybe it would be better to stick to the old path.
… Usually, they are more confortable when they have a specific version to link to.
… But this is not the case, then you can go for the other option - a living document.
… So, what do you want?
PWinstanley: Thanks, plh . Actually, this is something to be brought to the broad group.
… I think for DXWG it depends on which deliverable we are talking about.
plh: In case the group decides to stick to the old REC path for some deliverable, now the procedure is simpler.
<PWinstanley> AndreaPerego: thanks plh for the explanation. If we have a living doc that can be changed at any time, the community of users (e.g gov agencies for DCAT) might prefer the previous approach where there are defined releases. Is there any policy to follow to use a living doc but to prevent breaks?
<PWinstanley> plh: it is up to the WG to sort out a working approach for this
<PWinstanley> ... the WG needs to evaluate, and connect with its user community
<PWinstanley> riccardoAlbertoni: living standards have a publication structure
<PWinstanley> ... if in e.g. DCAT we have a new part, is the PWD a way to get feedback?
<PWinstanley> ... or are we forced to work at the level of CR Snapshot?
<PWinstanley> plh: CR Draft can be published - it just requires a WG decision. The CR Snapshot can only be made not less than every 6 months
plh: new features can only be added if authorised to do so when getting permission to publish the propsed recommendation
… which is not the case for DCAT
… once the PR authorises it then the authority is for ever
… When there are proposed changes, there will be a call for exclusion and AC review so that they can become normative
<plh> https://w3c.github.io/tr-design/p2020mockup/rec-6.2.11.3.html
plh: this is a sample of one that contains proposed changes
… The style is still open to change - the key is to ensure that the proposed changes are easily identified
… If you don't want to do all of that, then the CR Snapshot is the route to take
… These choices are not mutually exclusive
… The key is to ensure that the user community understands the process and the status of the product/s
AndreaPerego: we may want to discuss versioning to update alejandra
DCAT
<riccardoAlbertoni> https://w3c.github.io/dxwg/dcat/#dataset-versions
riccardoAlbertoni: the decision to merge changes related to versioning - see link. s10 has been updated
… the rationale is to show the summary of what is a very wide discussion and guides people to contribute to that discussion
… there is an agreement on the solution
… but there is a need for more discussion on issue #1251
… and #868
AndreaPerego: the extended version of the versioning section includes preliminary considerations in 3 sections. version information and resource lifecycle are not problematic. There is a consensus. Version types, on the other hand, is more problematic
… we have not solved this as the definition is community-specific
… There is github discussion on the use of FRBR - it has a rich model.
… by publishing at this stage we are looking to stimulate feedback from the group
alejandra: I think defining version types is useful in giving us a route to showing relationships between versions
… I also want to mention that there is a UN Data Forum recorded session where I promoted DCAT
… this forum will take place online -
<plh> https://unstats.un.org/unsd/undataforum/index.html
<alejandra> https://unstats.un.org/unsd/undataforum/index.html
<alejandra> yes, that's it
alejandra: my talk was about bridging between statistical and scientific data
<riccardoAlbertoni> Thanks all, have a good night!
<AndreaPerego> Thanks, bye bye!