Meeting minutes
proposed: accept minutes of last meeting
<PWinstanley> https://www.w3.org/2020/09/08-dxwg-minutes
<AndreaPerego> +1
<Ana> +1
<PWinstanley> +1
+1
<Rachel> +1
Resolution: accept minutes of last meeting
<PWinstanley> https://www.w3.org/2017/dxwg/wiki/Meetings:Telecon2020.09.22
approve agenda
TPAC
register and join in anything you want to. Our group will be pretty low key, doing a short breakout session.
PWinstanley: (above and here) anybody else have ideas for things they want to do?
<AndreaPerego> Not from my side.
PWinstanley: the other thing in the admin area was the Process 2020, which we're now one week into.
Process 2020
<PWinstanley> https://www.w3.org/2017/dxwg/wiki/Process2020Information
PWinstanley: This is where I put a synopsis of the changes.
… there's a little flexibility for making changes, some may be in the cards for DCAT v2
… we would be kind of the guinea pigs if we do errata.
PWinstanley: please familiarize yourselves with knowing where it is, link off main page of the wiki
AndreaPerego: what has been done recently, the respec template has changed. One of the things we did was try to use it for the specification, but there were some bugs.
… we can ask what the new version of the respec library does to the content that is automatically filled. There is no detailed description of the changes.
PWinstanley: we can ask PLH
… I'll write and ask
DCAT work
We met last week in the whole subgroup to discuss versioning. We took some actions, and we are now trying to draft an extended version of this section of DCAT. Discussion is still going on in GitHub. There will be some dramatic revision.
PWinstanley: yes, bikeshedding is one of these complicated issues. If you take the simple route, there's not a lot to say about it, otherwise it takes up a lot of the space of the rec. So we're thinking about how much goes in the rec and how much goes in primers.
AndreaPerego: Today Karen made an important comment. We recognize that versioning is very related to local practices. There is not a common definition of it. It depends on a lot of variables. On the other hand, leaving it out doesn't help. We at least want to have a definition that's helpful.
AndreaPerego: anybody who would like to chime in and comment, that would be more than welcome.
<AndreaPerego> Current discussion on versioning: https://github.com/w3c/dxwg/issues/1251
PWinstanley: anything anyone else wants to discuss?
crickets
open actions
PWinstanley: some of them should go in the primer. #381 is completely superseded by the new process; even the terminology is different. Are people happy to just close that?
crickets
PWinstanley: 428 has gone, hasn't it?
<AndreaPerego> close action-428
<trackbot> Closed action-428.
close action-381
<trackbot> Closed action-381.
PWinstanley: 425, annette, this is what we're talking about with versioning.
<PWinstanley> annette_g: I have been in touch with colleagues to see if anyone was interested in contributing to workflows, but in vain
<PWinstanley> ... It seems to be that workflows are unique and getting standardised terms for the stages is not going to yield useful results
<PWinstanley> ... The workflow stages depend on the instruments used, the datasets, the techniques used, etc, and so the merits of establishing a core vocabulary to describe these is questionable
AndreaPerego: Based on what you said, I wonder whether we should look at versioning just for the final stage, when the data are published, instead of looking at what happens behind and before that. I have several examples of a dataset published on platforms where you can have different versions, you can update it with a new DOI.
… the rationale is that the dataset may have a relationship to a specific publication. Do you think this would make sense? I understand probably it's more something that is not a concern for researchers.
yes
AndreaPerego: It may apply to published data. The researcher has to state that. We've had several cases where a dataset has been up for a while but we've changed the methodology, so the old dataset is deprecated and there is a new one.
right
PWinstanley: There are connections with prov, where something was done to a dataset, what process, what the raw material was. To have an illustration of the types of vocabulary that could describe status is potentially useful in this space.
AndreaPerego: this is also what we plan to do. We hope to find patterns for these vocabularies and properties because sometimes the definition is not crystal clear. Trying to guide people is difficult. We are trying to study them to understand the relationships between derivation, revision, etc.
<AndreaPerego> UKGOVLD Registry: https://github.com/UKGovLD/registry-core/wiki/Principles-and-concepts#status-and-life-cycle
AndreaPerego: for lifecycle and versionining there is an ISO standard, which is about the item registration--registries. This work has been reused by the UK government working group for their registry. That work is interesting. The documentation they provide is very interesting. An item needs to be approved, etc., the richest example I am aware of.
AndreaPerego: It's about the fact that these taxonomies are used by other organizations. One thing they support is to enable a specific version for a code release, to at least make sure the terms are there. It's very useful to see the workflow for the lifecyle. In my opinion, it can maybe be reduced depending on the use cases.
PWinstanley: That UK gov data working group registry got a couple of implementations in things like the meteorological office, CSIRO, and what have you, but I'm not aware of implementations that have really been pushed. It just seems to be things were people are just trying it out. Are you aware of anyone using it with higher energy?
AndreaPerego: INSPIRE registry is using it, but not the whole workflow. They are still out there but I guess something just didn't happen basically.
PWinstanley: yes, it's still useful archaeology. Some things are a bit ahead of their time.
<PWinstanley> s/ASPIRE/INSPIRE/
<AndreaPerego> https://inspire.ec.europa.eu/registry
AndreaPerego: I understand not everyone on the call knows what is INSPIRE. It's basically a registry providing access to ISO code lists for geospatial data. It is for environmental policies across Europe.
PWinstanley: Do we think the action can be closed now?
no objection here
<AndreaPerego> +1
<Rachel> +1
close action-425
<trackbot> Closed action-425.
PWinstanley: we need a few more people to do more with open actions.
… anything else people would like to say?
AndreaPerego: wanted to ask Rachel if she has interest in the work on DCAT, if there is something that she sees missing, needing elaboration?
Rachel: I'm still kind of trying to understand exactly what the use cases are, reading through the materials. I gave myself a month or so to get up to speed before jumping in. I do have experience in versioning of catalogs, but don't yet feel in a position to make suggestions or say what is missing. Maybe in a few weeks.
PWinstanley: Ana?
Ana: I'm okay
PWinstanley: let's meet again in a couple of weeks' time. Don't forget about TPAC stuff
<AndreaPerego> Thanks, and bye!
<Rachel> Thank you