Government Linked Data Working Group Teleconference -- 19 Dec 2013

<HadleyBeeman> trackbot, start meeting

<trackbot> Date: 19 December 2013

I will be there shortly

<MakxDekkers> On slow and unstable line today because of rain

Hadley and I are just finishing another call

<gatemezi> zamil, +33.4.93.00.aaaa is me

<HadleyBeeman> sorry, all — philA and I are nearly there

bhyland: Thanks to everyone for the input. My apologies for the confusion over Tuesday's meeting
... Sandro kindly called the vote

<scribe> scribe: PhilA

<scribe> scribeNick: PhilA

<bhyland> Frozen version of BP doc is https://dvcs.w3.org/hg/gld/raw-file/4dbafa673d70/bp/index.html

bhyland: We had 17 or so votes. Mostly with +1, but some with fractions
... But a 0 from Dave Reynolds and a -1 from MakxDekkers
... Do we have any room to address MakxDekkers' concerns

sandro: I think so. Procedurally we can make changes as long as everyone's happy with them. The question is how can we assess that everyone's happy with them within the time frame we have
... so if the changes discussed on this call are non-controversial then we should be OK

bhyland: I'll say on behalf of the editors that we'll make these changes. The suggestions are right, there's no push back on the changes proposed and we're happy to fix them

sandro: You mean in Makx's e-mail

bhyland: Yes

<HadleyBeeman> Makx

<HadleyBeeman> Makx's email http://lists.w3.org/Archives/Public/public-gld-wg/2013Dec/0093.html

bhyland: His and Dave Reynolds' concerns too
... Dave was questioning the 5 star scheme usage - and I agree

<HadleyBeeman> Dave Reynolds's email: http://lists.w3.org/Archives/Public/public-gld-wg/2013Dec/0080.html

bhyland: He was questioning the word Web site - I didn't know what else to call it. OK#
... He had other concerns with which I don't disagree

sandro: Several people talked about the 5 star - how do you plan to address that?
... I heard people being unhappy with 'diluting the 5 star brand'

bhyland: I'm happy to delete it altogether as long as there's a link to the 5 star system in the glossary then I'm happy to have it out of the BP doc

sandro: Anyone else?

<MakxDekkers> +1 for removing it

sandro: My inclination is just to remove it but it's not a show stopper for me

<bhyland> https://dvcs.w3.org/hg/gld/raw-file/4dbafa673d70/bp/index.html#VOCABULARIES

<bhyland> scroll down a bit to the "Vocabulary Creation"

<bhyland> section

MakxDekkers: I had 3 major points.
... in section 3 it says that the doc will highlight how LD modelling differs from other formats - but then it doesn't
... so either that needs to be removed or we need to add what it says

bhyland: I think step N needs to be in the BP doc. I'm disappointed that I didn't I didn't add more on that. My preference is to put in another paragraph
... I think there should be a paragraph there. I think there was some language that got cut out. So I want to propose some non-controversial language
... I'd like the opportunity to put it back in
... but I don't want to upset anyone at this stage
... but I think I can write something there that's not controversial

sandro: Do you think you have that para written somewhere?

bhyland: No, but I can follow up directly after the call
... I don't want to start searching around for the old version that has it

MakxDekkers: The second part is under provide basic metadata - the way it's formulated now, it sort of says that you must give those descriptors but they're not all always relevant
... I also know from experience that not all of things are always available or relevant. My proposal is to say 'may include' not must
... and MAY, not SHOULD

bhyland: So better qualification

<gatemezi> I don't see the SHOULD here "When modeling Linked Data, it is a best practice to provide basic metadata, including MIME type, publishing organization and/or agency, creation date, modification date, version, frequency of updates, contact email for the data steward(s). In subsequent sections, further guidance on the use of vocabularies, as well as a vocabulary "checklist" are provided as further informative guidance as they play a key role in the Linked Dat[CUT]

bhyland: This is a problem in real life is that agencies overlook this. I don't think using MAY not SHOULD will change behaviour. Some of those things get missed when they really shouldn't be optional

<gatemezi> Am I missing something?

MakxDekkers: The first argument is that you have some resources where those things don't exist. For e.g. I don't get updated
... The implication is a MUST - and it's context-dependent
... the basic metadata is different for differnet kinds of resource

bhyland: How about we say 'if it is available, one may include'

<HadleyBeeman> "… in can be helpful to itclude…"?

MakxDekkers: Say it as you like as long as people who read this afresh don't think that they MUST provide these descriptors
... And in section 8 - that's a show stopper for me. It really should say follow your nose

bhyland: I agree completely. That was an oversight. it must be in there

MakxDekkers: And then for consistency with DCAT - you should mention the Loc URI sets. I like Lexvo but it's maintained by one person. I'd be happy to keep it, but the LoC URI set is the most authoritative for this kind of info

PhilA: MakxDekkers - LoC and not EUPO NAL??

bhyland: I don't think we should link to a one person project
... Any other opinions about this?

sandro: I'm not familiar with Lexvo, but yes, we should link to LoC

bhyland: Thank you - I will go back and re-read your past messages. I need to do that

<gatemezi> http://www.lexvo.org/linkeddata/references.html uses library of Congress data

MakxDekkers: The spelling of my last name needs correcting in teh acknowledgements

<HadleyBeeman> While we're at it, LinkedGov isn't "UK LinkedGov"

sandro: I changed the header field - you should make it today's date

bhyland: I'll deal with Dave's issues

HadleyBeeman: I think it's great that you want to do this but I don't think you have to.
... you *could* not address dave's

bhyland: But I can, it's a one minute change and I'd like to do it since Dave took the time to review and write the mail
... We did this in 28 minutes ;-)

sandro: Who might have an issue with the modelling paragraph

bhyland: We want to highlight that LD modelling is about stating relationships between resources, you don't have a use in mind
... that's how I present LD
... you're working in an application-independent way

sandro: I don't think that's entirely uncontroversial

bhyland: Yeah, you're right,. I'll go softly softly. It was one if the things Dave highlighted
... Oracle has the 'convert to RDF button' - which doesn't do modelling as such

sandro: So let me propose that if you write a paragraph that Dave and I are both happy with - and anyone else that wants to see it
... I think the process here is that Bernadette is going to make some edits in the next couple of hours. I'm going to suggest she runs those past the relevant people - Makx, Dave, me and maybe others

<bhyland> I agree.

sandro: and if those people are happy then we circulate it to the group one more time. We believe it's OK but if not speak up before it's published in January

HadleyBeeman: I have a counter-suggestion...
... I am uncomfortable with introducing new content at this stage. Makx offered two ways of resolving his issue, one of which was to remove the relevant text
... i.e. remove the sentecne that says that that paragraph needs to exist. That seems to be an easier resolution

sandro: I'm inclined to agree with you that it reduces the risk

bhyland: My feeling is that it's a really important step and that it needs to be in the doc
... we've already neutered this to say that detail is out of scope. We've cut it down a lot. To not have anything else there makes it look inconsequential which means people may think why is it there?

PhilA: I suggest showing the doc with the new para and with that sentence that makes it necessary is removed

HadleyBeeman: So if all that is done, are we agreed that it's ready to go

proposed resolution: pending the changes discussed in this call, that the WG will publish the Best Practices document as a NOTE - accepting the e-mail vote

sandro: I'd like a 2 stage process. Can we get approval from the commentators first

<MakxDekkers> I am available tomorrow to look at it

proposed resolution: pending the changes discussed in this call being approved by the commentators by end of Friday 20th December, that the WG will publish the Best Practices document as a NOTE - as per the e-mail vote

<MakxDekkers> +1

<gatemezi> +1

<bhyland> +1

<gatemezi> +1

<MakxDekkers> +1

<HadleyBeeman> sandro: +1

bhyland: I think we're done...

<HadleyBeeman> +1

<MakxDekkers> bye

<gatemezi> thanks all

bhyland: On behalf of the editors thank you everyone

<bhyland> Happy Holidays, whatever you celebrate!

<HadleyBeeman> You too, bhyland

RRSAgent: make logs public

trackbot, end meeting

- DRAFT -

Government Linked Data Working Group Teleconference

19 Dec 2013

Attendees

Contents

Summary of Action Items

Scribe.perl diagnostic output