W3C

Data on the Web Best Practices Working Group Teleconference

05 Feb 2016

Agenda

See also: IRC log

Attendees

Present
phila, PWinstanley, yaso, antoine, newton, Caroline_, ericstephan, annette_g, laufer, hadleybeeman, RiccardoAlbertoni, deirdrelee
Regrets
Chair
Yaso
Scribe
PWinstanley

Contents



.present+ PWinstanley

password for webex?

<Yaso> is xGbzp445, PWinstanley

:-) thanks

<Yaso> no problem :-)

<phila> Yaso: Any volunteer to scribe this week?

<phila> scribe: PWinstanley

<annette_g> *waves back*

<Yaso> PROPOSED: Accept last week's minutes

<annette_g> Yaso, you are very quiet

<annette_g> better

<Yaso> https://www.w3.org/2016/01/29-dwbp-minutes

<phila> PROPOSED: Accept https://www.w3.org/2016/01/29-dwbp-minutes

<ericstephan> it may be your firewall PWinstanley

<annette_g> +1

<Yaso> +1

<Caroline_> +1

<phila> +1

<ericstephan> 0 (was absent)

<newton> +1

RESOLUTION: Accept https://www.w3.org/2016/01/29-dwbp-minutes

<PWinstanley_> phila: will start emailing minutes each week

Dataset usage Vocabulary

<PWinstanley_> Yaso: DUV

<Yaso> http://w3c.github.io/dwbp/vocab-du.html

<phila> latest published version

<ericstephan> https://www.w3.org/TR/dwbp/#DataIdentifiers

<ericstephan> https://www.w3.org/TR/dwbp/#feedbacksection

<PWinstanley_> ericstephan: We haven't made links between DUV and the best practices

<PWinstanley_> ...in the glossary there is mention of a citation, but we don't describe a reference

<PWinstanley_> ...separation of these is important and needs to be done

<ericstephan> https://www.w3.org/TR/vocab-duv/

<PWinstanley_> ericstephan: we have been very busy the past 2 weeks trying to get comments (comments from Robin haven't been responded to yet)

<PWinstanley_> .... trying to write in a collaboration journal

<PWinstanley_> ....opportunity to present a poster too

<PWinstanley_> ...these are good opportunities to publicise the DUV

<ericstephan> https://www.w3.org/2013/dwbp/track/issues/235

<PWinstanley_> ...Bernadette will be publishing it at meetings too

<phila> FORCE 11 Event, April - DUV has a poster session

<ericstephan> https://www.w3.org/2013/dwbp/track/issues/234

<PWinstanley_> ...issue 235, a note back to the editors to make sure we are finding the right namespaces

<phila> I'm planning to offer help with Issue-235

<PWinstanley_> .... JP had questions about the role of the usage tool. We are going to be routing ideas through to communities that have an interest in usage vocabularies

<PWinstanley_> ...questions on 235?

<phila> issue-234?

<trackbot> issue-234 -- Role of Usage Tool -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/234

<phila> issue-235

<trackbot> issue-235 -- Namespaces in DUV -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/235

<PWinstanley_> phila: I am offereing to help (235 - namespaces) . when I was getting doc ready for publication I needed to look through but was careful not to tidy up what I found. However, probably not this month

<ericstephan> https://www.w3.org/2013/dwbp/track/issues/236

<Yaso> issue-236

<trackbot> issue-236 -- agentClassification, usageClassification, skos:Concept -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/236

<PWinstanley_> ericstephan: 236 was more a general question abobut SKOS and usage classification.

<Yaso> ericstephan: almost can't hear you

<ericstephan> I have bad reception

<Caroline_> it is better now! :)

<PWinstanley_> ... 236 - JP had some concerns about the use of SKO Concept. The rationale was to be able to describe something beyond what was described for e.g. a Person (including type of Person etc)

<PWinstanley_> phila: the org ontonlogy has concepts of classification and purpose. I worry about type of person, we all fulfill multiple roles and ascribing a type to a person might be problematic

<PWinstanley_> ericstephan: we did have a usage role but were pushed into the FOAF corner.

<PWinstanley_> antoine: I have reservation about introducing new properties. It is the design principle I don't like.

<PWinstanley_> ...if there was a way to reuse from other vocabs I think that would be better

<PWinstanley_> ...we could recommend using vocabs from another namespace

<PWinstanley_> ericstephan: sounds like a pattern of recommendation rather than formal inclusion

<PWinstanley_> antoine: yes

<PWinstanley_> ericstephan: it sounds like we are trying to address corner cases, and that might be confusing to people. In order to be inclusive we could show patterns

<phila> +1 to limiting the scope

<PWinstanley_> antoine: it is a matter of determing core usage vs occasional use where the authoratative version lies elsewhere

<PWinstanley_> ericstephan: there could be an appendix to address these things

<ericstephan> https://www.w3.org/2013/dwbp/track/issues/237

<PWinstanley_> ...237: there was a question about the use of a term that we found for feedback.

<PWinstanley_> ...we found this class (recommended from a social networking vocab) and inserted this into the model. JP's concern is that this introduces another obscure concept to the model.. So, do we just creata a DUV term rather than importing only one term from this other vocab

<annette_g> +1 for keeping the number of referred vocabs lower

<PWinstanley_> phila: if it is just one term then minting is OK

<PWinstanley_> ericstephan: we can put a comment to refer it to the other

<PWinstanley_> hadleybeeman: +1 to phil's comment. the fewer references to other normative standards the better,

<PWinstanley_> ...for the sake of stability caution is better here

<PWinstanley_> antoine: I agree

<PWinstanley_> ericstephan: do we need a vote?

<phila> close issue-237

<trackbot> Closed issue-237.

<PWinstanley_> Yaso: no, it's OK

<PWinstanley_> ericstephan: I think there might be an opportunity to write some notes about vocab reuse in builfding the DUV - some best practice notes illustrating how to reuse vocabularies

<PWinstanley_> ...I think it is an interesting journy we are on

<Yaso> akc antoine

<PWinstanley_> antoine: I am involved in other groups keen on identify these guidelines, so we don't want too many developing BPs. This though might be brought into our own Best Practices

<PWinstanley_> ericstephan: I would like that - to document things and show the evolution of the vocabulary. I think it is something many go through when building vocabs

<PWinstanley_> antoine: can an action be recorded

<phila> ACTION: antoine to work with eric S on writing section on evolution of DUV wrt reuse of namespaces etc. [recorded in http://www.w3.org/2016/02/05-dwbp-minutes.html#action01]

<trackbot> Created ACTION-227 - Work with eric s on writing section on evolution of duv wrt reuse of namespaces etc. [on Antoine Isaac - due 2016-02-12].

<ericstephan> https://www.w3.org/2013/dwbp/track/issues/238

<Yaso> tks phila!

<phila> issue-238

<trackbot> issue-238 -- Should some of our properties be sub properties of a parent property? -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/238

<PWinstanley_> ericstephan: 238 - Carlos (not on the call) - in some cases we decided that instead of having 2 domains for dataset and distribution we break out the properties

<ericstephan> https://www.w3.org/TR/vocab-duv/#Vocab_Overview

<PWinstanley_> ...looking at the centre of the model I think that this concern about properties we have broken out - are they subproperties, or not

<PWinstanley_> laufer: in the way that was defined before, we have a conjunction of 2 domains. when someone defines a property there will be a distribution defined at the same time as a dataset.

<PWinstanley_> ...the solution implemneted was 2 properties, each with one domain. but we need another so that the vocabulary can describe things that are not dcat:dataset or dcat:distribution

<laufer> I can hear

<phila> acl p

<laufer> I think thta we have different definitions of dataset

<laufer> data cube, for example... or a datacube slice...

<PWinstanley_> phila: while Laufer is writing, I understood him to ask if we need to put domain and range restrictions everywhere. This ties people down to using the vocab in a narrowly specified way

<laufer> so, it will be interesting to have these propertises, like refersTo, with no ranges, for example

<ericstephan> I would prefer a simpler view with no domains or ranges

<laufer> so duv could be reused...

<antoine> +1

<PWinstanley_> ...where the vocab defines a dataset and a distribution, where it doesn't damage the vocab, I would support Laufer in not referring to domain & range

<PWinstanley_> ericstephan: I totally agree with simplifying. I think we were trying to mimic other vocabs that mentioned these things, but I would prefer not to specify domain & range

<laufer> we can, in our examples, show the use for a dcat dataset or distribution... but others used could be nice too...

<phila> PROPOSED: Do not include domains and ranges on properties unless it genuinely adds to the semantics

<laufer> If duv want to define subproperties for specific uses, I think is ok too...

<PWinstanley_> Yaso: next item is BP doc, the table of issues

<Caroline_> +1

<RiccardoAlbertoni> +1

<Yaso> +1

<phila> +1

<PWinstanley_> +1

<ericstephan> +1

<laufer> +1

RESOLUTION: Do not include domains and ranges on properties unless it genuinely adds to the semantics

<hadleybeeman> +1

<newton> +1

<annette_g> +1

Best Practices, table of issue

<Yaso> http://w3c.github.io/dwbp/bp-status.html

<PWinstanley_> Yaso: next agenda item is the table of issues that the editors sent recently

<PWinstanley_> Caroline_: Newton prepared a table to visualise what needs to be done for each BP

<Caroline_> https://www.w3.org/2013/dwbp/wiki/BP_Plan_for_CR

<PWinstanley_> ...we have prepared target dates as per last call

<PWinstanley_> ...we can allocate work from this

<PWinstanley_> ...so take a look

<phila> Just to record, looking at the table, I am feeling smug

<PWinstanley_> ...most important thing is to get people assigned

<PWinstanley_> ...some names have been added, but change/add as you think appropriate. There are still some empty places in the allocation

<PWinstanley_> ...we put Feb 19 as a date

<PWinstanley_> antoine: put me on 16 & 17

<phila> Table of duties

<hadleybeeman> Ah, thanks phila! I was on the BP_plan_for_CR

<PWinstanley_> ... and a question about 18. JP is there. Is there scope for distinguishing between tentative and confirmed assignments?

<PWinstanley_> newton: we just made some suggestions. if you are ok then we keep

<PWinstanley_> antoine: but how do we distinguish between proposed and confirmed assignments?

<PWinstanley_> Caroline_: please can people confirm their assignments

<annette_g> I'm happy to help where my name shows up

<RiccardoAlbertoni> let's put in green the people who has confirmed ..

<phila> Like annette_g, I'm happy with my assignments

<ericstephan> oops I am very delinquent looking at the table...my apologies...I am happy with my assignments

<PWinstanley_> antoine: I have a quesiton about assignment, did you use the table prepared some weeks ago?

<newton> https://www.w3.org/2013/dwbp/wiki/Call_for_BP_example_contributors

<PWinstanley_> Caroline_: we created another table - it is easier to see things

<PWinstanley_> ...newton used the one on the wiki as the basis to make this more detailed version of the table

<PWinstanley_> ...we are focusing on the examples, we used that table as a basis for assignment

<PWinstanley_> ...but things are not fixed - you can choose to work on other things

<newton> who is not comfortable to contribute in one specific BP, we can change it...

<RiccardoAlbertoni> I confirm my contribution in Bp 7

<PWinstanley_> Caroline_: can people on this call attend to confirming, or altering their assignment

<newton> thanks RiccardoAlbertoni

<RiccardoAlbertoni> yes.. i can

<RiccardoAlbertoni> whatever i will start

<PWinstanley_> phila: I am happy with the assignments - and unusually I have lots of green on my assignments

<PWinstanley_> Yaso: Newton sent an email a few days ago - we could use the github assignment

<annette_g> does needs review mean review by editors?

<newton> @annette_g, not only by the editors, but from the group, because we need to make sure that the tests are deterministics

<PWinstanley_> ericstephan: I am on the opposite end of the spectrum - lots of red - but am comfortable with my assignments. I think it is OK as it is, but will think about versioning with Phil

<PWinstanley_> phila: send me an email

<newton> @annette_g and the editors could help with who was assigned to the tasks in what is necessary

<annette_g> yes

<PWinstanley_> Caroline_: maybe annette could confirm her assignments

<annette_g> yes

<annette_g> I think I could help with versioning

<PWinstanley_> I could help with 16 & 17

<laufer> yes

<newton> @PWinstanley_ would you like to contribute in another one, this way we can replace the "?" :-)

<laufer> I think it is ok... my timetable is full...

<PWinstanley_> ok .... let me know another

<phila> I can ping Christophe who wrote those BPs

<phila> He's still reachable

<ericstephan> bp 6 is pretty easy

<PWinstanley_> I will take 28 and 29

<ericstephan> someone should be able to pick that up

<PWinstanley_> antoine: question about contribution - what has happened to the contributor listing?

<PWinstanley_> ...the previous version had a list of contributors

<PWinstanley_> Caroline_: there is a coding issue that Phil is sorting out

<PWinstanley_> ...the generation of the first page has a problem that is being resolved in due course

<phila> ACTION: phila to fix bpconfig.js to restore contributors to BP doc [recorded in http://www.w3.org/2016/02/05-dwbp-minutes.html#action02]

<trackbot> Created ACTION-228 - Fix bpconfig.js to restore contributors to bp doc [on Phil Archer - due 2016-02-12].

<phila> Good to see such focussed progress!

<PWinstanley_> Yaso: all covered. Thanks for making yourselves available. Editors are available if you need specific help

<PWinstanley_> newton: Do we need to create actions for each piece of work?

<ericstephan> annette_g are you going to CoDa in Santa Fe March 1-2?

<PWinstanley_> Yaso: we should perhaps use github. I will send an email.

<annette_g> @ericstephan, I don't even know what that is

<PWinstanley_> phila: if we come back to the table every week then we don't need an action

<ericstephan> annette_g http://www.cvent.com/events/coda-2016-conference-on-data-analysis-2016/event-summary-a11ed42531524891a3ebeb626147a980.aspx

<PWinstanley_> Caroline_: next and the following week can we have this on the agenda

<PWinstanley_> Yaso: no problem

<ericstephan> It might be an interesting place to talk about some topics

<ericstephan> data versioning etc

<annette_g> @ericstephan whoa! maybe...

<phila> Zagreb F2F

<PWinstanley_> phila: if you are going to Zagreb fill in the wiki

<ericstephan> Its really limited in terms of who can go, but would be interesting for you to go

<laufer> bye all... nice wknd... abraços...

<Yaso> bye all!

<PWinstanley_> bye

<RiccardoAlbertoni> bye .. thanks ..

<annette_g> @ericstephan are you going?

<ericstephan> Ywa

<ericstephan> yes

Summary of Action Items

[NEW] ACTION: antoine to work with eric S on writing section on evolution of DUV wrt reuse of namespaces etc. [recorded in http://www.w3.org/2016/02/05-dwbp-minutes.html#action01]
[NEW] ACTION: phila to fix bpconfig.js to restore contributors to BP doc [recorded in http://www.w3.org/2016/02/05-dwbp-minutes.html#action02]
 

Summary of Resolutions

  1. Accept https://www.w3.org/2016/01/29-dwbp-minutes
  2. Do not include domains and ranges on properties unless it genuinely adds to the semantics
[End of minutes]