Data on the Web Best Practices Working Group Teleconference -- 14 Nov 2014

<trackbot> Date: 14 November 2014

<hadleybeeman> Hello!

<deirdrelee> Hi hadleybeeman

<hadleybeeman> Hello, deirdrelee!

<deirdrelee> i'm trying to invite him

<deirdrelee> "/invite zakim #dwbp"

<deirdrelee> that's what i'm using

<deirdrelee> nope, not working for me

<yaso_> haha

<phila> OK, who's able to use Skype?

<deirdrelee> i'm on skype

<yaso_> we can use skype from here also

<hadleybeeman> I'm on Skype.

<yaso_> Me, Nathalia, Caroline and Newton

<phila> It seems to be a recent security update on the Windows box that Zakim runs on

<AdrianoC-UFMG-INWeb> me too, I can use Skype!

<hadleybeeman> phila, were you suggesting a group Skype chat?

<phila> OK, let's get everyone's Skype handle

<phila> yes, sorry hadleybeeman

<nathalia> deception

<phila> Zakim is down

<deirdrelee> ah

<phila> Ralph is looking at it but it looks like a security patch has killed zakim

<deirdrelee> but i could hear eric kauz and yaso no the call

<yaso> “the day zakim went down” seems like a good name for a movie

<phila> +1 yaso

<RiccardoAlbertoni> hi everyone!

<phila> OK, I;m going to start calling people on Skype in 1 minute

<hadleybeeman> k

<phila> If we're not already linked on Skype, please enter your Skype handle here

<phila> prefix with /me if you don't want it in the minutes

<yaso> ouquei. Call Caroline Burle, plz @phila

<laufer> hello all

<phila> I would if she were online yaso

<yaso> Ops. Call Newtoncallegari then, please

<AdrianoC-UFMG-INWeb> My skype id is adriano.machado

<laufer> \me is carlos.laufer

<Caroline_> Is the meeting on Skype?

<phila> Wa hay!

<hadleybeeman> Hi Caroline_ — we're on zakim now. You can dial in as normal.

<laufer> \mute me

<deirdrelee> scribe: annette_g

<deirdrelee> scribe: antoine

Last week's minutes

<deirdrelee> http://www.w3.org/2013/meeting/dwbp/2014-11-07

<phila> PROPOSED: Accept last week's minutes

s/acceot/accept

<phila> +1

<hadleybeeman> +1

<newton> +1

<nathalia> +1

<RiccardoAlbertoni> +1

<laufer> +1

<annette_g> +1

<Eric_Kauz> +1

<AdrianoC-UFMG-INWeb> +1

<Caroline_> +1

RESOLUTION: Accept last week's minutes

<yaso> +1

Issues

<deirdrelee> https://www.w3.org/2013/dwbp/track/issues/open

Focusing on the Use Cases first

<phila> issue-48?

<trackbot> issue-48 -- Phil to look at whether the ucr doc sufficiently covers code lists -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/48

phila: it doesn't
... suggests to close the issue :-)
... We have already 30 UCs
... Perhaps an existing UC could be completed with code lists

deirdrelee: what are code lists?

phila: controlled vocabularies with list of values to be used
... I think the same reasoning (on requirements and use cases) could be applied more widely.
... it's quite easy now to do this

<cgueret> Would it be interesting to look at how schema.org solved the issue ? I remember Peter Mika explaining something about it but forgot what their solution was

<adler1> don't we run the risk of having too many use cases to digest?

phila: I can review the use cases that I've put it to check if some are related to code lists
... Issue 49 about reference vocs is about the same issue
... same for Issue 50

deirdrelee: I could take an action
... on writing to indiviual case authors about things they should check

phila: and put an action on them!

<adler1> intellectually it seems we are seeking new use cases to justify new ideas instead of investigating existing use cases more deeply to discover what we already may have

<phila> ACTION: Deirdre to review issues around UCR and contact individual use case authors to review them in the light of current issues and create relevant actions [recorded in http://www.w3.org/2014/11/14-dwbp-minutes.html#action01]

<trackbot> Created ACTION-130 - Review issues around ucr and contact individual use case authors to review them in the light of current issues and create relevant actions [on Deirdre Lee - due 2014-11-21].

[ annette_g asks about the rational for publishing UCs, deirdrelee answers]

[phila digs up what other groups have done - CSVW ]

adler1: UC are important for relevancy
... the publishing is less important
... and it seems we're in a never ending process of revising UCs

<phila> http://www.w3.org/TR/ldp-ucr/ http://www.w3.org/TR/skos-ucr/ http://www.w3.org/TR/webaudio-usecases/

adler1: It seems that we've done a superficial documentation job on existing UCs

<annette_g> +1 to working use case to requirements, not the other way

adler1: The intent of light use cases was to establish contact
... with owners, start a more organic process than what we have.

<nathalia> +1

adler1: UC like Palo Alto and NYC are really wide. They probably have discussed everything we're discussing

deirdrelee: we agree
... we're in process of refining, finding more details
... what we need is not necessarily new UC, but diving deeper in existing UCs

annette_g: we could look at them from the perspective of data re-user
... not only the perspective of publisher

deirdrelee: seems a sensible idea
... the scope is data on the web, so before it's used, but it's published to be used.
... We seem to have general agreement
... BAck to issue-48. Phil should we close?

phila: we should have a list of current issues that UC owner should look at
... if we find a UC that says something that someone else has said, the UC could be removed

deirdrelee: way to relate this to actions?

phila: your action should be enough
... Including ISSUE-X in the mail you send seems ok.
... same for issue-48, issue-49, issue50...

deirdrelee: We could list the issues now

<phila> issue-49?

<trackbot> issue-49 -- Phil to either improve on the definition of "reference vocabulary", or point to existing definitions -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/49

<phila> issue-50?

<trackbot> issue-50 -- Bernadette to help us find more use cases on the vocabulary itself (including creating a vocabulary) -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/50

<phila> issue-51?

<trackbot> issue-51 -- Phil to clarify the use of the word "provenance" any potential confusion it causes -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/51

phila: it's part of the action of looking back at UCs

[ phila looking at issues]

<phila> http://w3c.github.io/dwbp/usecasesv1.html

phila: there's a need to publish a new version no later than Dec 2!
... it would be good to make enough progress on the document so that we can release a new version

<BernadetteLoscio> I think we should discuss the new requirements from the new use cases

phila: The kind of discussion we're having is useful for this.

<BernadetteLoscio> yes... im talking for the next version

phila: annette_g and Eric Stephan have put new UCs

<phila> issue-91?

<trackbot> issue-91 -- New req: Locations (countries, regions, cities etc.) must be referred to consistently -- open

<trackbot> http://www.w3.org/2013/dwbp/track/issues/91

phila: which resulted in new Reqs

<phila> R-Location

<phila> Locations (countries, regions, cities etc.) must be referred to consistently

phila: two new UCS make a lot of ref to locations
... it's not rocket science, but it is relevant, and testable
... it has the three (meta) requirements we need

<phila> http://w3c.github.io/dwbp/usecasesv1.html#R-Location

phila: to declare it relevant. If the group agrees

deirdrelee: something about code lists?

phila: yes

deirdrelee: maybe we could add "for example, a location code list" for this

phila: we could recommend to use URIs for countries
... some don't like this, but the data should be consistent in any case

<BernadetteLoscio> location is a kind of metadata, no?

<phila> antoine: On code lists... my own project is an example. We just use text for countries. Code lists like ISO codes 3166-X would prove useful

<phila> ... so I'd be in favour of adding this example of a locatiomn code list to the use case

laufer: we will find a lot of things particular to each of the publication.
... it will be difficult to define the exact way to do this

<BernadetteLoscio> i think that the recommendation of URIs should be general, not just for countries

<hadleybeeman> a+

<phila> antoine: laufer - think you're right. We prob cannot force people to use one list for everyone, but we can recommend some patterns

<phila> ... like ISO codes, URI sets etc. and you might want to encourage their use in a given community

<phila> +1 to antoine

<Eric_Kauz> +1 to antoine

<RiccardoAlbertoni> +1 to antoine

hadleybeeman: list of countries are very political

<phila> http://w3c.github.io/dwbp/bp.html#bp-license

deirdre + phila : we agree

phila: pointing to the license BP: it doesn't specify one single way to do it
... rather, it says what the intension is, and gives possible implementation
... sthg like "here's one acceptable way you could do it"

<BernadetteLoscio> I guess that some metadata values should be defined bt URIs, for example: location

deirdrelee: should we incorporate the req?

Steve: it's a specific field: should it be included explicitly?
... GIS is super important
... but does it deserve specific formal requirement?

phila: if it's not included then it won't be in the BP doc

steve: I'm good with it!

<hadleybeeman> +1 to PhilA

<phila> Locations (countries, regions, cities etc.) must be referred to consistently

<BernadetteLoscio> +q

phila: this is the current phrase, the intension was to link it to the use of code lists

adler1: most of the UCs I wrote have GIS/maps in them
... which location: location in map or physical location?

phila: many cases
... as long as the case have consistent practice

BernadetteLoscio: not sure the requirement should be specific to location
... other type of metadata could have the same

adler1: is there a downside?

<laufer> +1 to bernadette

<Zakim> phila, you wanted to say it's not all about metadata

adler1: I don't think it's exclusive

phila: yes, spatial is relevant for metadata. Also for structural metadata and data itself itself.

<BernadetteLoscio> ok... i understand

<BernadetteLoscio> but again...

<BernadetteLoscio> it should be a general requirement, no?

<BernadetteLoscio> for both data and metadata?

<nathalia> s/lomng/long

<BernadetteLoscio> data and metadata values should be referred to consistently

deirdrelee: technically location is included in the code list but there's no harm making it a separate requirement

annette_g: there maybe other metadata type that raise to the same level
... maybe we can convert this a best practice that lists the different types that should be concerned

makx: I want to add language!

<Eric_Kauz> currency too.

deirdrelee: this is getting close to a list of critical data types

<MakxDekkers> +1 to currency

<BernadetteLoscio> so... this should be a general requirement

phila: deirdre took an action on asking people to look at their UC
... we could ask people with specific interest to look at a wider range of UC
... trying to detect whether their favorite topic could be represented in the UC

<annette_g> +1 to phil

<MakxDekkers> I think language applies to each and every use case

deirdrelee: what would be the end result. E.g. "language should be described consistently"?
... separate requirements for everything?

<MakxDekkers> at least when there is any text involved

<BernadetteLoscio> and for the data?

deirdrelee: where do we draw the line?

<yaso> +1 to Phila

adler1: BP should apply to all the data points we've discussed
... but I don't think to spend time on each BP (location etc)
... we would imagine how the world should be based on our standards

MakxDekkers: we're talking about a number of things that could be expressed as code lists
... maybe we could cover that aspect as an horizontal issue
... in the BP document
... we should have a general view.

<BernadetteLoscio> +1 to Makx

<Zakim> hadleybeeman, you wanted to respond to Makx about requirement s not in the use cases

annette_g: if it helps people we can put at higher level

hadleybeeman: I'm a bit worried about 'horizontal' phrasing

<BernadetteLoscio> +1 to Makx to have a general view

deirdrelee: good discussion, even though we've not closed issues!

<MakxDekkers> bye

<yaso> Bye

<nathalia> bye

<BernadetteLoscio> entendi que deveria ser tratado bye

<newton> bye

<RiccardoAlbertoni> thanks, bye

<annette_g> bye

<laufer> bye all, nice wk

<jerdeb> BYE

<adler1> nice call!

<AdrianoC-UFMG-INWeb> Good bye!

<phila> trackbot end meeting

- DRAFT -

Data on the Web Best Practices Working Group Teleconference

14 Nov 2014

Attendees

Contents

Issues

Summary of Action Items

Scribe.perl diagnostic output