W3C

- DRAFT -

SV_MEETING_TITLE

26 Jan 2012

See also: IRC log

Attendees

Present
galway, HadleyBeeman, fadi, PhilA, dvilasuero, cygri, mhausenblas, GofranShukair, BartvanLeeuwen, csarven, BenediktKaempgen, gatemezin, sandro, olyerickson, GeraldSteeman, boris, Washington, t_gheen, spyroskotoulas, DeirdreLee, DaveReynolds, stasinos, +1.703.201.aaaa
Regrets
Chair
SV_MEETING_CHAIR
Scribe
cygri, gatemezi, fadi

Contents


<cygri> ACTION: cygri to move dcat-related content from eGov wiki to GLD wiki [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action01]

<trackbot> Created ACTION-37 - Move dcat-related content from eGov wiki to GLD wiki [on Richard Cyganiak - due 2012-02-02].

<PhilA> close Action-29 Completed 26/1/12

<PhilA> action-29 Close

<PhilA> close Action-29

<trackbot> ACTION-29 Add products on issue tracker closed

<PhilA> ACTION: PhilA to convert editors' draft of DCAT to make use of respec [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action02]

<trackbot> Created ACTION-38 - Convert editors' draft of DCAT to make use of respec [on Phil Archer - due 2012-02-02].

<cygri> ACTION: cygri to update http://vocab.deri.ie/dcat to point to the new dcat ED [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action03]

<trackbot> Created ACTION-39 - Update http://vocab.deri.ie/dcat to point to the new dcat ED [on Richard Cyganiak - due 2012-02-02].

<HadleyBeeman> Does zakim know that this IRC channel is connected to our call?

<cygri> (ghislain, boris, spyros and deirdre are not yet on IRC)

thanks, cygri

<gatemezin> @cygri, thanks!!

<HadleyBeeman> :)

<HadleyBeeman> Morning, Washington

<HadleyBeeman> Wasn't there going to be some breakout work from eGov as well (on DCAT)?

<PhilA> scribe: cygri

gatemezin: ghislain atemezing, EURECOM, france

<mhausenblas> ACTION-31?

<trackbot> ACTION-31 -- Boris Villazón-Terrazas to create a Wiki page on multi-lingualism of vocabs -- due 2012-02-01 -- OPEN

<trackbot> http://www.w3.org/2011/gld/track/actions/31

Report on work since yesterday

<boris> http://www.w3.org/2011/gld/wiki/Multi-lingualismOfVocabs

<mhausenblas> close ACTION-31

<trackbot> ACTION-31 Create a Wiki page on multi-lingualism of vocabs closed

boris: we created a wiki page on multilingual issues

<mhausenblas> close ACTION-32?

mhausenblas: all please contribute to this page

<mhausenblas> ACTION-32?

<trackbot> ACTION-32 -- Michael Hausenblas to compile first version of vocabulary selection quality checklist -- due 2012-02-01 -- OPEN

<trackbot> http://www.w3.org/2011/gld/track/actions/32

<mhausenblas> close ACTION-32

<trackbot> ACTION-32 Compile first version of vocabulary selection quality checklist closed

<boris> http://www.w3.org/2011/gld/wiki/VocabularySelectionQualityChecklist

mhausenblas: we did a first cut of vocabulary selection checklist
... vocabulary selection might be too restrictive, so we reinterpreted as "dealing with vocabularies"
... to cover vocabulary discovery, selection and creation
... we expect that most users will be well-served by info on discovery+selection
... but some will need to create new ones
... discussing ontology creation methodologies are out of scope but might point to them informatively
... versioning ...
... flow chart ... driven by usage ... add/remove terms ...
... deprecating vocabularies
... some cross-cutting issues
... like stability

<PhilA> cygri: This was all about vocab selection?

<olyerickson1> When I browse to the Vocab Selection wiki, it doesn't show me as logged in...even though I'm logged in

<PhilA> mhausenblas: No, part 1 was selection and discover, part 2 was about management

<PhilA> mhausenblas: We said we want to provide a checklist, not a list of recommended vocabs

<mhausenblas> [[Vocabulary Selection. The group will provide advice on how governments should select RDF vocabulary terms (URIs), including advice as to when they should mint their own. This advice will take into account issues of stability, security, and long-term maintenance commitment, as well as other factors that may arise during the group's work.]

<PhilA> cygri: The charter doesn't mention vocab creation. It seems that we're stretching the charter

<sandro> "advice as to when they should mint their own"

<sandro> (but not *how* :-)

<PhilA> cygri: So my preference would be to stop at the point where selection and discovery fails

<DaveReynolds> Exactly, don't try to give pointers to advice on vocabulary creation. There's lots out there already and doing a comprehensive enough job seems out of scope.

<bhyland> ‰q+

<PhilA> mhausenblas: There aren't any established BPs for vocab creation, stuff like deprecating terms and so on.

<PhilA> mhausenblas: If you don't talk about deprecation at all then people may think that it isn't possible

<DaveReynolds> So explicitly state what issues are NOT covered, but don't try to partially cover then through "informative" sections.

<PhilA> mhausenblas: But people should be aware of the issues around creation

<PhilA> mhausenblas: So I agree with cygri - but we should focus on selection and discovery but then include a paragraph talking about what the issues are?

<PhilA> cygri: The other things was stability, versioning etc.

<PhilA> cygri: Is it worth breaking those out in the vocab selection? They get discussed in the BP document anyway so can we just refer to that?

<olyerickson> I think we need to make stakeholders aware (a) that vocab selection matters (b) that there are best practices for selecting vocabs, based on common usage (c) that custom creation is possible but should be deferred if prior existing vocabs exist (d) that long-turn "stability" of vocabularies may be a factor

<PhilA> sandro: I agree with what's being said

good point sandro

sandro: when we say what consumers are looking for in a vocabulary, this also helps producers to understand what vocabulary consumers are looking for
... so it's discovery/selection/evaluation

<olyerickson> We should provide guidance (or pointers) to "evaluation criteria" rather than specific evaluation criteria. Also, evaluation criteria (and ratings systems0 for vocabs are evolving

mhausenblas: eager to capture experience of vocabulary maintainers, like danbri

bhyland: creation of vocabularies should be a book of its own
... better to restrict to selection criteria. pay attention to who maintains the vocabulary, is there a long-term maintenance plan in place etc

<George> ack

PhilA: i wrote stuff for the EC last year that covers much of what we just talked about. i'll contribute that

<PhilA> ACTION: PhilA to reflect on SEMIC advice on vocab selection etc. to see if there is more to contribute to the BP doc [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action04]

<trackbot> Created ACTION-40 - Reflect on SEMIC advice on vocab selection etc. to see if there is more to contribute to the BP doc [on Phil Archer - due 2012-02-02].

<mhausenblas> https://www.w3.org/2011/gld/track/actions/open

<bhyland> @Michael, et al, are we all agreed this is the main wiki page for Vocabulary discussion from which all other sub pages will be linked? See http://www.w3.org/2011/gld/wiki/Vocabulary_Discussion_Summary

http://dvcs.w3.org/hg/gld/raw-file/default/dcat/index.html

<mhausenblas> Michael: Yes, bhyland, I think so

<bhyland> ta

http://www.w3.org/2011/gld/track/products/2

<PhilA> I have linked the newly public Editors; draft from today's agenda

http://www.w3.org/2011/gld/wiki/Data_Catalog_Vocabulary

<Zakim> mhausenblas, you wanted to talk about on how to use products (very quickly)

cygri: we have a dcat editor's draft, updated list of dcat issues, and a stub wiki page for dcat on the gld wiki

mhausenblas: we now have separate products for dcat, best practices, cube in the tracker

<bhyland> What is the list of "products" in Tracker for us to assign?

mhausenblas: so please assign new issues to the respective product

http://www.w3.org/2011/gld/track/products

mhausenblas: when you create/edit an issue, there's a product dropdown
... i think you can't assign issues to products from IRC

<olyerickson> bhyland you can create in IRC then fine-tune in the hyperinterface

Standard Vocabularies

<olyerickson> PLEASE specify pages when we're talking about specific pages

bhyland: there's a lengthy list of people here http://www.w3.org/2011/gld/wiki/Vocabulary_Discussion_Summary#Vocabulary_5:_Geography.2C_Spatial_Information
... whom do i chase?

PhilA: for dcat, that's me, fadi, john, richard

cygri: for data cube, we have a list from yesterday

mhausenblas: looking at the charter, how do we break this down into specs?

sandro: no umbrella spec necessary

bhyland: we talked about it a bit yesterday

<DaveReynolds> Surely vocab guidance is in Best Practices document

bhyland: i think we might want to have an umbrella spec about selection etc

sandro: the umbrella spec should be the Best Practices

mhausenblas: so for each of the five areas in Standard Vocabularies, what do we do?

<olyerickson> What "5 points" where are we looking...

olyerickson, look at the charter

<olyerickson> Okay thanks

mhausenblas: charter, section 2.3
... metadata corresponds to dcat, statistical cube data to data cube, don't know about the rest

<olyerickson> Thanks mhausenblas (just trying to make sure we have links with words like "these" ;) )

PhilA: the work i'm currently doing for the EC covers people, legal entities
... i'd like to put some energy into organization ontology

<mhausenblas> Michael: I'd be interested in learning from DaveReynolds if he'd be willing to/interested to move ORG ontology into W3C (see 4. in section 2.3 of our charter)

DaveReynolds: transferring into w3c namespace and w3c care would be good
... i can do very little work on that though
... it's completely public domain, so no ip issues

<cmusialek> I'd be interested in putting some energy into the organization ontology as well.

<mhausenblas> PROPOSAL: Moving http://www.epimorphics.com/public/vocabulary/org.html into W3C namespace and consider it as addressing 4. of section 2.3 of our charter

<cmusialek> Data.gov is interested in implementing an organization ontology for US govt entities in the short term

http://www.epimorphics.com/public/vocabulary/org.html

<Zakim> DaveReynolds, you wanted to ask about timescales

DaveReynolds: this ontology is aimed at generic re-usable concepts. so it has to be extended for particular uses
... so data.gov.uk had to extend it for UK gov use

charter: http://www.w3.org/2011/gld/charter

DaveReynolds: what's the timescale?

<olyerickson> Wonder if there are some good Data.gov datasets with org data that could be useful test cases

charter says: FPWD in Dec 2011, LC Oct 2012, CR Dec 2012, PR Mar 2013, Rec Apr 2013

PhilA: should be not much work given the pedigree and uptake of this work

<George> I used Site from Org to get to VCard:Address stuff

<cmusialek> Cygri: where is the UK govt's ontology implemented? Do you have a link to share?

cmusialek, that's a question for DaveReynolds

DaveReynolds: the issue is not the amount of time writing stuff; it's about public review, coordination etc
... ideally we could serialize this, make progress on the data cube first, then org etc

<sandro> DaveReynolds: I'd prefer to serialize this work, doing QB first.

<Zakim> mhausenblas, you wanted to clarify his proposal

mhausenblas: i disagree with serialization. we should send a signal soon, by doing FPWDs of whatever we have now

<George> +1 mhausenblas

mhausenblas: so that the world sees: these are the things in our scope
... i agree that we can do serial focus once FPWDs are out
... we should get FPWDs out in a matter of weeks

<BartvanLeeuwen> +1 mhausenblas

bhyland: i agree

<dvilasuero> +1 mhausenblas

<George> cmusialek: see http://health.data.gov/doc/hospital/393303/site/1

<mhausenblas> PROPOSAL: Moving http://www.epimorphics.com/public/vocabulary/org.html into W3C namespace and consider it as addressing 4. of section 2.3 of our charter

<bhyland> +1

<George> +1

<BenediktKaempgen> +1

<gatemezi> +1

<BartvanLeeuwen> +1

<DaveReynolds> +1

<HadleyBeeman> +1 to schema.org not being appropriate for this

<sandro> +1 (reconsider if we find any other viable options)

<boris> +1

+1

<olyerickson> make what happen?

RESOLUTION: Moving http://www.epimorphics.com/public/vocabulary/org.html into W3C namespace and consider it as addressing 4. of section 2.3 of our charter

<PhilA> ACTION: PhilA to create W3C format of Org ontology specification document [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action05]

<trackbot> Created ACTION-41 - Create W3C format of Org ontology specification document [on Phil Archer - due 2012-02-02].

<bhyland> @Sandro - do we write APPROVED for the proposal to capture in minutes?

olyerickson: we've done some work on schema.org extensions in other areas

bhyland, we write RESOLUTION like i did

<bhyland> @PhilA - thanks!

<bhyland> @PhilA, I particulary like that due date: 2012-02-02 ;-)

olyerickson: just because schema.org doesn't cover something right now, this doesn't mean we can't extend it

<sandro> bhyland, cygri did it fine, "RESOLUTION:" or "RESOLVED:" or "APPROVED:" should all work.

<bhyland> @sandro ta

<mhausenblas> Michael: Wondering if anyone fancies an action for cleaning up http://www.w3.org/2011/gld/wiki/Vocabulary_Discussion_Summary and sync w/ today's resolution

ADMS discussion

<Zakim> mhausenblas, you wanted to note that we have not discussed 3. People of 2.3

<scribe> scribenick: BenediktKaempgen

ADMS

EU Commission, Asset Description Metadata Schema (ADMS)

<dvilasuero> I can take an action for cleaning up http://www.w3.org/2011/gld/wiki/Vocabulary_Discussion_Summary

bhyland: any wishes to rearrange agenda?

<olyerickson> FWIW: People interested in RPI's DCAT-driven Schema.org extension for Catalogs and Datasets can check out demo at: http://logd.tw.rpi.edu/schemaorg_dataset_extension

PhilA: Comment - hopes to make the most of the time
... Interoperable Solutions - big EU activity
... two strangs: ADMS, Section on Core Vocabularies
... Fadi, Gofran more on ADMS
... UML diagram in the wiki - http://philarcher.org/isa/adms.png
... describing meta data. Meeting in March: Using ADMS to interchange metadata.
... lots of similarities with dcat: repository, asset, release of asset (in particular format)

GofranShukair: asset can be dataset but reusable one
... release (actual document representing asset)

<bhyland> @dvilasuero, thanks for cleaning up the vocab discussion summary page. We made some good progress today and it would be great to reflect that as a status update on that page and link to all the links provided in IRC today.

<bhyland> ping

GofranShukair: asset, release have statuses (e.g., already published)

<bhyland> ping

GofranShukair: idea behind ADMS: semantic assets making reusable. E.g., asset metadata, federation (professionals can find assets)

<sandro> much better :-)

GofranShukair: federation (explore, reuse of assets)

PhilA: model is actually in public review
... has well pedigree

<bhyland> @dvilasuero, thanks very much for tidying the vocab discussion summary page and reflecting what was discussed today.

PhilA: through ?

<bhyland> ping

<cygri> ACTION: cygri to update http://www.w3.org/2011/gld/wiki/Data_Cube_Vocabulary [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action06]

<trackbot> Created ACTION-42 - Update http://www.w3.org/2011/gld/wiki/Data_Cube_Vocabulary [on Richard Cyganiak - due 2012-02-02].

PhilA: reuse of dcat repos, asset, relesase, lots of references to dcat, also to void, foaf
... issue of vcat: particular use of term/class, and we have ranges: it can be that we are making relationships that we are not allowed to make.
... wants to know whether this is of interest to US people
... published as XML, RDF...
... no problem to change example.org to w3c.org (TBL did not mind)
... EU has concerns, e.g., about stability (change control)
... politicians probably will have concerns.

george: Probably, it would be easier for US standards

<sandro> (that wasn't sandro, BenediktKaempgen... didn't catch who it was.)

then it was george

I mix you often ;-)

cmusialek: Interested.

<rreck1> yeah i wanted google+ yesterday

<George> US Gov agency adoption of ADMS will be greatly facilitated by having ADMS namespace be provided by a 'voluntary consensus standards organization' (of which W3 is an important one)

PhilA: Will be elaborating on relationship to dublin core.
... later this day.
... e.g., will inferencing change concepts that are linked by ADMS.

<DaveReynolds> Technically correct practice is to use OWL restrictions :)

<George> +1 DaveReynolds

s\dublin core\dublin core

<olyerickson> Arghh my tax dollars at work

<sandro> can someone in washington try g+ on their laptop, to see if it sticks to port 80 or something?

PhilA: Second section: Core vocabularies.
... EU Commission wants to recommends vocabulary to be used cross-country; possible to specialize.

<rreck1> noone here is willing to try

PhilA: based upon lots of resources: First core vocabularies: person, business, location

<cygri> http://philarcher.org/isa/corevocs.png

<olyerickson> Link please

<olyerickson> thanks

PhilA: background: experts on vocabularies, Euro justice, business registrars, INSPIRE directive secretar (environmental data)
... all three things are interlinked.

<DaveReynolds> ISADate?

PhilA: properties should be familiar. No relationships between persons. Only most important features.
... without reference to other vocabs, but implicit references should be obvious (e.g., foaf)

<stasinos> Greek too

PhilA: cultural differences covered e.g., patronymic Name
... dates raise big problem:

<bhyland> ping

PhilA: many people do not know their birthdate or place

<sandro> pong

PhilA: recording dates: most dates in public databases not clearly defined
... to represent those, raises big issues.
... Cleaning would sometimes even make dates inaccurate (e.g., from August 1984 to midnight 1. August 1984)

<olyerickson> PhilA: If you know dat, use xsd:date

<olyerickson> ...if you can't type to xsd: date, use a literal with whatever you have

<sandro> sigh. what a pain. not sure if there's a better solution.

<rreck1> sounds like a sane approach

<cygri> i'd prefer xsd:gYear or xsd:gYearMonth over w3cdtf

PhilA: actual recommendation would be to represent the date depending on the situation (e.g., representing parts separately).

<DaveReynolds> Yes, there is a difference between date and datetime. Much gov data is actually about intervals, not dateTime points, hence the UK reference time service for time intervals.

<sandro> I'd be inclined to use different properties, I think. one if you know it, one for a comment, especially if you dont know the actual value.

<olyerickson> @sandro @cygri Our guys have suggested xsd: date then xsd: g* for date parts that are known

PhilA: Next businesses: also issues to represent legal entities
... Legal entities must have a legal identifier (registrar).

<DaveReynolds> +1 to sandro, semantics are different if you are giving constraint or hints instead of value, so use different property

PhilA: eg.., tax registration
... Therfore identifier as an own class. Person: People have identifiers, also.
... Diagram: http://philarcher.org/isa/corevocs.png
... Dbpedia is used for location.
... Address is also difficult: In the end we use vcard.

<olyerickson> PhilA: Address gets all the attention, in the end we use vcard

thanks cygri, you were faster

<olyerickson> ... "addresses are a mine field"

<olyerickson> ...none of these things record change

<olyerickson> ...everything can change, we don't know how to do it

scribe: recognize need to record changes

<olyerickson> ..."more research is necessary" (it is left as an exercise ;) )

DeirdreLee: INSPIRE is complex, but this is their standard, how to resolve issues with relationship with INSPIRE and vcard.

PhilA: INSPIRE and vcard do not much diverge.

<olyerickson> PhilA: INSPIRE directive has legal force

<olyerickson> ..."you shall use it..."

PhilA: will make sure to make proper reference.

<olyerickson> ...usage guidelines will specify how "they" map

<olyerickson> george: Instance examples...illustrating mapping?

george: examples available for mapping to recommended vocabs

<olyerickson> PhilA: usage guiamplesdelines will include instance data/ex

PhilA: does effort match to things you do in the US

DanG: Recommends to look at something. Will give link.

<George> ISO 19773

<PhilA> 19773 ISO

PhilA: Only small amount of uri's needed to be minted in Core Vocabs.

<DaveReynolds> Doesn't seem to be free: http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber=41769

bhyland: title for vocab diagram?

<olyerickson> + to "Thanks PhilA"

<GofranShukair> this MDR standard is only relted to schemas

<HadleyBeeman> sorry — google plus is failing me

<bhyland> Title for the model Phil described "ISA Core Vocabulary Combined Conceptual Model"

<HadleyBeeman> I just wanted to know if member state representation has been… representtative. Are you happy with what you have?

<HadleyBeeman> Ahhh… no wonder. I've signed off from Zakim's telecon

PhilA: some but not enough member state representations.

<HadleyBeeman> Fair enough, PhilA. I was just wondering about status on that. Thanks

PhilA: is difficult to get them involved. Talking to various folks. Open to recommendations to whom they should be talking to.

<DanG> Link to ISO/IEC 19773 is http://webstore.iec.ch/preview/info_isoiec19773%7Bed1.0%7Den.pdf

<DanG> Let me know if this doesn't work.

<olyerickson> WAIT

Procurement use cases.

<olyerickson> is DERI on break

<olyerickson> or just Michael?

MichaelPendleton: Issues, e.g., scope: guidance to people that needs to make that procurment happen (e.g., contract officers). Could be broader, also.
... some text available but can be improved upon.
... community directory part will be talked about later.

<olyerickson> Point of order: could we please provide links before commencing with presos?

MichaelPendleton: Slides: http://www.slideshare.net/michaelpendleton1/best-practices-procurement-of-linked-data-services

<sandro> olyerickson, the link is in the agenda

MichaelPendleton: possible questions regarding procurement, e.g., connection to community directory.
... Should scope be broader?

PhilA: So many ways where procurement is relevant (not only for government).
... How to know about own or external expertise?

bhyland: Institutions need to know what is different in publishing Linked Data from publishing content on the web of documents (which they are doing for a long time already).

<olyerickson> bhyland: "let me help you with some collateral et.al. to get you up-to-speed on different types of services"

<olyerickson> phila: heard a comment, it all sounds terribly US/UK

PhilA: Regarding, internationalizing it. UK is interested in it.

<olyerickson> ...UK public consultation coming up "soon"

PhilA: UK wants to include open standards in their procurement process

<HadleyBeeman> Just to add:

PhilA: but first they need to know what standard

<olyerickson> ...to define what an "open standard" is

<olyerickson> bhyland: need other resources besides John Sheridan

<HadleyBeeman> the reason that this is a big deal NOW is that Cabinet Office now hava new centralised control over procurement— so the outcomes of this consultation can be rolled out to new effectiveness

<HadleyBeeman> I'm here— need to restart G+

<HadleyBeeman> be right back

bhyland: not got john sherridan involved, yet.

<olyerickson> PhilA: Hadley is the best "line of action"

<HadleyBeeman> I'm back now… Can I help?

bhyland: can ask around but does not have a specific person in mind.

<DaveReynolds> Question (since I can't use audio): Is it necessary to have the procurement checklist? Having a model of types of services makes sense. The checklist though seems to be based on a lot of assumptions which aren't universal.

<HadleyBeeman> Thaaaaaanks Bernadette

<DaveReynolds> Not convinced W3C should be advising on procurement principles.

<sandro> DaveReynolds, not principles, just products.

bhyland: glossaries part of recommendation?

<cygri> +1 for glossary

<PhilA> W3C Glossary http://www.w3.org/2003/glossary/

<DaveReynolds> sandro: the principles of technology are OK but principles for *procurement*, NO!

PhilA: Not sure whether W3C glossary is still maintained.

<sandro> DaveReynolds, agreed, absolutely.

<DaveReynolds> sandro: so do you agree with having the current material from the wiki on a rec?

<bhyland> ping

sure, nP

sandro: does not understand procurement

<sandro> george: it's about defining what we're procuring. Defining buying a "linked data server", like they buy a "database server"

<sandro> george: Procurement officers want a checklist.

george: what does it mean to provide a Linked Data server?

<DaveReynolds> Much of current checklist is not that LInked Data specific

<PhilA> Having looked, it seems that the W3C glossary is no longer an option. Neither, at present, is the cheat sheet (http://www.w3.org/2009/cheatsheet/) as to get the data in there is quite a hurdle and I'm not sure it's the best medium anyway so, scrub that idea

<DaveReynolds> Agree with importance, VERY worried about unintended consequences

<olyerickson> Nothing from me; I am relatively clueless on 'procurement"

who creates the action item?

<George> would be very interesting in how you're imagining or examples of unintended consequences DaveReynolds

<bhyland> Summary of Procurement Topic: We need to get from point A to point B... what are the barriers to remove and allow the technical officer to get LOD to the public.

PhilA: Recommends to use the wiki for glossary.

<olyerickson> PhilA: rec'd "nice page on the wiki"

bhyland: takes action to create glossary for best practice document

<olyerickson> bhyland: it's 9:16

<PhilA> ACTION: bhyland to gather terms for a Glossary section in the BP document, then link that static list to the live wiki page where these will be repeated and updated over time [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action07]

<trackbot> Created ACTION-43 - Gather terms for a Glossary section in the BP document, then link that static list to the live wiki page where these will be repeated and updated over time [on Bernadette Hyland - due 2012-02-02].

Does someone want to scribe?

<GeraldSteeman> *Far out

Stability

<PhilA> scribe PhilA

Thanks, Phil.

<PhilA> rreck: Introduces himself

<olyerickson> DC: Lots of noise...

<PhilA> rreck: What could influence the stability of LOD?

<PhilA> rreck: If the info is extremely valuable, it will last a long time

<olyerickson> What are we looking at?

<annew> slides are linked on the F2F page

<olyerickson> Links please before we start presos!!!!

<annew> http://iama.rrecktek.com/html/sites/iama.rrecktek.com/files/Data_Properties.pdf

<PhilA> rreck: is talking through http://iama.rrecktek.com/html/sites/iama.rrecktek.com/files/Data_Properties.pdf

mhausenblas, can you invite me to the cool kids?

<annew> second part of stability presentation will be http://www.w3.org/2011/gld/wiki/images/0/06/W3C-Washington2012_F2F2.pdf

<DaveReynolds> George: examples include "government approved contract vehicle" which tend to be biased towards large companies, "interface to load data" which encourages a batch/static mindset instead of live integration, "Is the vendor or provider an active contributor to Standards groups" which again biases towards large companies.

I thought I had invited to one of my circles.

<olyerickson> @annew yes but it is hard to juggle...we need "we're now looking at..."

<George> thnx DaveReynolds understood

<annew> understood olyerickson

<PhilA> I'm not scribing this as rreck's slides are a good record

<bhyland> Should HTTP URIs and attention to proper MIME types by the Linked Data client manage all this correctly?

<PhilA> PhilA: Pointing out that recommending not to use diacritic characters is not an internationally acceptable recommendation.

<dvilasuero> +1 PhilA

<PhilA> bhyland: Not sure whether this is relevant to the Gov Linked data WG - it's more generic and not specific to LD

<PhilA> sandro: I share bhyland's thoughts

<olyerickson> pop me from queue

<George> sandro: stability here is about URI stability

<PhilA> sandro: not sure the archive format is relevant. What is relevant in this space is URI persistence

<bhyland> Sandro:Stability is about maintaining persistent of a link, over time, that is the problem we are trying to solve.

<PhilA> PhilA: +1 - that's what I thought this session would be on

<PhilA> olyerickson: I wanted to disagree slightly with the point just taken

<PhilA> olyerickson: This touches on one of the conversations that we had yesterday

<olyerickson> pop me from queue

<bhyland> Olyerickson: Disagrees slightly. Some of this touches on yesterday's conversation re: using good data management practices, but we should provide treatment of LOD differences.

<PhilA> olyerickson: It doesn't hurt to say that stakeholders should be using good data management lifecycle practices and then say that there are specific issues wrt LD that may not otherwise be obvious

<bhyland> ... The issue of different archive files is a problem RPI deals with daily. GB sized datafiles for download are non-trivial problems that we deal with somewhat uniquely within the LOD community.

<PhilA> DanG: I think I agree with olyerickson

<olyerickson> The last speaker was me ;)

<sandro> yeah, but we don't have the expertise/right to speak to data in general.....

<PhilA> DanG: Some of these recommendations are general and not specific to LD but they are specific to data and much of what is going to be said does still matter

<olyerickson> My point was there are *some* practices that are indeed unique to Linked Data

<George> +1 sandro

<HadleyBeeman> Are there other, more general W3C practices for data management that we can take in reference?

<sandro> mhausenblas, yeah, I haven't been working on that this morning.

<olyerickson> ...a provider can't trivially choose between formats they make available, etc...

<PhilA> bhyland: Suggest that we deal with - recognising that when you put data out ...

<PhilA> sandro: When i use RDF I refer to other people's URIs and I need some confidence in their management or my stuff will break

<PhilA> annew: That's part of the stability issue

<PhilA> bhyland: The other is dealing with very large file sizes.

<PhilA> bh: People don't download to their laptopns they work with them in the cloud

<PhilA> s/laptopns/laptops/

<olyerickson> @sandro I think I see the point --- hard for me to articulate, but this isn't about back-end stability, more about keeping the Web of Data stable

<annew> good question Hadley

<PhilA> Mike_Pendleton: I'm happy to go back and make some revisions based on this interesting discussion

<George> http://www.w3.org/2011/gld/wiki/images/0/06/W3C-Washington2012_F2F2.pdf

<PhilA> annew: Talking through http://www.w3.org/2011/gld/wiki/images/0/06/W3C-Washington2012_F2F2.pdf

<bhyland> Summary of this topic (stability): 1) Stability is about maintaining persistent of a link, over time, that is the problem we are trying to solve. 2) Dealing with large file sizes.

Stability (continued) by AnneW

<sandro> (Great, this stuff looks more in line with what I was expecting.... )

<bhyland> +1

<bhyland> AnneW: Best practices for Stability is all about "Making data available in perpetuity, persistently archived if necessary."

<bhyland> ... It is predictable, follows a logical format.

<bhyland> ... It is Externally visible and has stable consistent locations

<bhyland> ... provides a Legacy - uses earlier naming schemes, formats, data storage devices

<bhyland> ... Has stewards, someone needs to take responsibility - people are committed to consistently maintain specific datasets

<olyerickson> annew hadleybeeman was that "We have to have the Ability for Stability?"

<bhyland> ... If this is tied to support of a business unit, or legislative function, that is key.

<HadleyBeeman> Yep, olyerickson. That's the one. /cc annew

<bhyland> Contact and data consistency are related to success. Consistent infrastructure, and separation of internal politics and external stability.

<bhyland> .... It is a good to have a data "last will & testament" to ensure your wishes are maintained in perpetuity.

<olyerickson> The problem that estates have is that even if they are 'required' to persist (by policy or statute) they still might not b sustainable (no unfunded mandates)

<bhyland> ... I don't know what model we're going to recommend yet, but there are some for consideration.

<bhyland> Some models for stability include: Estate, Archives, Private Foundation, Government stewards, Internet organizations

<bhyland> Key concepts around interconnections. These are sources that establish a context for the production and/or use of an artifact. W3C Provenance. There are mechanims for persistence. Some are PURLs, handles.

<bhyland> Key considerations: Preservation of Content, Preservation of Access, Conservation of media

<bhyland> ... Other considerations include how long term is long term? 2 years, 10 years, forever? Organizations need to consider this.

<annew> Rock on 8 track tapes!

<PhilA> http://www.w3.org/2011/gld/wiki/images/0/06/W3C-Washington2012_F2F2.pdf

<PhilA> http://www.w3.org/2011/gld/wiki/images/0/06/W3C-Washington2012_F2F2.pdf

<PhilA> Let me try again http://www.w3.org/2001/tag/2011/12/dnap-workshop/notes.html

<annew> chia ching!

<bhyland> Questions: PhilA mentioned workshop held earlier this month. Longevity differs depending on who you are speaking to. For example, a librarian things in centuries.

<cygri> .cz has been gone *twice*

<HadleyBeeman> PhilA: we were reminded this weekend that the average life span of a government department in the UK is 5.5 years.

<bhyland> Various level domains have disappeared due to geographic boundaries changing, including countries.

<HadleyBeeman> Big implications for URI persistence/stability

<bhyland> ... Agreement with an institution in advance. For example, if W3.org goes away, then MIT.edu will pick it up.

<olyerickson> RE "one of the registrars, it may have been someone from CNRI (representing Handles) or crossref (representing DOIs, based on Handles)

<bhyland> ... Top level domain for .arpa, only available if you have an RFC (Request for Comment). If you coordinate with IETC, you can get a .arpa domain.

<bhyland> .... someone suggested W3C should have w3.arpa, it is so important.

<annew> 8 track = google +

<bhyland> ... DNS has been around for many decades, even longer than 8-track cartridges (which the cool kids know prevailed as a better technology solution)

<George> 'for the life of the republic'

<bhyland> s/know prevailed/know should have prevailed/

<Zakim> PhilA, you wanted to highlight persistence WS http://www.w3.org/2001/tag/2011/12/dnap-workshop/notes.html

<PhilA> olyerickson: Thanks annew for the talk.

<PhilA> olyerickson: I was involved in the handle.DOI discussion

<bhyland> olyerickson: Thanks to AnneW for a very comprehensive talk that highlights her library science expertise/perspective.

<PhilA> olyerickson: When people set up institutional archives, a concern has been around what naming structure to use

<bhyland> ... Handles system - single source is a concern. What if CNRI went out of business, or Bob Kahn got hit by a bus.

<annew> DOI = Digital Object Identifier. A consortium of publishers created for commercial handles

<bhyland> ... people were concerned about (one or two) publishers backing DOIs ... too much central control.

<bhyland> ... Believes our WG's recommendations should help people make decisions.

<annew> handles http://handle.net from CNRI

<bhyland> ... PURLs infrastructure mentioned.

<bhyland> bhyland

<olyerickson> I think Web of Data (and persistence of links) adds new meaning to "preservation"

<bhyland> Yigal: NSF had a research program circa 2005 that looked into preservation of data. I remember a few things of note: 1) in order for things to be preserved, you must retain *interest* in the data. Maintenance of electronic data requires "care and feeding", a continued expense.

<olyerickson> ...continuity of access <-> stability

<bhyland> ... if there is no perceived interest, these things will not be made available.

<cygri> very good point - data won't survive loss of interest

<bhyland> @GeraldSteeman - can you chime in here about STI?? NASA's work???

<PhilA> PhilA: Notes that W3C does not print its Recs. I believe that the IETF does print its RFCs

<Cory> I doubt storing paper is cheaper!

<annew> storing paper is different than migrating data.

<annew> migrating data required continuous activity and pruning and consideration

<annew> :-)

<bhyland> GeraldSteeman: Pressure within scientific & technical information community to digitize and provide guidance on longevity, access and of course, persistence.

<bhyland> ... If there is no interest, is there no value?

<bhyland> ... perceived value may change over time. Interest may change over time.

<annew> and the first electronic census is lost

<bhyland> DanG: shares the story about the circa 1960's data on some tape format that cannot be read. The 9 track tape readers have all gone to the dump and cannot get at the data.

<HadleyBeeman> That's really sad

<bhyland> ... census data from 1960's data is lost while we have paper based records from the 19th centry that persisted.

<annew> this is the danger of going electronic without considering the long term storage of it.

<Zakim> cygri, you wanted to comment on preservation of content vs preservation of access

<PhilA> cygri: There was a question about the disctinction between persistence of content and access

<PhilA> cygri: not sure where the draw the line between the 2

<bhyland> Cygri: persistence of access and content ... where do you draw the line.

<PhilA> cygri: I create a bunch of files with the RDF and structure to support it

<olyerickson> cygri: where to draw the line, persistence of data/persistence(??) of access

<PhilA> cygri: Once the content that I set up no longer dereference, a big part of it has been lost

<Yigal> s/19th/18th

<bhyland> ... for content to persistent because someone has a copy of these files. But once the URIs don't de-reference any more, the value has been lost.

<HadleyBeeman> +1 to cygri's point that we need to preserve access, in addition to the content itself. Preserved content that no one can access isn't useful.

<annew> olyerickson... you just aren't cool ;-)

<PhilA> bhyland: Tries not to get into the nit picking between URI and URL. After lucnh we'll have a session on HttpRange14

<bhyland> Thus preserviing the value of URIs is very important, URIs must remain resolvable (agreed to not open URI vs URL discussion here ...).

<PhilA> cygri: The 2nd question concerns the stages that might kill the stability of deployed data

<bhyland> We can say there are staged events to help stability of deployed content ...

<PhilA> cygri: Can the thing outlibe the individual who set it up departing the organisation?

<olyerickson> HadleyBeeman, the point the Handle/DOI community has made for years is that having "persistent IDs" in the infrastructure is useless if there is no commitment to maintain continuity of access

<PhilA> s/outlibe/outlive/

<bhyland> 1) Can the data outlive the steward. Someone leaves and can their organization be a good steward?

<olyerickson> ...and that requires back-end maintenance etc

<PhilA> cygri: Can it outlive the death of DNS? Western civilisation?

<bhyland> 2) Can it outlive the DNS and death of Western Civilization as we know it on the Internet?

<bhyland> ... Of course DNS has been stable for some decades, but change happens.

<PhilA> cygri: Individuals leave all the time, but DNS seems pretty stable for now

<annew> zen philosophy... zen and the art of rdf maintainence

<bhyland> ... It would be good to include advice on moving top level domains, and there are strategies for handling this. We should provide guidance on this.

<bhyland> ... Mention there are costs associated with URIs.

<bhyland> AnneW: It is about a stability of infrastructure that we're talking about.

<olyerickson> annew: it's "stability of infrastructure' we're talking about

<PhilA> annew: I'll take the point about accepting that it's about managing change, not working to avoid it altogether

<bhyland> AnneW: I'll fold in guidance on gracefully handling change ... save the end of Western Civilization or the entire economy changes ...

<olyerickson> ...we need to bound conversation (no questions about economy collapsing, etc)

<PhilA> sandro: has said a lot of what I wanted to say

<bhyland> Sandro: I think this WG needs to be concerned about keeping a given URL working.

<olyerickson> sandro: this group needs to be concerned about "keeping that URL working"

<cygri> +1 sandro

<PhilA> sandro: The idea that some copy of the data survives somewhere is separate. What we need to focus on keeping the URL live

<olyerickson> ...Need new second-level domain for any new "project" that needs to persist

<PhilA> sandro: Every time you have a new project, set up a new domain that can be moved from person to person, organisation to organisation without any change

<bhyland> ... everytime I hear this conversation, I think stability will be assured by assigning a new atomic, unique top level domain. I know this is not a consensus position.

<olyerickson> ...not a consensus, but to be considered

<bhyland> For example, thing1.org, thing2.org.

<olyerickson> ...Domains that can move and have their own "life"

<annew> +1 .data

<PhilA> I want to see if people have strong feelings on purls?

<olyerickson> bhyland: want's to here if people have strong opinions on esp PURLs

<PhilA> bhyland: It was news to me that the LoC is using handles/DOIs and not just pruls

<PhilA> s/pruls/purls

<bhyland> AnneW: For FDSys, all the new documents are handles not PURLs.

<PhilA> annew: basically new stuff is being given handles, purls are seen as old system although is maintained

<PhilA> sandro: Speaks against the use of PURLs

<PhilA> sandro: They may be a necessary evil but they add another point of failure

<annew> PURLS tradmarked = (Persistent Uniform Resource Locators) http://purl.oclc.org

<PhilA> sandro: If you can't get your own domain name, you have your own problems...

<olyerickson> ...Note that Handle System is parallel to DNS, not dependent on it.

<cygri> i like purls.

<annew> i like pearls...

<PhilA> PhilA: Notes that bhyland has a business that sells PURLZ servers

<PhilA> PhilA: There is a difference between PURLs and purl.org

<olyerickson> 90% of Handle System use is through rpoxies, but HDL infrastructure is independent

<olyerickson> s/rprox/prox/

<PhilA> GeraldSteeman: I wanted to agree with sandro on the pre-PURL discussion

<olyerickson> s/rpox/prox/

<annew> Government Printing Office GPO http://www.gpo.gov/fdsys/

<PhilA> GeraldSteeman: If it can live on its own then its value can be passed from one person to another

<bhyland> @PhilA, note, I don't have a business that sells PURLs servers. PURLs is a FLOSS project. I run a company that provides commercial support for a given instance of a production server running PURLs by the US GPO.

<PhilA> GeraldSteeman: The US has an 'official record' and gov officials are bound to submit it to the national archives so that there is an archived version of that info

<PhilA> s/Notes that bhyland has a business that sells PURLZ servers//

<bhyland> It is the difference between a company running a production Apache server for a customer and the Apache Project. We are not the PURLs project, just a user of it :-)

<Zakim> cygri, you wanted to mention danbri's law

<cygri> danbri's rule: "Rule of thumb - when wondering what info to include in a namespace URI, ... try to leave *out* as much as possible"

<annew> The items printed by the GPO, official record, can be accessed through handle technology

<PhilA> cygri: I wanted to say +1 to sandro's point about getting a new domain for a project

<bhyland> Cygri: Danbri's advice is 'leave it out if possible', any thing that can change should be left out if possible.

<PhilA> cygri: schema.org URIs are short, easy to remember and a good example

<annew> Lunch is HUNGRY

<HadleyBeeman> calendar says 1.5 hours

<annew> Thanks all for a great discussion

<HadleyBeeman> so— 1 hour?

<mhausenblas> reconvene at 11:15 and 4:15

<olyerickson> What was the question?

<cygri> thank you annew!

<t_gheen> \quit

<HadleyBeeman> bye

<olyerickson> mhausenblas I lost your G+ audio...

<mhausenblas> Michael: The latest ED of the BP document is now available via https://dvcs.w3.org/hg/gld/raw-file/a1cb0d2a8fc9/bp/index.html

<olyerickson> whoa, that was noisy...

<mhausenblas> Galway's ready

<PhilA> Here's a piece relevant to the discussion earlier about procurement and what the UK Cabinet Office is doing http://www.computerweekly.com/news/2240114381/UK-Governments-prepares-to-open-up-IT-procurement

<cygri> dcat slides: http://www.w3.org/2011/gld/wiki/images/2/28/Dcat-gld-f2f2.pdf

DCAT

<Mike_Pendleton> Richard covering DCAT

<PhilA> scribe: gatemezi

scribeNickname: gatemezi

cygri: slides are on the wiki...
... http://dvcs.w3.org/hg/gld/raw-file/default/dcat/index.html
... is not a publicly editor draft

<Mike_Pendleton> DCAT Editors Draft available

cygri: have dataset, can have distribution...

<PhilA> Editor's draft *is* now public, however, do not take any notice of the URIs for it - this has not been agreed

cygri: with the different versions of the data (csv, xml,..)

<Mike_Pendleton> many properties from FOAF

cygri: DCAT has been shapered along withi the eGov .
... still in the process to move stuff in the W3C domain

<Mike_Pendleton> connecting use cases and requirements

cygri: collected some requirements...

<olyerickson> cygri: Slide 5, use cases inherited from earlier eGov discussions

<olyerickson> ...Q: what to do with these use cases

<olyerickson> ...may make sense to separately publish

cygri: what will we do with those requirements? publish them as note?

<Mike_Pendleton> publish use cases as notes is a possibility

cygri: certainly have things to be added...some works up-to-date this morning

<Mike_Pendleton> identifed open issues in GLD tracker

cygri: Some issues were raised up...
... four people are interested in this work so far

<Mike_Pendleton> Phil, John, Fadi and Richard are woking this issue

cygri: next steps for the following weeks are also described in the last slide

<Mike_Pendleton> Richard asked for those interesting to join

<Mike_Pendleton> interested not interesting...

<olyerickson> FPWD => "First Public Working Draft..."

cygri: work through the tracker issue...

<Mike_Pendleton> thanks PhilA

cygri: people aked how does it related to ADMS, etc...?
... there are of course of alignments but till now not really good answer for that

<mhausenblas> Michael: The diffs/overview etc. re DCAT, VoiD, etc should be explained via the BP deliverable

<Mike_Pendleton> Richard: how can we ready as many publishers of data catalogs as possible; need outreach

cygri: outreach to publisher should be to take in consideration
... comments ? questions?

olyerikson: issue in complexity into the draft that we don't necessary need
... we have issue for example about language

<Mike_Pendleton> Q from olyerickson: draft has no official status, but stable for some time; concern about legacy issues discussed add complexity to process and draft very late that we don't need; we have had little issue with DCAT (other than language); concerned about working issues we don't need to

s/olyerikson/olyerickson

<olyerickson> cygri yes, that was my largest point

cygri: we should be aware of existing deployment

<Mike_Pendleton> Richard: We need to be aware DCAT is being used and we should be aware of that when considering changes; however, it is a draft and not comfortable with saying we shouldn't change it

<olyerickson> ...and yes, the issues were tabled, not actually dealt with

cygri: it's still a draft, but we need to talk about the issues raised by people

philA: an issue does not need to follow the changes in the draft

<Zakim> PhilA, you wanted to talk about outreach

philA: we can put the same time org, best practices in the same time.

<George> would like to hear more about CKAN DCAT 'funniness'

<olyerickson> PhilA: "CKAN uses DCAT, but in a 'funny way'"

philA: CKAN is a big use case of the use of DCAT.

cygri: need to articulate what we need in other to them to use (CKAN)

<mhausenblas> Michael: I think we can summarise this - we (DERI) can take care of CKAN/OKF liaison as we work together on project basis (LATC, LOD2)

<bhyland> @mhausenblas, thank you. This is really important. Can you update the GLD WG on whatever frequency makes sense as we're really interested outreach to OKF.

<mhausenblas> PROPOSAL: Richard C acts as liaison point for OKF/CKAN

<bhyland> +1

philA: if DCAT is not attractive to CKAN, there may be a problem somewheere

<Zakim> mhausenblas, you wanted to comment on the process

<HadleyBeeman> Happy to help, Cygri, if you need someone to chat to them on the ground in London.

<HadleyBeeman> (though it sounds like you have it all in hand)

<PhilA> thanks HadleyBeeman - prob take you up on that

<olyerickson> CKAN "funny" use of DCAT doesn't mean DCAT is 'wrong' any more than RPI use of DCAT means it's right ;)

<Mike_Pendleton> mhausenblas: goal to get out working draft; define processes

<cygri> HadleyBeeman, it will be good if they hear about dcat from multiple directions

<bhyland> +1 mhausenblas

<mhausenblas> PROPOSAL: Richard C acts as liaison point for OKF/CKAN regarding DCAT

<bhyland> +1

<stasinos> +1

mhausenblas: propose cygri to be the liaison to CKAN

<BartvanLeeuwen> +1

<boris> +1

<olyerickson> +1 to mhausenblas proposal

<sandro> ( I don't think we need a WG resolution on this kind of thing. :-)

<BenediktKaempgen> +1

<HadleyBeeman> you're welcome :-D

<sandro> +1

<Mike_Pendleton> Hadley will support Richard

<bhyland> Noted: Hadley also in a good position to liaise with OKF re: DCAT

<olyerickson> Do we need a proposal w.r.t. specific work on issues?

<Zakim> fadi, you wanted to comment on reach out

<mhausenblas> RESOLUTION: Richard C acts as liaison point for OKF/CKAN regarding DCAT

fadi: CKAN uses DCAT in two services

<olyerickson> ... need a proper and clear statement of DCAT

<olyerickson> Repeat the question?

<Mike_Pendleton> George: Next step questions posed - do we want to work on them?

george: do we discuss the next steps now?

<PhilA> The question was about non-CKAN usage of DCAT and what we could learn from that. Fadi noted that a number of Spanish catalogues use DCAT but all do it differently

<cygri> http://www.w3.org/egov/wiki/Data_Catalog_Vocabulary/Use_Cases_and_Requirements

cygri: For the requirements, they have text for that, it is just to pass them to WG

<mhausenblas> +1 to turn the UC into a WG Note, that is non-REC-Track

<fadi> +1 to mhausenblas

<olyerickson> Wondering specifically what spanish catalogs Fadi is referring to. We haven't seen (much if any) dcat use in the wild

<PhilA> cygri: Arguing that publishing the UCs as a draft WG Note helps to attract more review and interest

<olyerickson> ...is fadi referring to CTIC?

<Mike_Pendleton> Bernadette: Process for note?

<bhyland> W3C process for Working Note vs just having a more mutable wiki page.

<fadi> olyerickson: catalonia http://www20.gencat.cat/portal/site/dadesobertes

philA: need a resolution for DCAT use case..

<cygri> olyerickson - catalonia, barcelona, gijon and balearic islands use it in their local catalogs

<bhyland> Per PhilA, it is a shorter process than Recommendation. Circulate email to TLR, chairs with intension to publish Working Notes with some descriptive context as to why we're we doing this.

<bhyland> ... Working Notes that supports Recommendation and carries slightly more heft than the wiki page.

<mhausenblas> Michael: http://www.w3.org/2005/10/Process-20051014/tr.html#q75

<sandro> to me "resolution" == "decision" == "written record of an agreement"

<bhyland> +1 @sandro

<olyerickson> -1 to tracker issues now

<sandro> trackbot, how the heck are you?

<PhilA> PROPOSED RESOLUTION: That the DCAT Use Cases be formatted as a draft WG Note and that the chairs seek permission to publish at /TR/dcat-use-cases

@philA: thanks!

<olyerickson> +1 to The DCAT Group having separate call to thrash through, preceded by email discussion

<DanG> I think there needs to be a general discussion before some group goes off by itself to look at an issue

<Mike_Pendleton> How do we want to address tracker issues? Can we use thursday call or separate call

<fadi> +1 to separate call

<trackbot> Sorry, sandro, I don't understand 'trackbot, how the heck are you?'. Please refer to http://www.w3.org/2005/06/tracker/irc for help

<stasinos> +1 to mostly emails, with separate call at the end

<olyerickson> +1 to The DCAT Group reporting back to larger group

<Zakim> mhausenblas, you wanted to talk about WG Notes

<HadleyBeeman> +1 to more content on the weekly calls.

<Mike_Pendleton> Bernadette recommends working on specific issues on our regular weekly calls

bhyland: should there be a specific telecon for DCAT group?

<Zakim> PhilA, you wanted to make a proposal regarding what we talk about today cf. other occasions

<olyerickson> arrggghh

<fadi> https://www.w3.org/2011/gld/track/issues/raised

<bhyland> Cygri: If others think it is a good use of our weekly telecons to dig deeper into DCAT to help streamline it moving forward, please indicate your interest.

<bhyland> +1 by bhyland

<Mike_Pendleton> PhilA: looking at the well catalogued issues, some are detailed about individual properties, and thos should be handled among interested folks on a separate call; others are of broader interest

<olyerickson> phila: is referring to the "distribution" stuff

<Mike_Pendleton> PhilA: mechanism for what is in a zip; like CKAN has

<annew> are boiled frogs legs necessary to jump the queue?

olyerickson: there are some issues to be talked by emails and after if necessary others in the regular call

<Mike_Pendleton> olyerickson: if there are high level issues which should be discussed today or on reg. call; other low level issues on email/breakout call; leave it to chairs/rest of group

<George> i think that's consistent with PhilA 's suggestion

<Zakim> mhausenblas, you wanted to talk to fragmentation vs. efficiency

@Mike_Pendleton: thanks

<Mike_Pendleton> @ gatemezi: it takes a village

mhausenblas: how details of the discussion do we like ? Efficiency arises when the small group meets together

<Zakim> sandro, you wanted to suggest (re dates) we *not* merge data with comments on the data, in the same field

<Mike_Pendleton> Sandro: date question - approach of having typed value if known or string if you don't know the date

<csarven> IMO better to make it easier for the publisher than for tool builders (writing code around stuff)

<olyerickson> sandro is concerned about best way to handle "uncertainty" around dates

<Mike_Pendleton> Sandro: Handling uncertainty is dates - labeling and annotation

<olyerickson> ...clear what to do when it's known e.g. xsd: date but what to do when NOT known

sandro: is there any research going on in handling uncertainty?

philA: in half the time publishers don't type very well the data

<sandro> *nod* this is about whether the burden goes on the publisher or consumer....

<Mike_Pendleton> PhilA: half the time data publishers use the wrong date type - better to use two versions

<Mike_Pendleton> George: Make it easier on the publisher

<stasinos> http://placetime.com/

<olyerickson> +1 @george being TOO specific e.g. with dateTime or gYear tec can b a problem

cygri: if you have 2 properties, you have to write how to use them, no complexity in term of coding
... DCAT takes existing metadata of data and expose them in RDF

<mhausenblas> allow 30 sec

<bhyland> cygri: Purpose of DCAT is to take existing databases of metadata and exposing them as RDF. Anything that requires recoding of dates is COMPLETELY out of the question for cygri.

<bhyland> ... quantfy uncertainty is too difficult.

s/quantfy/quantify

<olyerickson> +1 to cygri being persuasive ;)

<bhyland> Sandro: Is persuaded by cygri's argument.

<olyerickson> The use case of legacy dates MUST NOT be ignored

<Zakim> PhilA, you wanted to suggest a third way that I don't like but recognise it's a possibility

<bhyland> Discussed: Data accurary using XSD Date which is the most ubiquitous date format

<olyerickson> That's up to the consumer

s/accurary/accuracy

<olyerickson> bhyland: chair interrupt

<Mike_Pendleton> Bernadette: Gov't orgs have to accept whatever; they accept data that sometimes they can't clean up.

<cygri> (but: unclean data in - unclean data out)

<olyerickson> bhyland that's the point, we're talking about a string-bucket as a roll-over

<bhyland> Discussion is about how do we handle and not fall over with systems give us ill-behaved/ill-formatted data.

<bhyland> ... Is our recommendation 'leave it as a string' or data type it if we DO know it is a date.

<olyerickson> +1 to xsd: date preferred, string literal otherwise

<bhyland> +1

<George> +1

<boris> +1

<olyerickson> ;)

<dvilasuero> +1

sandro: if you know the date, use xsd:data; if not use it as string

<sandro> +1 (untll some folks come with a compelling rdfs:range use case)

<csarven> A literal doesn't necessarily imply anything more than what it really is (unless further context is give). The interpretation is up to the consumer. 2012-01-26 (would be on that day, or the whole day, and not necessarily on the midnight of that day).

stasinos: propose not to allow string

<bhyland> Stasinos recommends looking closely at PlaceTime.com - URIs for Places and Times. PlaceTime.com is intended to be a URI space containing URIs that represent places and times.

<stasinos> stasinos proposed to not allow *unstructured* strings

philA: it is difficult to ask providers to reformate their data

<Zakim> mhausenblas, you wanted to propose close queue

<Mike_Pendleton> Mhausenblas: are there other things we want to cover?

mhausenblas: do we have another important issue to discuss?

<olyerickson> +1 to shaken, not stirred

<olyerickson> +1 to it being 5 O'Oclock Somewhere

philA: how do we use DCterms issued?

<Mike_Pendleton> PhilA: Issue 2 :-) In definition of DCAT, cube data, etc, we will be saying 'use this term' and this is how we want them to use it; is it right to make range statements about others terms?

<mhausenblas> Michael: I'd prefer a soft advise as we did in VoID such as in http://www.w3.org/TR/void/#dublin-core

<olyerickson> +1 to "anyone can say anything about anything" http://bit.ly/yZvdpk (RDF def'n)

<BenediktKaempgen> +q

<Zakim> mhausenblas, you wanted to comment on usage advises on external vocabs

<fadi> +1 mhausenblas

<boris> +1 usage notes michael

+1 to the AAA slogan

@olyrickson: is not "anyone can say anything about any topic" ? :-)

<olyerickson> @gatemezi I'd like to understand why not. The way it is being presented, it sounds like that to me; this is why I'd like some examples

<PhilA> olyerickson: There's a specific case cited in the issue http://www.w3.org/2011/gld/track/issues/4

<Zakim> cygri, you wanted to mention third (OWL) option

<Mike_Pendleton> DanG: in standards community; conformity is the notion, not compliance, and extends to a technical spec. For example, US Open Govt vocab working group; as long as vocab conforms to the spec., all is good; takes us out of the business of telling people what and how in particular circumstances

<olyerickson> Thanks @PhilA, I think I get it. Problem being discussed is, cases where a 3rd party term is used but with an added restriction

@Mike_Pendleton: thanks!! it is sometimes difficult to scribe alone :-)

<olyerickson> ...ie "we use foo: xyz BUT only with range xsd: fooBar "

<DanG> I think my statement is consistent with what Richard is saying.

cygri: actually OWL is powerful to make explicit context about datasets, or DCAT

<PhilA> I like what cygri is saying about using OWL - that sounds like a solution to me (along with a textual note)

<stasinos> q

<stasinos> OWL might be overkill for this particular use case

DeirdreLee: saying in prose in open to interpretation, formalise it is better; agrees with cygri

<Mike_Pendleton> deirdreLee: DCAT usage is open to interpretation

<sandro> there's a difference between conflicts due to errors and conflicts due to people redefining other people's stuff

<cygri> yes - defining a subproperty is always possible and avoids the problem

+1 to defining a subproperty

<stasinos> +1 to backward-chaining -1 to tableaux

<cygri> yes - defining a subproperty means inventing a new property.

philA: Ok for the defining a subproperty; but problems come when using a reasoner using inferencing
... using owl is the wright thing

<Mike_Pendleton> PhilA: stasinos idea is the most accurate solution; however Richard pointed out that most linked data systems don't involve reasoners, so it doesn't matter

<olyerickson> Can someone please explain why dcat's use of dc dates is not an example?

<olyerickson> dc doesn't say anything about xsd: date

<PhilA> Possible Ways forward: 1) define the sub domain and range as we see fit and don't care 2) only define domain and raneg in a usage note specific to DCAT etc, 3) define ranges in OWL

@cygri: do we have a link of the slides?

<cygri> gatemezi, which slides?

<BenediktKaempgen> slide 8: http://www.w3.org/2011/gld/wiki/images/2/28/Dcat-gld-f2f2.pdf

<cygri> ah ok

@benediktKaempgen: thanks

<cygri> option 4) define subproperties and subclasses

+1 for option 4)

<olyerickson> PhilA: By "Combination of 2 & 3" do you mean prose + RDFS, or RDFS + OWL, or???

<cygri> option 5) it depends.

<PhilA> olyerickson: I mean use OWL and text, not RDFs

<csarven> 6) all of the above

<annew> Can we have a vote after we have a short break? and a clarification of the question

<stasinos> 7) do nothing

<olyerickson> PhilA ah okay

<mhausenblas> 7) none of the above?

<stasinos> 8) go for dinner

<olyerickson> me notes it's waaay after 5p in Galway ;)

<mhausenblas> s/dinner/beer

<mhausenblas> yeah, like 51min

<annew> s/dinner/martini

<George> 1) define the sub domain and range as we see fit and don't care

<cygri> -0.8

<BenediktKaempgen> -1

<mhausenblas> +0.2

<boris> -1

<Mike_Pendleton> rreck +1 1

+0.3

<George> +.03

<bhyland> abstain

<olyerickson> Who is "we" (I'm serioues)

<stasinos> + 1E-1

<csarven> nil

<annew> Anne will give promises of preservation and stability to whoever buys drinks

<olyerickson> No we're not

<olyerickson> Trying to understand who "we" is

<DanG> Will someone please write down a succinct atement of the question?

<annew> +1 dang

<George> 2) only define domain and raneg in a usage note specific to DCAT etc

<fadi> +1

<DanG> s/atement/statement

s/atement/statement

<cygri> s/atement/drinkment/

<Mike_Pendleton> abstain

<stasinos> this was 3 before, wasnt it?

<cygri> +0.8

<PhilA> DanG: When publishing the machine readable version of a vocabulary, how should we indicate that in the context of that vocabulary, terms from other namespace should be used in a way that is more restrictive than is defined in the original definition

<stasinos> -1

<BenediktKaempgen> -1 (then we do not need to call it domain/range)

<DeirdreLee> -1

0.2

<mhausenblas> +1

<csarven> +1

<George> 3) define ranges in OWL

<fadi> -1

<olyerickson> WHO IS "WE"????????

<stasinos> +0.5

<DeirdreLee> +1

<csarven> --0

-1

<GofranShukair> -1

<cygri> +0.5

<mhausenblas> Michael: On a serious note - can we delegate this to the DCAT TF - this doesn't make sense to me at all

<annew> </childishBehavior>

philA: WE=vGLD group

<olyerickson> Thanks! Sorry for exploding

<cygri> +1 mhausenblas. we've collected the options and heard some opinions and arguments, that's perhaps as good as we can do here.

s/vGLD/GLD

<stasinos> here comes (8) again: do nothing

<HadleyBeeman> +1 to bhyland and mhausenblas

<annew> +1 break

<olyerickson> Let's package it up as a proposal in email and thrash it there

<BartvanLeeuwen> +1 for break

+1 break if we close the "debate"

bhyland: break 10 minutes

<HadleyBeeman> okay.

<olyerickson> wondering what galway is looking at (curious)

<olyerickson> hey Galway, unmute

<mhausenblas> galway coming back

Review of all outstanding Documents that the WG is updating

<PhilA> scribe: fadi

bhyland: let's review outstanding documents we are updating

<Mike_Pendleton> Bernadette: How to keep momentum and make progress on a montly basis

bhyland: we need to keep the great momentum we have
... what's our strategy for the upcoming tele-calls?

<mhausenblas> Michael: Can we first agree on FPWD please? And then have the meta-discussion?

bhyland: we already have consumers involved
... how to keep the moentum then, suggestions/comments?

s/moentum/momentum/

FPWD Best Practices timing

<Zakim> mhausenblas, you wanted to update people about BP and to suggest a timeline for FPWD 9 Feb and 23 Feb

<mhausenblas> See https://dvcs.w3.org/hg/gld/raw-file/default/bp/index.html

mhausenblas: we had a quick session in the moorninng updating an editor draft

<cygri> can one of the google plus kids invite me again?

mhausenblas: feel free to raise issues and provide feedback

<mhausenblas> PROPOSAL: The GLD WG decides to publish BP, DCAT, QB and ORG as FPWD on the REC-Track by end of Feb 2012 with 9 Feb ready for WG review and final decision to be taken on 23 Feb.

PhilA: do we need the two weeks between 9 and 23?

mhausenblas: probably yes... based on experience, people need this time to review
... talking about WG members

<bhyland> +1 to proposal timing and 1 week review cycle. IF needed, we can extend, but better to put a fire under people's bum

<mhausenblas> Michael: I suggested 2 weeks review cycle, bhyland, just to clarify

bhyland: we are all here, so we can make a decision
... anyone has particular input to people here who are working on the best practice?

<mhausenblas> PROPOSAL: The GLD WG decides to publish BP, DCAT, QB and ORG as FPWD on the REC-Track

mhausenblas: any other document should go the REC-track?

bhyland: this looks a lot giving that some people expressed that they are busy currently

<olyerickson> bhyland yes DaveReynolds was either working QB or ORG

<sandro> Yes -- it's okay to have placeholders for sections of the document.

cygri: is it acceptable for the FWPD to have big gaps like entire section not written yet?

mhausenblas: yes

<bhyland> DaveReynolds is key person on ORG. What is being proposed is that 4 documents as FPWD.

<sandro> the key thing to to give people an idea which way we are heading, and give them something interesting to review.

cygri: standard vocabulary doesn't seem covered in the list

<sandro> right, People Vocab is missing.

cygri: referring particularly to the "people"
... mentioned in the charter

PhilA: as part of the core vocabulary work we are working on a "person" vocabulary
... there will be soon a version available in the public domain
... this might be used as basis for representing people
... not going with FWPD
... but just pointing that we tackled the "people" representation topic

<mhausenblas> PROPOSAL: The GLD WG decides to publish BP, DCAT, QB and ORG as FPWD on the REC-Track, while people area will be addressed later (based on ISI draft)

PhilA: the ISA person vocabulary can be, when available, a candidate

<mhausenblas> PROPOSAL: The GLD WG decides to publish BP, DCAT, QB and ORG as FPWD on the REC-Track, while people area will be addressed later (based on ISA draft)

<bhyland> PhilA: Hoping that people in this WG would consider as a candidate for endorsement the ISA Person Core Vocbulary ..

<mhausenblas> PROPOSAL: The GLD WG decides to publish BP, DCAT, QB and ORG as FPWD on the REC-Track, while people area will be addressed later (based on ISA-core draft)

PhilA: for the WG to consider

<Zakim> PhilA, you wanted to just check we want the vocabs on the Rec Track

<Mike_Pendleton> PhilA: to get to REC you have to prove it is being used correctly

<bhyland> PhilA: Part of being a Recommendation, is a reference implementation to provide it is being used, and used correctly.

PhilA: by publishing vocabulary on the REC-Track, it needs to be proved that the vocabulary is used

<olyerickson> Link to ISA Core Person Vocabulary http://joinup.ec.europa.eu/asset/core_person/description

PhilA: and is used in a correct manner... a note doesn't have this restriction

mhausenblas: with dcat for example we can prove it

<bhyland> mhausenblas: DCAT, Cube and ORG are all used and we can prove it.

<Mike_Pendleton> Proof comes from signature logs, etc

mhausenblas: using mailing lists, Sindice log, answer.semantic.org

<olyerickson> Link to all ISA Core Vocabularies (reference) https://joinup.ec.europa.eu/software/isa-cv/description

PhilA: you need to prove that people use *all* of it

<HadleyBeeman> This is sounding like a significant bit of research.

PhilA: every element of the specification has been implemented more than once in a right manner

cygri: the charter said that the vocabularies will be on the REC-track

<Zakim> cygri, you wanted to say that the charter says the vocabs will go REC

cygri: looking at SKOS is helpful

<Mike_Pendleton> bhyland: We are not in the business of writing vocabs; we do checklists and specify what is used

<mhausenblas> Michael: I disagree. We do do vocabs

<mhausenblas> ... see sec 2.3

bhyland: with BP, the focus is on providing checklist help people choose vocabularies, etc...
... Core vocabularies sounds out of the charter
... yesterday decision was not to give a particular recommendation on a vocabulary

<mhausenblas> PROPOSAL: The GLD WG decides to publish BP, DCAT, QB and ORG as FPWD on the REC-Track, while people area will be addressed later (based on ISA-core draft)

bhyland: except for items that are explicitly mentioned in the charter
... which includes people

<mhausenblas> PROPOSAL: The GLD WG decides to publish BP, DCAT, QB and ORG as FPWD on the REC-Track, while people area will be addressed later

cygri: it is early to decide/vote on the ISA core draft as we are fully aware of it

<bhyland> Cygri: couldn't be supportive of ISA Core Core vocab until we've had a chance to review it.

<cygri> +1

<George> + VCard?

<bhyland> +1 cygri

<BenediktKaempgen> +1

<boris> +1

<sandro> +1

<SimpsonTP_> +1

+1

<dvilasuero> +1

<gatemezi> + 1

<gatemezi> +1

<DanG> +1

<mhausenblas> PROPOSAL: The GLD WG decides to publish the first batch of FPWD by end of Feb 2012 with 9 Feb ready for WG review and final decision to be taken on 23 Feb.

mhausenblas: the proposal gives two weeks for the members to comment

<mhausenblas> + where is my brain

PROPOSAL: The GLD WG decides to publish the first batch of FPWD by end of Feb 2012 with 9 Feb ready for WG review and final decision to be taken on 23 Feb.

<SimpsonTP_> +1

<rreck1> predictable

<sandro> +1

<boris> +1

<cygri> +1

<dvilasuero> +1

+1

<Mike_Pendleton> +1

<olyerickson> Well that was my question

<mhausenblas> RESOLUTION: The GLD WG decides to publish BP, DCAT, QB and ORG as FPWD on the REC-Track, while people area will be addressed later

<mhausenblas> RESOLUTION: The GLD WG decides to publish the first batch of FPWD by end of Feb 2012 with 9 Feb ready for WG review and final decision to be taken on 23 Feb.

PhilA: I will work on putting both dcat and ORG into respec

<Mike_Pendleton> PhilA: it is a mechanical job

<olyerickson> I'm +1 to DCAT (I think very little other than what PhilA is talking about)

mhausenblas: the two weeks comment period gives us a back-up

<olyerickson> I'm frankly unclear what additional work reqd for section changes to BP

mhausenblas: in case the deadline of Feb 9 is not met

bhyland: a glossary is missed from the charter
... this should be section 1.10

olyerickson: re. provenance I started working on the respected section

<mhausenblas> Michael: I suggest to keep it simple for the FPWD - our charter doesn't require provenance

olyerickson: so far it is not clear what the requirements are

bhyland: I will work on the provenance section in the cook book

<olyerickson> +1 to keeping what we have so far

mhausenblas: this doesn't look a requirement in the chapter

s/chapter/charter/

<mhausenblas> Michael: it isn't a requirement ;)

mhausenblas: may be we'd better focus on the "must do" parts

<olyerickson> bhyland that is NOT provenance

bhyland: I already have some contents ready

olyerickson: provenance need more discussion later

<Zakim> cygri, you wanted to ask about volunteers for html churning

<olyerickson> I'm more concerned about narrowing URI construction

<Zakim> mhausenblas, you wanted to comment on what is required for the BP

<mhausenblas> +1 to cygri

<dvilasuero> +1 to cygri

cygri: it would be great if someone is willing to help with the editorial work

<Mike_Pendleton> Bernadette: Any issues yet to be raised?

<rreck_again> +q validation

mhausenblas: shall we have a meta-discussion? re. organizing, time management, etc.

<rreck_again> oops imean the topic

bhyland: now that we have the deadline set to Feb 9, we've got the motivation/target to work for

<cygri> +1 to bernadette

<HadleyBeeman> +1 to bernadette and mhausenblas's points on big bang publishing approach

<Zakim> George, you wanted to talk about why VCard is getting no GLD love

<Mike_Pendleton> George: Pulse check on VCARD

George: what people think of vcard?

<Mike_Pendleton> George: ORG seems to work with it pretty well

<Zakim> mhausenblas, you wanted to talk about the non-Editor's role in the process

George: I'm curious to know about people's opinion about using it for represnting people

<rreck_again> +q

<Mike_Pendleton> mhausenblas: need folks to review

mhausenblas: people who are not editors would help in reviewing the drafts
... they have an equivalent important role

<mhausenblas> http://www.w3.org/TR/vcard-rdf/

<gatemezi> s/represnting/representing

mhausenblas: re VCard
... there are a lot of options currently
... FOAF, VCard, the ISA vocab, etc.
... further research is needed to decide

<bhyland> mhausenblas: vCard needs a bit of research before we give de facto blessing ... if it is vCard, ok, but it deserves review in light of alternatives.

PhilA: my opinion is that VCard looks very American-specific

<olyerickson> Is this a vcard vs ISA Person discussion?

<HadleyBeeman> +1 to philA: vCard is difficult when you're in the UK (or outside the USA)

PhilA: but it is hard to define a vocabulary that fit everybody

<Mike_Pendleton> PhilA: vCard format is alien to many (except Americans)

<HadleyBeeman> s/difficult/awkward

<olyerickson> HadleyBeeman what does the 'rest of the world" use?

<Mike_Pendleton> PhilA: No one format will work for everyone

<bhyland> I'm Amercan and find vCard overly simplified.

<HadleyBeeman> olyerickson we've been talking about that this week at LinkedGov. We're shoe-horning stuff into vCard— but it is awkward. And hard to explain to non-techie users.

<Mike_Pendleton> George: thought it was an easy win; this is the pulse check we are looking for

<olyerickson> HadleyBeeman has e.g. hcard come up?

<mhausenblas> Michael: In fact, as I 've been doing some work already in this area I'd volunteer to contribute to the 'people area' once we have the FPWD of BP out of the door

<bhyland> 1?

PhilA: VCard is still one of the best available options

<csarven> +1 to PhilA: implementation > convenience

<HadleyBeeman> olyerickson: yeah, it has. (I can't remember right now what the problem was there, to be honest) I think we'll end up using vCard but relabelling the fields for UK-based users (for input only).

rreck_again: valid RDF is important and woth being added to the check list

<PhilA> s/VCard is still one of the best available options/VCard is still one of the most readily available options and is no worse than any of the others/

<olyerickson> HadleyBeeman hCard discussed here http://microformats.org/wiki/hcard could be lack of official status

rreck_again: re. blank nodes

<Mike_Pendleton> rreck: two topics: 1. When I get RDF, it doesn't always validate against 'X'; 2. Anonymous nodes - tell govt's not to use B nodes

rreck_again: I'd suggest to governments not using it

<stasinos> I like bnodes when they are right for what I am doing!

<sandro> genid uris

bhyland: more suitable point for the RDF WG

<Mike_Pendleton> ... suggest replacing B nodes with URIs

<bhyland> cygri: Commented on decissions / discussions within RDF WG for RDF 1.1

<cygri> http://dvcs.w3.org/hg/rdf/raw-file/default/rdf-concepts/index.html#section-skolemization

<olyerickson> Who is talking?

cygri: in RDF WG we have a note on how to replace blank nodes with identified ones

<csarven> Avoid where possible.

<sandro> @ olyerickson it was rreck

<olyerickson> @sandro thanks

<HadleyBeeman> olyerickson: I'm quite familiar with microformats. :) It may well be that the Google Refine code we're using to import data defaults to vCard? (We're still building a proof-of-concept; that stuff will be refined as we go)

cygri: looks like blank nodes are not "liked" in general
... re. validation there is one available from the W3C
... but it only supports RDF/XML and not Turtle
... it is worth raising the point to the RDF WG

<sandro> +1 this group can say how important an RDF validator (with Turtle, etc) mght be for GLD.

cygri: and asking for Turtle support in the validation
... talking about the validator operated b the W3C

<olyerickson> HadleyBeeman RE Refine defaulting to vCard I wouldn't be surprised. Makes me wonder what e.g. socrata defaults to when it interprets data types

s/ b / by /

<Zakim> cygri, you wanted to answer on validation and bnodes

<bhyland> cygri: Would be helpful if the GLD WG liaised with RDF WG that improved validation in .ttl would be very useful.

<Mike_Pendleton> Bernadette: I can bring this up with the RDF working group (David Wood)

<mhausenblas> Michael: charter [[People, such as elements of FOAF or vCard in RDF. This is an area for particular attention to privacy considerations.]] - reminds me on the NeoGeo approach, see http://geovocab.org/doc/neogeo.html

sandro: when each of us votes to accept a FPWD it means that to the best of our knowledge
... the working draft is correct
... so each one is encouraged to review the working draft
... and provide feedback

s/is correct/ has the WG consensus/

<bhyland> Sandro: Appearance in a FPWD means there is consensus among the GLD WG. It is an accurate representation of the published view of the GLD WG.

<olyerickson> When you vote, you are agreeing that it should be published. There can be sections that are "undecided"

<olyerickson> s/When/My understanding is that when/

<mhausenblas> Michael: Iff we resolve on 23 Feb we can go public on 28 Feb

<mhausenblas> ... which means we show up on http://w3.org/TR/

<bhyland> PhilA: W3CComm team should be advised of the upcoming publication of the batch of FPWD from the GLD WG.

<sandro> "Government Linked Data Working Group proposes specs which will destroy the Internet, unless you comment now."

sandro: after the working drafts are out, we still have to iterate and refine them towards
... the last drafts

<cygri> s/last drafts/last call drafts/

<cygri> (charter says LC in oct 2012)

PhilA: with vocabulary we can get to the last call draft faster than the rest

<olyerickson> +1 to mhausenblas "febuary" proposal. Now I must jump off call...

PhilA: to speed up the process, we use the fact that the vocabularies are used currently

<mhausenblas> Michael: thanks olyerickson and cya

<olyerickson> bye!

<cygri> thanks olyerickson!

PhilA: so we might be able to reduce the implementation face

<HadleyBeeman> Bye olyerickson

s/face/phase/

<gatemezi> @olyerickson : bye!

<Zakim> mhausenblas, you wanted to talk about people vocab contribution

mhausenblas: re. representing people which is mentioned in the charter
... I am afraid of ending up with a very "shallow" vocabulary if we care too much about being compliant with every previous candidate
... I'd be glad to contribute there

<PhilA> +1 to bhyland

<HadleyBeeman> I suspect this may be some of the conversations that are coming out of PhilA's group… re unique identifiers for people (and the privacy implications)

<Zakim> sandro, you wanted to talk about people vocab

<bhyland> mhausenblas: reviewed GLD WG charter and noted: People, such as elements of FOAF or vCard in RDF. This is an area for particular attention to privacy considerations. Michael is willing to participate after FPWD re: privacy aspects ...

sandro: "people" topic is particularly sensitive because it implied privacy concerns

<bhyland> s/Michael is willing/Michael is not willing

sandro: schema.org is strong in regards of representing people

<stasinos> If the vocab forces that fields about people are filled in, the vocab might break privacy regulations

<bhyland> [14:15] <bhyland> s/Michael is willing/Michael is *not* willing/

<mhausenblas> Michael: yes, contribute to people and thanks to sandro I now understand what the privacy part means

sandro: a lot of publishers would like to use schema.org fro representing people

<mhausenblas> Michael: I AM willing (after FPWD)

<bhyland> Sandro: Discussed schema.org's people vocab and we should give it due consideration.

<HadleyBeeman> sandro, don't Google and Bing still index other forms of mark-up? (RDFa and microformats)?

cygri: one of the privacy concerns is the use of IFP

<Zakim> PhilA, you wanted to try and answer the privacy point made by mhausenblas

<sandro> HadleyBeeman, RDFa and microformats are syntaxes --- the only vocab they plan to use is their own (I think -- that could perhaps change, too).

cygri: for example foaf:homepage

PhilA: there is an important distinction between describing person as individual animal vs. describing the identity

<mhausenblas> Michael: In fact, Google+Bing+Yahoo, parse microdata and RDFa lite

<Zakim> mhausenblas, you wanted to talk about the 'usefulness' of people vocab

<HadleyBeeman> sandro, right, but if the argument for using schema.org's vocabulary is that it is more indexible, then wouldn't it also hold for the others?

<sandro> HadleyBeeman, it would if schema.org had terms for these other uses cases, but I don't think they do (yet).

mhausenblas: people are involved in many areas for example in organization description you need a point of contact

<sandro> we COULD argue to schema.org that they should include these other terms we want.

mhausenblas: we need to know what makes sense from government perspective

<mhausenblas> http://xkcd.com/927/

<HadleyBeeman> Sandro: we could ask schema.org to include those other things… for futher conversation in the future.

sandro: we could consider trying to add the ontologies as extension to schema.org

<HadleyBeeman> +1 to bhyland for having a serious discussion and an informed opinion on what our relationship should be with schema.org for use in government.

<sandro> bhyland: We at least need a good answer for why not use schema.org, if we don't.

mhausenblas: we might invite danbri as he is involved in schema.org
... AFAIK it is still very early

<cygri> +1 to having danbri talk to this

mhausenblas: as the process of extending schema.org is not finalized yet

bhyland: in a previous webcast they showed interest in adding government related vocabs to schema.org
... and encouraged people to act early

<sandro> +1 bhyland We need to have a conversation with the schema.org folks about whether do to do these vocabs in schema.org, or whatever.

<bhyland> What is the extension mechanism for adding gov't vocabs to schema.org.

<HadleyBeeman> +1 to inviting Guha to speak to us

<mhausenblas> ACTION: bhyland to invite Vocab TF chair to talk about Schema.org extensions [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action08]

<trackbot> Created ACTION-44 - Invite Vocab TF chair to talk about Schema.org extensions [on Bernadette Hyland - due 2012-02-02].

<bhyland> ACTION: bhyland to invite RV Guha to speak at GLD WG telecon on extension mechanism for adding to schema.org [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action09]

<trackbot> Created ACTION-45 - Invite RV Guha to speak at GLD WG telecon on extension mechanism for adding to schema.org [on Bernadette Hyland - due 2012-02-02].

sandro: there is a distinction between adding it and extending the schema through an extension mechanism

<mhausenblas> Michael: Exactly, sandro, the *Extension* mechanism is not (yet) defined

sandro: I think the former is at this time better

<mhausenblas> close ACTION-44

<trackbot> ACTION-44 Invite Vocab TF chair to talk about Schema.org extensions closed

<bhyland> PhilA: Any events happening in next 6 mos that we could leverage for next F2F.

PhilA: an upcoming event... on the 21 June
... and 22nd "the digital agenda summit" in Brussel

<bhyland> ... June 21-22, 2012 in Brussells Digital Agenda Summit ... good place to meet possibly? Govies from EU in Brussels.

PhilA: plenty of government people will be in Brussel

<HadleyBeeman> lanyrd.com?

<mhausenblas> http://xkcd.com/927/

<HadleyBeeman> that's for any conference

<dvilasuero> http://epsiplatform.eu/

<mhausenblas> http://epsiplatform.eu/news

mhausenblas: a list of government-related events http://epsiplatform.eu/

<HadleyBeeman> lanyrd.com does track by keywords… open data is one, linked data is another that theytrack

Mike_Pendleton: Smetech is also interesting

<mhausenblas> Michael: Agreed, also outreach into WWW in Lyon

bhyland: I submitted a talk there

<dvilasuero> +1 to WWW

<HadleyBeeman> Also, I'm doing an open data workshop at WWW2012

<HadleyBeeman> http://www2012.wwwconference.org/

<HadleyBeeman> http://lanyrd.com/search/?context=future&q=open+data

<gatemezi> @HadleyBeeman: could you provide the link of the workshop on open data?

<HadleyBeeman> (sorry to bang on about this— but lanyrd.com is so easy to use!)

<George> thnx HadleyBeeman !

<bhyland> Mike_Pendleton asked about whether WG is giving a talk at SemTech West in June. Answer: Yes, bhyland submitted one on behalf of the WG.

<HadleyBeeman> For Gatemezi: Open Data in Practice workshop at WWW2012 (last on the page): http://www.w3.org/2012/04/tuto-track.html

mhausenblas: we can have a Google+ hangout for outreach

<mhausenblas> Michael: The nice thing is that we can record is as well, we can share screen and docs - once FPWD are out we can try this for an hour

<gatemezi> Thx bhyland!!

bhyland: we need to be more active on the egov blog as well

<HadleyBeeman> Great to see you all!

<mhausenblas> glawway out!

<BenediktKaempgen> bye, thanks.

<PhilA> sandro: Are you taking care of the minutes?

<DanG> bye

<HadleyBeeman> Bye, all. Enjoy the pub(s)!

<sandro> PhilA, uh, sure.

<PhilA> sandro: I should have said "please"

<bhyland> Suggested: Google+ Hangout once the the various FPWD's are released.

<bhyland> ping

Summary of Action Items

[NEW] ACTION: bhyland to gather terms for a Glossary section in the BP document, then link that static list to the live wiki page where these will be repeated and updated over time [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action07]
[NEW] ACTION: bhyland to invite RV Guha to speak at GLD WG telecon on extension mechanism for adding to schema.org [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action09]
[NEW] ACTION: bhyland to invite Vocab TF chair to talk about Schema.org extensions [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action08]
[NEW] ACTION: cygri to move dcat-related content from eGov wiki to GLD wiki [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action01]
[NEW] ACTION: cygri to update http://vocab.deri.ie/dcat to point to the new dcat ED [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action03]
[NEW] ACTION: cygri to update http://www.w3.org/2011/gld/wiki/Data_Cube_Vocabulary [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action06]
[NEW] ACTION: PhilA to convert editors' draft of DCAT to make use of respec [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action02]
[NEW] ACTION: PhilA to create W3C format of Org ontology specification document [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action05]
[NEW] ACTION: PhilA to reflect on SEMIC advice on vocab selection etc. to see if there is more to contribute to the BP doc [recorded in http://www.w3.org/2012/01/26-gld-minutes.html#action04]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.136 (CVS log)
$Date: 2012/01/26 19:44:08 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.136  of Date: 2011/05/12 12:01:43  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/cygri:/mhausenblas:/
Succeeded: s/otehr/other/
Succeeded: s/sai/said/
Succeeded: s/particular used/particular uses/
Succeeded: s/sandro:/george:/
Succeeded: s/?/cmusialek/
Succeeded: s/void/dublin core/
Succeeded: s/we use dcat/we use vcard/
Succeeded: s/dcat/vcat/
Succeeded: s/procuding/procuring/
FAILED: s/laptopns/laptops/
FAILED: s/know prevailed/know should have prevailed/
FAILED: s/19th/18th/
FAILED: s/outlibe/outlive/
FAILED: s/pruls/purls/
FAILED: s/rprox/prox/
FAILED: s/rpox/prox/
FAILED: s/Notes that bhyland has a business that sells PURLZ servers//
Succeeded: s/interested/interesting/
FAILED: s/olyerikson/olyerickson/
FAILED: s/quantfy/quantify/
FAILED: s/accurary/accuracy/
FAILED: s/dinner/beer/
FAILED: s/dinner/martini/
FAILED: s/atement/statement/
FAILED: s/atement/statement/
FAILED: s/atement/drinkment/
FAILED: s/vGLD/GLD/
FAILED: s/moentum/momentum/
FAILED: s/chapter/charter/
FAILED: s/represnting/representing/
FAILED: s/difficult/awkward/
FAILED: s/VCard is still one of the best available options/VCard is still one of the most readily available options and is no worse than any of the others/
FAILED: s/ b / by /
FAILED: s/is correct/ has the WG consensus/
FAILED: s/When/My understanding is that when/
Succeeded: s/ MarComm/Comm/g
FAILED: s/last drafts/last call drafts/
FAILED: s/face/phase/
FAILED: s/Michael is willing/Michael is not willing/
Found Scribe: cygri
Inferring ScribeNick: cygri
Found ScribeNick: BenediktKaempgen
Found Scribe: gatemezi
Inferring ScribeNick: gatemezi
Found Scribe: fadi
Inferring ScribeNick: fadi
Scribes: cygri, gatemezi, fadi
ScribeNicks: BenediktKaempgen, cygri, gatemezi, fadi
Default Present: galway, HadleyBeeman, fadi, PhilA, dvilasuero, cygri, mhausenblas, GofranShukair, BartvanLeeuwen, csarven, BenediktKaempgen, gatemezin, sandro, olyerickson, GeraldSteeman, boris, Washington, t_gheen, spyroskotoulas, DeirdreLee, DaveReynolds, stasinos, +1.703.201.aaaa
Present: galway HadleyBeeman fadi PhilA dvilasuero cygri mhausenblas GofranShukair BartvanLeeuwen csarven BenediktKaempgen gatemezin sandro olyerickson GeraldSteeman boris Washington t_gheen spyroskotoulas DeirdreLee DaveReynolds stasinos +1.703.201.aaaa

WARNING: No meeting title found!
You should specify the meeting title like this:
<dbooth> Meeting: Weekly Baking Club Meeting


WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Got date from IRC log name: 26 Jan 2012
Guessing minutes URL: http://www.w3.org/2012/01/26-gld-minutes.html
People with action items: bhyland cygri phila

[End of scribe.perl diagnostic output]