SV_MEETING_TITLE -- 25 Oct 2012

<Mike_Pendleton> John Erikson: clarify - it is inconsistent with Dublin Core; John's recommendation be that it is easy for implementers; use a typed literal as the choice

olyerickson: The issue is inconsistency between how Dublin Core specifies range of dcterms.

<PhilA> PROPOSED RESOLUTION: The we recommend that the value of dcterms:language be given as a typed literal, e.g. en^^xsd:lang

<Mike_Pendleton> Accepted

<Mike_Pendleton> Question from Phil: Are we knowingly making an inconsistency?

… This is incorrect according to the DC spec which gives the range as http://purl.org/dc/terms/LinguisticSystem - i.e. a resource.

DaveR: Has no problem with what is proposed but isn't sure that it is consistent with the spirit of what DC language as a resource.

<atemezin> http://lov.okfn.org/dataset/lov/details/vocabulary_dc.html

<james> see http://dublincore.org/schemas/rdfs/

<jmynarz> rapper -i rdfxml -o turtle http://purl.org/dc/terms/ > dcterms.ttl

<Mike_Pendleton> Bernadette: We can't close issue 26. We need to get it right.

<scribe> ACTION: PhilA to get clarification from Maxx Deckers on what DC meant in what they specified. [recorded in http://www.w3.org/2012/10/25-gld-minutes.html#action01]

<trackbot> Created ACTION-83 - Get clarification from Maxx Deckers on what DC meant in what they specified. [on Phil Archer - due 2012-11-01].

<Mike_Pendleton> Phil will take an action item to resolve this with Max from the Dublin Core

<jmynarz> This is how dcterms:language is used in practice: http://bit.ly/RYd4Ab

<DaveReynolds> Agree with Richard, not inconsistent with letter of spec but worth checking with expert about spirit of DC spec

<james> +1

<fadmaa> remove dcat:dataQauality, dcat:granularity

<cygri> PROPOSAL: Drop dcat:dataQuality and dcat:granularity

DCAT (continued)

1) drop dcat:dataQuality and dcat:granularity https://www.w3.org/2011/gld/track/issues/10

<olyerickson> + to nuking data quality

<atemezin> +1 to drop

<fadmaa> +1

<Mike_Pendleton> Phil: The issue is not going away, but we can close it here

<Mike_Pendleton> Phil: We need this, but this is not the place for it

<olyerickson> +0 to granularity (it's not wrong)

<cygri> RESOLUTION: Drop dcat:dataQuality and dcat:granularity

To close Issue-10, We agreed to remove but the issue doesn't go away. This isn't the place for it.

<olyerickson> +1 to resolution

2) format of distribution https://www.w3.org/2011/gld/track/issues/12

<olyerickson> "Proposal: define dcat:mediaType as a sub-property of dc:format. Publishers should use dcat:mediaType to provide a Media Type as defined by IANA. if this is not available, then publishers may use dc:format"

<PhilA> +1

<olyerickson> +1

<DaveReynolds> +1

<George> +1

<martinAlvarez> +1

<Mike_Pendleton> Thanks John!

<cygri> +1

<fadmaa> +1

Next, 3) size of distribution https://www.w3.org/2011/gld/track/issues/44

<Mike_Pendleton> Issue 12 proposal accepted

<cygri> RESOLUTION: define dcat:mediaType as a sub-property of dc:format. Publishers should use dcat:mediaType to provide a Media Type as defined by IANA. if this is not available, then publishers may use dc:format

<fadmaa> :dist dcat:size [dcat:bytes 1024; rdfs:label 1k]

<olyerickson> cygri's proposal on dist. size: "...Replace dcat:size and dcat:bytes with a single property dcat:byteSize, range is xsd: decimal, stating that the value can be approximate..."

<olyerickson> RE: dcat:byteSize I very much like this, esp. that the size is "approximate." There can be platform-specific ways to document exact size, digital signatures, etc which we won;t get into

<olyerickson> Please re-state the resolution

<cygri> RESOLUTION: define dcat:mediaType as a sub-property of dc:format (in the old DC namespace). Publishers should use dcat:mediaType to provide a Media Type as defined by IANA. if this is not available, then publishers may use dc:format

<Mike_Pendleton> PROPOSAL: replace properties dcat:size and dcat:bytes with one property dcat:byteSize

https://www.w3.org/2011/gld/track/issues/44

… byteSize can be approximate

<olyerickson> value is "guidance"

<cygri> for example: </dataset123/distribution1> dcat:byteSize 547000.

<Mike_Pendleton> Bernadette: point is that you verify you have the right file

<atemezin> +1

<cygri> +1

<martinAlvarez> +1 (easier to use)

<fadmaa> +1+1

<Mike_Pendleton> RESOLUTION: replace properties dcat:size and dcat:bytes with one property dcat:byteSize

<fadmaa> http://www.w3.org/2011/gld/wiki/Dcat_examples

Two issues re: distribution subclasses, see Distribution subclasses https://www.w3.org/2011/gld/track/issues/9 https://www.w3.org/2011/gld/track/issues/8

<fadmaa> http://www.w3.org/2011/gld/track/issues/9

<fadmaa> http://www.w3.org/2011/gld/track/issues/8

<Mike_Pendleton> Fadi has one more; related to issues 8 and 9

<Mike_Pendleton> fadmaa discussed DCAT examples http://www.w3.org/2011/gld/wiki/Dcat_examples

Some examples:

-> Direct: accessURL points to a WebService, RSS, XLS, or XML, which

offers the distribution directly.

-> Indirect: accessURL points to a REST WebService or API

documentation (how to use it, parameters, etc) | an XML zipped

<fadmaa> http://www.w3.org/2011/gld/track/issues/43

<Mike_Pendleton> bernadette: This goes to issues around ADMS if I understand correctly; Phil, is this DCAT distrubution related to ADMS?

<olyerickson> fadmaa's proposal on distribution sub-typing makes sense to me

1++ Fadi

Discussed: Does dcat:Distribution address one of the use cases of ADMS, i.e., 4 x zipped PDF files ...

Fadi: I think of DCAT describing raw data, not PDF

Cygri: this is all pretty complicated, we all 'know' what a data catalog is, but we really aren't defining it explicitely.

… thus hard to make hard & fast rules about what is "in" vs "out" of a data catalog

<cygri> http://dvcs.w3.org/hg/gld/raw-file/default/dcat/index.html#class--dataset

… everyone has different views of what can & cannot go in a data catalog using DCAT

Per the ED, "DCAT is an RDF vocabulary designed to facilitate interoperability between data catalogs published on the Web. This document defines the schema and provides examples for its use.

<atemezin> @cygri: could we *exclude* those type of *things (eg PDF) in the spec ?

… whether a zipped file can be described using DCAT, it isn't currently addressed by DCAT sufficiently

Sounds to me like we need to review DCAT & ADMS in context of common use cases

cygri: I have yet so see handling of archive files sufficiently handled by gov data catalog sites … it is a case that hasn't been fully addressed.

<Mike_Pendleton> Bernadette: we need a dedicated session on this....it goes to the heart of ADMS too.

cygri: Invoked Rufus Pollock in relation to use of DCAT for CKAN. There are use cases where there is a desire to describe associated documentation of a given dataset, or scripts that were used to create a dataset, or representative sample data from a dataset that isn't a data set per se.

… DCAT as designed currently doesn't describe supplementary documentation. We need to figure out how to do this.

<Mike_Pendleton> cygi: core issue (also addressed by Rufus P.) CKAN use cases also describe other things (e.g., scripts used to create the dataset). DCAT doesn't discuss decribing these things.

<fadmaa> http://www.w3.org/2011/gld/wiki/Dcat_examples#5._A_dataset_with_an_example_record is an example that uses dc:reference for an example which is too generic

<PhilA> Is DCAT for 3* + and ADMS for 1 and 2 star?

… This gets to the core of ADMS and DCAT, issues around interoperability of downloadable resources.

<Mike_Pendleton> cygi: if we had the answers to this; things would fall into place

<Mike_Pendleton> Bernadette: This makes the point that we need a separate session on this.

<fadmaa> +1 to a dedicated session

<Mike_Pendleton> Bernadette: This is at the core of 3, 4, and 5 star data we are trying to publish

<Mike_Pendleton> Bernadette to Fadi: any other issues that need to be addressed to move DCAT forward?

Issues 8,9 and 43 all related to bigger distribution topic that we want to follow up with in a focused DCAT WG call.

PhilA: We can move FPWD to LC even with a handful of open issues.

<Mike_Pendleton> Phil: Some of these issues may be left open and we can still publish

Propose: Move DCAT to LC

<Mike_Pendleton> Phil: you can get a new version out w/o closed issues; but we cannot't move to LC with open issues

<Mike_Pendleton> LC is last call

<Mike_Pendleton> Phil: would like a new version published with open issues

<Mike_Pendleton> fadmaa: can do this Monday

<scribe> ACTION: Fadi to update DCAT FPWD with revisions from this past 2 weeks discussion [recorded in http://www.w3.org/2012/10/25-gld-minutes.html#action02]

<trackbot> Created ACTION-84 - Update DCAT FPWD with revisions from this past 2 weeks discussion [on Fadi Maali - due 2012-11-01].

<Mike_Pendleton> bernadette: Next

cygri: Recommends that distribution issues be solved before next PWD.

<PhilA> PROPOSED RESOLUTION: That we aim to resolve to publish a new version of DCAT in 2 weeks' time. Whether that is LC or not remains open

<atemezin> +1 to solve the distribution issue before the next working draft

<cygri> +0.5

Fadi & Cygri agreed to talk over the next two weeks and edit the DCAT FPWD so we can publish another working draft within 3 weeks.

<atemezin> +1

<olyerickson> +1...I wasn't opposed to the previous proposal, and I think we will either solve this quickly or we need to rethink it

Linked Data Glossary, W3 Working Note or wiki?

<Mike_Pendleton> RESOLUTION: resolve to publish a new version of DCAT in 2 weeks' time; another draft in the next 2-3 weeks

Last week mhausenblas mentioned glossary going here http://docs.webplatform.org/wiki/Main_Page

<sandro> +1 Working Group NOTE

<Mike_Pendleton> Bernadette requesting opinion on glossary as working note or wiki

<sandro> (not strong opinion)

Question: Should it be a Working Note or wiki

<DaveReynolds> No strong opinion. Weak preference for Wiki to simplify future update.

cygri: suggested that language will evolve with current usage

<Mike_Pendleton> Richard: There will be updates after the WG terminates; language will evolve;

<Mike_Pendleton> Bernadette: I've not gotten input since June

<Mike_Pendleton> Bernadette: There will always be new things, I propose the working note

<Mike_Pendleton> Sando: there is not rule about who owns a note once published

Sandro: Procedurally, there is no reason another WG or even say eGov IG couldn't take on editing the Note.

<Mike_Pendleton> Sandro: Semantic Web Interest Group could take over the Note (not WG Note)

<PhilA> I prefer a Note personally

<Mike_Pendleton> Bernadette: Objections to a Note?

No objection to the Linked Data Glossary being published as a Note.

<DaveReynolds> No objection, still mild preference for wiki.

<atemezin> No objection

<Mike_Pendleton> Bernadette: Thanks for your time; send projects announcements around publishing LOD by today for tomorrow; Bernadette leaves for France Monday.

<Mike_Pendleton> Adjourned

<martinAlvarez> thanks, bye

<Mike_Pendleton> Not meeting next week; next meeting Nov 8.

thanks all & safe travels.

- DRAFT -

SV_MEETING_TITLE

25 Oct 2012

Attendees

Contents

DCAT (continued)

Linked Data Glossary, W3 Working Note or wiki?

Summary of Action Items

Scribe.perl diagnostic output