See also: IRC log
<ivan> minutes of last meeting
<mhausenblas> +1
RESOLUTION: accept http://www.w3.org/2011/12/14-swcg-minutes.html as minutes of last week
<ivan> 2012 jan 25
next script: tom or david
regrets for next next meeting from me and Bern.
ivan: congrats
... RDFa, SPARQL, ...
... Provenance
... prov-o and prov-primer FPWD
<Luc> thanks, Ivan
ivan: RDFa lite. alas, no Manu today.
... you probably heard about the schema.org discussions, happy ending is
this simple subset of RDFa
... a few attributes, and some advice to publishers, to stick to these
simple attributes for simple things.
... cannot set datatypes or do complex structures
... for simple things it's very simple
... similar complexity to microdata
... schema.org has publically said they will encourage their users to
use RDFa lite
... will encourage to the same degree as microdata
davidwood: Alas, that's not actually reflected in their documentation yet.
ivan: Goal is for RDFa lite to be its own Recommendation. For people authoring RDFa, it should stand on its own.
<Ashok_Malhotra> Jeni Tennison ran a task force on RDFa and Microdata
<Guus> about RDFa & schema.org: DanBi is very optimistic, but he suggests not to shout about it yet
<Luc> Provenance Data Model (prov-dm): http://www.w3.org/TR/prov-dm/
http://www.w3.org/TR/prov-aq/ http://www.w3.org/TR/prov-primer/
<Luc> Provenance Ontology (prov-o): http://www.w3.org/TR/prov-o/
sandro: prov-o is a mapping of prov data model to OWL 2
<Luc> Provenance Primer (prov-primer): http://www.w3.org/TR/prov-primer/
<Luc> Provenance Access and Query (prov-aq): http://www.w3.org/TR/prov-aq/
luc: Once you have obtained a resource, aq
tells you how to find out about its provenance.
... including SPARQL querying
ivan: On prov-aq it raises some concern,
because there has been a lot of discussion of HTTP protocols, SPARQL
REST, Linked Data Platform
... have you checked/seen/sync'd about all that?
... eg in LDP charter, ways to get info about datasets through HTTP.
Luc: I'm aware of HTTP protocol for SPARQL.
Not sure about other points.
... it would be useful to get feedback from other WGs
... to see how what we're proposing integrates with other systems
... any specific people
<LeeF> Sandro is at the nexus of all of this :-D
sandro: I'll probably review at least while I'm en route to Prov F2F
lee: I'll ask the SPARQL group for volunteers to review http://www.w3.org/TR/prov-aq/
<LeeF> "503 Service Unavailable"
sandro: rdf wg too
<LeeF> better now
luc: not the right time yet. we haven't settled our notion of "accounts", which is likely to related to named graphs
<LeeF> yup
Prov F2F in Amsterdam Feb 2/3.
ivan: I heard Linked Data Patterns workshop was lively and good
<davidwood> Linked Enterprise Data Patterns Workshop: http://www.w3.org/2011/09/LinkedData/
<davidwood> Workshop report: http://www.w3.org/2011/09/LinkedData/Report
ivan: advance notice went to AC that we are working on this.
<ivan> announcement to the AC
<davidwood> TimBL also attended both days
<ivan> very initial charter
ivan: I tried to summarize from the outside what I saw as a possible initial charter
davidwood: Cochairs were Ashok, Arnaud, and
myself, with lots of help from EricP
... Martin Nally, CTO of IBM Rational
david: Brad Allen from Elsevier
... enterprise participation
... theory was that as enterprises began to use Linked Data techniques
internally, to solve problems,
... they would fall into same problems/patterns
... 1. data integration
... integration of enterprise software, never meant to work together,
but with APIs
... lingua franca without enterprise service bus
... this is what IBM Rational is doing
... others are doing just data integration.
... workshop reports lists some of the useful patterns
... see position papers for the rest
<davidwood> Workshop papers: http://www.w3.org/2011/09/LinkedData/Papers
<LeeF> auth/auth
davidwood: Eg, how do people do
authentication and authorization?
... people are still thinking in their enterprise mentality, not the Web
way
... clear exceptions to that, but avg enterprise is mired in older
solutions
... once we got to thinking Web way, we got into some patterns about how
to update data, separate data from metadata, patch information
... surprosing interest in defining exactly what Linked Data means in an
enterprise context
... defining a Minimal Linked Data service
Ashok_Malhotra: What was really wonderful was that we had people there USING linked data to do REAL and complicated things
<davidwood> +1 to Ashok
Ashok_Malhotra: and they were running into
some difficulties
... the IBM folks had come up with some ideas about extending REST,
doing different ATOM-linke things with RDF
<davidwood> IBM also said that DB2 will support RDF "sometime in 2012"
Ashok_Malhotra: the feeling was that we ought to be able to look at these patterns, these people's experiences, then codify them. Write them down, to help other people
sandro: It was very exciting to see this RDF+REST as universal API theory finally manifest as real products
<Zakim> mhausenblas, you wanted to ask re time line and chairs
bern: this should be a keynote at SemTech SF for Martin Nally
<mhausenblas> Michael: Two remarks and questions re the upcoming LD WG - 1. questions re timeline and chairs, 2. liaison/alignment with http://www.w3.org/community/networked-data/
<bhyland> Correction: I suggested to Tony that there be a keynote at SemTech West by the three co-chairs from the Linking Enterprise Data Patterns workshop, namely Ashok Malhotra, Arnaud Le Hors (IBM) and Dave Wood.
ivan: prelim timeline is workshop report to come out this week
<bhyland> Shortly after I suggested this to Tony, I heard Dave & Ashok spoke and Martin Nally has been approached for the keynote
<bhyland> by Tony presumably.
ivan: Ian Jacobs did an interview with
Arnaud, also for publication this week
... when & how consensus on charter?
... possibly to AC by end of month... maybe.... voting takes a month.
... Could potentially start in March.
sandro: Member submission expected from IBM.
ashok: It requires some legal agreements, holding us up.
sandro: So I think Ivan's timeline may be somewhat optimistic given that.
<davidwood> Networked Data Community Group: http://www.w3.org/community/networked-data/
ivan: This community group is interesting.
... Ideally, my preference would be this CG folks into WG, if the WG
materializes
mhausenblas: Makes sense.
ivan: The draft charter should be member
visible or public soon
... So this is a time the CG can give us the comments, to make that
synergy easier
... on the charter itself, based on my reading of some of the emails, so
I'd like feedback about whether this really reflects what was discussed
there. feedback, please!
... comments in email are welcome too, of course
Ashok_Malhotra: I have a couple of comments
<Ashok_Malhotra> "provide a standard environment"
Ashok_Malhotra: I think it's a reasonable
charter. On thing jumped out at me: "provide a standard environment"
... it sounds like we're building a development envionrment
<mhausenblas> Michael: One quick comment re the charter ... copy & paste error in section 10: "This charter for the Government Linked Data Working Group ..."
Ashok_Malhotra: and we're not. we're just
codifying some patterns.
... so "standard environment" is slightly misleading.
... and as David pointed out, one of the big problems we're going after
is Application Integration, and that's not here.
ivan: Can one of you guys suggest, soon,
some language to help make that clear?
... I learned today that the software integration part is not clear in
the charter, and should be made clearer.
... I'm not sure how to exactly formulate that.
<LeeF> This is my big concern as well!!
ivan: one more thing: I don't want this to crowd out the other Semantic Web work, or to make people concerned that it might
<Zakim> mhausenblas, you wanted to add two comments re the content
<mhausenblas> Michael: re "Define a mechanism for advertizing SPARQL endpoints" ... 1. new HTTP headers are costly, almost impossible to *deploy on the Web*. 2. re SPARQL SD and VoID
mhausenblas: I don't see a reason to define a new vocab here...
<mhausenblas> Michael: http://www.w3.org/2009/sparql/wiki/CommentResponse:MH-1
sandro: mhausenblas let's you and I take this off line and see if we can make this service description bit happen within SPARQL.
<mhausenblas> Michael: I agree and thanks for this, Sandro
bhyland: This WG is about the use of Linked
Data in Enterprise, and with Patterns.
... I associate with "Linked Data" a certain set of stuff, and we don't
want to throw all this at them.
... GLD is very prescriptive, esp on Linked Open Data.
... needs to be narrowed down from all of "Linked Data". comm group also
very broard.
... Charter needs some iteration to get the focus clear. Enterprise
product managers to be involved.
<LeeF> FWIW, I thought that at the workshop there was a great deal of non-enterprise linked data stuff (in addition to some enterprise linked data stuff)
<Zakim> Luc, you wanted to ask if provenance should be mentioned
Luc: Was there any mention of provenance?
... Prov WG wont standardize everything; maybe this group needs to do
some for Prov with linked data.
davidwood: There was lots of talk, the desire for Prov. Revelytix talk covered it a bit. It would be interesting, but how it fits into the charter isn't clear to me.
luc: I don't have concrete ideas at this stage, but it would be interesting.
ivan: One of the threads of this charter ... (lost)
luc: rev doing prov-o spec
http://www.w3.org/2001/sw/wiki/REST
<bhyland> Sandro: Balancing technical ideas for this proposed group and W3C role. RDF + REST as "universal API", a new version of a distributed object system (without baggage).
sandro: That's my intuition for the kind of thing the group should be producing.
ivan: Sandro and I will work on this, going
forward.
... When we get to member feedback, we will need a good address list of
people we should be sure know about this work,.
<davidwood> I just mailed the LD WG mailing list noting that we need language for the inclusion of software/application integration: http://lists.w3.org/Archives/Public/public-ldp/2012Jan/0000.html
ivan: If you know of folks who should be involved, be sure to let us know, please.
<bhyland> What I really like about this proposed group is that it commercial companies solving real complex problems with RDF and RESTful approach. Very cool. Congrats to all that attended & are moving this forward :-)
ivan: There is a mailing list, for comments on the charter: public-ldp@w3.org
<bhyland> bye all.
mhausenblas, email for now please.
<mhausenblas> OK