RDF Web Apps WG telecon minutes for 2011-07-21

Thanks to Stéphane for scribing! The RDF Web Apps WG telecon minutes for
July 21st 2011 are now available here:

http://www.w3.org/2010/02/rdfa/meetings/2011-07-21

If you would like to read minutes from this or previous meetings, the
public record of all RDF Web Apps WG telecons is available here:

http://www.w3.org/2010/02/rdfa/wiki/Meetings

Full text log follows:

----------------------------------------------------------------------------

RDF Web Applications Working Group Teleconference

Minutes of 21 July 2011

Agenda
    http://lists.w3.org/Archives/Public/public-rdfa-wg/2011Jul/0042.html
Seen
    Gregg Kellogg, Henri Bergius, Knud Möller, Manu Sporny,
    Niklas Lindström, Sebastian Germesin, Shane McCarron,
    Steven Pemberton, Stéphane Corlosquet, Ted Thibodeau, Thomas Steiner
Guests
    Stéphane Corlosquet, Henri Bergius, Niklas Lindström
Chair
    Manu Sporny
Scribe
    Gregg Kellogg, Stéphane Corlosquet

Topics
1. structured-data.org update
2. Microdata/RDF conversion
3. Alternate @profile proposals

http://www.w3.org/2005/Incubator/lld/Scribing.html

(Scribe set to Stéphane Corlosquet)

Steven Pemberton: agenda?
Steven Pemberton: http://www.doodle.com/87nkyax5q5bma698#table
Manu Sporny: any change to the agenda?

Manu Sporny: everyone ok for having a call on Aug 4th?

Henri Bergius: I *may* be able to join Aug 4th, but not sure
Manu Sporny: Aug 4th is the only telecon which might be cancelled

1. structured-data.org update

http://linter.structured-data.org/

Manu Sporny: I owe you guys a front-page...
Gregg Kellogg: Linter - http://linter.structured-data.org/
Stéphane Corlosquet: hope to launch early next week

Gregg Kellogg: mimic google rich snippets. schema.org is todo, plus 
other schemas like FOAF, SIOC

Manu Sporny: fantastic to work on that, should be beneficial for lots of 
people

Manu Sporny: we want to get Microdata and microformats folks on board, 
ensure that they have edit/mod privileges to the website, but I have not 
been successful so far (very busy and missed them on IRC)

Thomas Steiner: ok w/ me
Henri Bergius: looks good
Manu Sporny: any concerns with including other groups? and launch early 
next week anyways?

Gregg Kellogg: we don't have microformats now (no parser available). if 
anyone has pointers to parser (XSLT maybe)

Manu Sporny: not much available

2. Microdata/RDF conversion

Gregg Kellogg: Topic with Hixie comment: 
https://plus.google.com/u/0/115203359751471044302/posts/92VKitpppB4
Manu Sporny: the RDF conversion steps could be removed the RDF 
processing rules from the microdata specs

Henri Bergius: if the RDF processing were to be removed from microdata, 
we would end up with 2 separate formats - that might be bad.

Henri Bergius: then would people bother with RDFa at all for SEO?

Henri Bergius: the advantage for microdata not being RDF would of course 
be clarity: you do schema.org microdata only for SEO, RDFa for linked data
Shane McCarron: I do agree that if microdata stops having an RDF mapping 
it would be a good thing
Niklas Lindström: could there be one monolithic format for schema.org as 
a subset of RDFa 1.1?

Manu Sporny: if the RDF steps are removed from microdata, this might be 
enough to avoid forming the W3C TAG RDFa/Microdata TF.

Manu Sporny: but then, people might think that there is no reason to 
implement RDFa for SEO because it seems more complex (but has roughly 
the same level of complexity as Microdata for the schema.org use cases). 
However, there are other important use cases for RDFa - universal data 
model for the web, publishing data via your website in a way that is 
compatible with Linked Data and with larger systems that integrate data. 
There are good reasons for picking each technology - Microformats if you 
want to publish basic information or dip your toes into the lower-case 
semantic web. Microdata if you want something a bit more advanced than 
Microformats and something that will work with schema.org w/ a simple 
API. RDFa if you want something that is designed for Linked Data, the 
upper-case Semantic Web, allows you to create your own Web vocabularies, 
allows you to do vocabulary mixing easily, works with big data.

3. Alternate @profile proposals

Niklas Lindström: 
http://lists.w3.org/Archives/Public/public-rdfa-wg/2011Jul/0048.html
Niklas Lindström: profiles are complex. primary suggestion (item #2) is 
to move mapping of terms from a syntax level to a semantic level

Niklas Lindström: use the vocab attr only. and instead describe 
vocabularies which import other terms

Niklas Lindström: like schema.org and ogp, they define every single term 
that they think people will need

Niklas Lindström: it's vocabulary design, as opposed to embeding vocab 
definitions in RDFa

Niklas Lindström: markup would look more like microdata

Manu Sporny: at the time, we didn't see problems or any danger with profiles

Thomas Steiner: looking at your gist, lindstream: 
https://gist.github.com/1092350
Manu Sporny: in your email, you provide a mechanism to define terms. 
solution #2 is interesting for the RDF people (the rest don't care 
because it's close to microdata syntax)

Thomas Steiner: is this something like inlined grddl?
Niklas Lindström: solution #2 specifies a new emerging pattern for the 
semantic web (broader than RDFa)

Niklas Lindström: great value in linking terms from your vocabs to other 
vocabs

Niklas Lindström: defining this mapping vocabulary is for the benefit of 
the general RDF community

Gregg Kellogg: http://www.w3.org/TR/rdf-mt/#RDFSRules
Gregg Kellogg: some of the RDFS entailment rules would accomplish the 
same thing. we could generate a subset vocab for class and properties.

Niklas Lindström: good point, will include that in my next email. 
chicken and egg: you produce triples intended to be remapped, which 
won't be remapped until other use the same mechanism

Manu Sporny: there were a few concerns about removing profiles

Manu Sporny: people who want to keep profiles are ShaneM and the ePub folks

Niklas Lindström: I looked at the ePub spec, it didn't seem too tricky 
to tweak their work and avoid profiles, and use vocab instead. I will 
look at that.

Manu Sporny: negative: the initial RDF graph you get from the RDFa is 
not as complete as the one you would using RDFS

Niklas Lindström: from asking people to reference multiple vocabs in 
your page, we ask them to reference these other vocabs via an 
intermediary vocabulary (defined according to the mapping mecahnism)

Manu Sporny: what are you thoughts Shane?

Shane McCarron: I've sent an email about my concerns. main point: there 
has to be a way for authors what they mean to say when using a term

Manu Sporny: 1. Why is waiting for all @profile documents to load and 
then proceeding a bad thing? What makes it technically challenging to 
implement in a browser?
Manu Sporny: 2. Is there an announcement mechanism for RDFa Core 1.1? We 
removed @version and pseudo-replaced it with @profile. Do we need to 
re-introduce @version? If we don't do this, an RDFa 2.0 processor may 
accidentally corrupt the intent of an RDFa 1.1 document.
Gregg Kellogg: re relying on another mapping document. the original 
reason for this discussion is the need to load the profile during the 
parsing of the HTML document.

Gregg Kellogg: if you have the need to operate on the inferred triples, 
you still have a (weaker) dependency on the vocabulary

Ted Thibodeau: there has to be external dependencies no matter what, 
which have to be dereferenced later

Niklas Lindström: main point: these dependencies are on the semantic 
level, not at the parsing level. might break when using the follow your 
nose, but we always have triple we can operate on

Gregg Kellogg: this mechanism processing rules and remove dependencies 
during that processing

Manu Sporny: right, it does not remove the need to do follow your nose, 
but it puts in a the background, and people who want to use it can just 
do it

Shane McCarron: q+ to ask about follow your nose
Zakim IRC Bot: ShaneM, you wanted to ask about follow your nose
Gregg Kellogg: if we do use something like RDFS, the original statement 
does not get erased, it gets added

Manu Sporny: how does the RDFa API handle the RDFS rules? We'll have to 
discuss that in the future.

Manu Sporny: STRAW POLL: Drop @profile as it is defined now, and replace 
it with Niklas' @vocab proposal (#2 item in his e-mail - proxy vocabularies)
Manu Sporny: +1
Gregg Kellogg: +1
Sebastian Germesin: +1
Steven Pemberton: +0
Niklas Lindström: +1
Henri Bergius: +1
Stéphane Corlosquet: +1
Thomas Steiner: 0 (no opinion really)
Shane McCarron: +0
Knud Möller: +1
Ted Thibodeau: +0 insufficiently considered for me to decide
Manu Sporny: looks like a consensus, but we should ask other groups like 
IPTC, Google and Facebook

Niklas Lindström: we should consider the proposal #1

Manu Sporny: Nathan said it was not possible in RDFa, but I can't recall 
his reasoning....

Manu Sporny: prefix="title: http://purl.org/dc/terms/title dc: 
http://purl.org/dc/terms/"
Manu Sporny: property="title" or property="dc:title"
Shane McCarron: prefix=":title http://purl.org/dc/terms/title dc: 
http://purl.org/dc/terms/"
Niklas Lindström: prefix=":title http://purl.org/dc/terms/title dc: 
http://purl.org/dc/terms/"
Manu Sporny: can't we remove the : in the prefix list? Let's kick this 
discussion onto the mailing list.

--manu

-- 
Manu Sporny (skype: msporny, twitter: manusporny)
Founder/CEO - Digital Bazaar, Inc.
blog: Uber Comparison of RDFa, Microformats and Microdata
http://manu.sporny.org/2011/uber-comparison-rdfa-md-uf/

Received on Wednesday, 27 July 2011 03:25:27 UTC