[participants] [actions] [decisions] [minutes] [irc logs] [venue] [objectives] [agenda] [reading]

25th-26th February 2002 Meeting of the W3C RDFCore WG

Technical Plenary and WG Meeting Event Hotel Royal Casino - Cannes Mandelieu FRANCE

Participants

Dave Beckett, ILRT
Jeremy Carroll, (hplabs)
Dan Connoly, W3C
Mike Dean, BBN
Jos De Roo, AGFA
Martyn Horner, Profium Sarl
Graham Klyne, Baltimore Technologies
Brian McBride, Hewlett Packard (co-chair)
Eric Miller, W3C
Patrick Stickler, Nokia

Telecon Participants

Telecon partipants attended the two telecon's at 4pm each day.

Pat Hayes, IHMC
Frank Manola, MITRE
Aaron Swartz, HWG

Regrets

Dan Brickley, University of Bristol (co-chair)
Ron Daniel, Interwoven
Jan Grant, ILRT, University of Bristol
Sergey Melnik, Stanford University
Steve Petschulat, IBM

Absent

Frank Boumphrey
Bill dehOra, InterX
Rael Dornfest
R. V. Guha
Yoshiyuki Kitahara
Michael Kopchenov
KWON, Hyung-Jin
Ora Lassila
Satoshi Nakamura
Pierre G. Richard

Objectives

to resolve as many outstanding issues as possible
to plan for getting to last call and beyond

Agenda

irc channel: irc.openprojects.net #rdfcore

Monday, 25th Feb 2002 (all day)

         0830 Welcome, Introductions, meeting objectives, agenda review
         0845 The current state of play (bwm)
         0900 Issues session 1
                rdfms-not-id-and-resource-attr
                rdfms-nested-bagIDs
                rdfms-rdf-names-use
         1030 Coffee
         1045 short presentation
         1100 issues session 2
                rdfms-replace-value
                rdfms-editorial
                rdfms-xml-base
                rdfms-fragments
         1230 lunch
         1330 issues session 3
              preparing for I18N meeting
              Datatypes
         1515 coffee
         1545 short presentation
         1600 telecon
                review of the day
                rdfms-identity-of-statements
                rdfms-fragments
                datatypes         
         1730 plans for next day
         1745 close

              Dinner
              Homework - read Pat's latest datatyping proposal

Tuesday, 26th Feb 2002 (all day)

         0900 Split Session
                  Meeting with I18N irc #rdfcore-i18n
                  breakout group - test cases
                     DaveB's detailed test case for rdfms-nested-bagID
                     Jeremy's XML base test cases (+ve test cases for error2 and error3)
                     generate test cases for illegal names use e.g. <rdf:foo>
                     Discuss RDF 2396 - rdf:ID="foo" is equiv to rdf:about=".#foo" 
         1030 Coffee
         1045 issues session 4
                rdf-charmod-literals
                rdf-charmod-uris
                rdfms-xmllang
                rdfms-literal-is-xml-structure
                rdfms-xml-literal-namespaces 
                Test cases from morning session               
                RFC 2396 - rdf:ID="foo" is equiv to rdf:about=".#foo" 
                rdfms-assertion
                rdfms-not-id-and-resource-attr
         1230 lunch
         1330 Issues session 5
                rdfms-assertion
                rdfms-not-id-and-resource-attr
                datatyping
         1515 short presentation - Patrick
         1530 coffee 
         1600 Telecon
              Getting to Rec
              Goals and Next Steps
         1800 close

Actions


2002-02-25#1	jjc	Formalize test case for <eg:prop rdf:ID="issue" eg:p1="foo" />
2002-02-25#2	JanG	Add test case for <eg:prop rdf:ID="issue" eg:p1="foo" /> to test cases draft
2002-02-25#3	JanG	Review test cases in case changed by <eg:prop rdf:ID="issue" eg:p1="foo" /> decision
2002-02-25#4	DaveB	rdfms-nested-bagID: update syntax WD to reflect decision
2002-02-25#5	bwm	review test case for rdfms-nested-bagID
2002-02-25#6	jjc	generate more test cases for rdfms-nested-bagID
2002-02-25#7	DaveB	Emphasise in syntax WD that rdf:li is not allowed as an attribute
2002-02-25#8	DaveB	Update syntax doc to reflect
2002-02-25#9	JanG	Ensure test cases for rdf:foo added to test cases WD
2002-02-25#10	DaveB	Update syntax doc to reflect xml:base same doc ref decision
2002-02-25#11	JanG	Add test case for xml:base same doc reference to test cases WD
2002-02-25#12	DanC	Contact RFC 2396 editors to confirm that RDF will use a different algorithm for xml:base
2002-02-25#13	DaveB	Update syntax WD to reflect xml:base="http://example.org"> decision
2002-02-25#14	JanG	update test cases to reflect xml:base="http://example.org"> decision
2002-02-25#15	DanC	Confirm with RDF 2396 that xml:base="http://example.org"> decision is correct
2002-02-25#16	jjc	replace xml:base error cases 1 and 2 with +ve test cases
2002-02-25#17	EricM	Ensure primer has appropriate description of use of rdf:value
2002-02-25#18	bwm	ensure model theory updated to reflect semantics of rdf:value
2002-02-25#19	Graham	Draft text for the primer on use of frag id's with appropriate warnings re mime types
2002-02-25#20	DanC	Highlight frag id semantics issue with the TAG
2002-02-25#21	EricM	Ensure M&S has an erratum to say the M&S will be superceded by new documents
2002-02-26#1	bwm	Add to postponed issues list, URI's to represent languages in literals
2002-02-26#2	DanC	Create test case for daml:collection behaves like Literal
2002-02-26#3	DaveB	Update syntax doc to reflect decsision about structure of literals
2002-02-26#4	bwm	Contact Pat re effects of literals decision on model theory
2002-02-26#5	jjc	Produce proposal for use of canonicalization for parseType literals
2002-02-26#6	bwm	send draft text for #rdfms-assertion issue to TimBL for comment
2002-02-26#7	DaveB	Update test cases to reflect not-id-and-resource decision
2002-02-26#8	PatrickS	Create XML examples for datatypes doc
2002-02-26#9	FrankM	Primer WD ready for review by 8th Mar
2002-02-26#10	bwm	Organise f2f around Sardinia timescales in Bristol
2002-02-26#11	bwm	Consider arrangements for RDF 2 workshop
2002-02-26#12	DaveB	Propose n-triples changes to represent the new form of rdf literals.

Decisions

d-2002-02-25-1

Issue: rdfms-not-id-and-resource-attr

  <rdf:RDF>
    <rdf:Description>
    <eg:prop rdf:ID="issue" eg:p1="foo" />
    </rdf:Description>
  </rdf:RDF>

should generate

  _:a <eg:prop> _: b.
  _:b <eg:p1> "foo" .
  <#issue> <rdf:type> <rdf:Statement> .
  <#issue> <rdf:subject> _:a.
  <#issue> <rdf:predicate> <eg:prop>.
  <#issue> <rdf:object> _:b.

This is a change to some readings of M&S and is made on the on the grounds that M&S is confusing and this decision adopts a consistent interpretation of rdf:ID on property elements. This issue is now closed.

d-2002-02-25-2

Issue: rdfms-nested-bagID

The WG decided that:

A bagID reifies the property attributes on the same element as the bagid, the type node and statements immediately arising from property elements that are immediate children of the element containing the bagId. In particular a property element whose statement is part of the bag, which has property attributes, those statements are not part of the bag.

Specifically:

  <rdf:Description about="a" bagID="bag1"> 
    <some:prop rdf:ID="st1"> 
      <rdf:Description about="b" bagID="bag2"> 
        <some:otherProp rdf:ID="st2"> A literal </some:otherProp> 
      </rdf:Description> 
    </some:prop> 
  </rdf:Description>

generates two bags. Bag1 contains st1 only. Bag2 contains st2 only. This issue is closed.

d-2002-02-25-3

Issue: rdfms-rdf-names-use

The WG reaffirmed its decision not to restrict names in the RDF namespaces which are not syntactic. The WG decided that an RDF processor SHOULD emit a warning when encountering names in the RDF namespace which are not defined, but should otherwise behave normally.

Specifically:

  <rdf:Description>
    <rdf:foo>foo</rdf:foo>
  </rdf:Description>

is equivalent to:

  _:a <rdf:foo> "foo" .

This issue is closed.

d-2002-02-25-4

Issue: rdfms-xml-base

RFC 2396 states that self document references, such as rdf:about="", are not relative URI"s are thus not subject to being converted to an absolute URI using xml:base. It was also noted in section 4.2 of RFC 2396 it states:

However, if the URI reference occurs in a context that is always intended to result in a new request, as in the case of HTML's FORM element, then an empty URI reference represents the base URI of the current document and should be replaced by that URI when transformed into a request.

It can be argued that this case should cover RDF's use of URI's.

The WG decided that RDF will convert such references to absolute URI's and will take in scope xml:base attributes into account in such conversions.

Specifically:

  <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
           xmlns:eg="http://example.org/"
           xml:base="http://example.org/dir/file"> 
    <eg:type rdf:about="" />
  </rdf:RDF>

is equivalent to:

  <http://example.org/dir/file> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://example.org/type> .

d-2002-02-25-5

Issue: rdfms-xml-base

The algorithm in RFC 2396 for creating an absolute URI is flawed in the case where the base consists of a scheme, such as http:, followed by a domain name with no terminating '/'.

Specifically, the WG decided that:

  <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
              xmlns:eg="http://example.org/" 
              xml:base="http://example.org"> 
    <eg:type rdf:about="relfile" /> 
  </rdf:RDF>

is equivalent to:

  <http://example.org/relfile> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://example.org/type> .

d-2002-02-25-6

Issue: rdfms-xml-base

M∓S states that it is not legal to have two rdf:ID attributes in a document with the same value. The WG decided that any scope xml:base attributes should be taken into account in this test.

Specifically, the WG decided that

  <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
              xmlns:eg="http://example.org/" > 
    <rdf:Description xml:base="http://example.org/dir/file"
                        rdf:ID="frag" eg:value="v" /> 
    <rdf:Description rdf:ID="frag" eg:value="v" /> 
  </rdf:RDF>

is not an error

All outstanding questions on this issue have been resolved. It will be closed when all test cases have been approved.

d-2002-02-25-7

Issue: rdfms-replace-value

The WG resolved:

that rdf:value is a property defined in the rdf namespace
the model theory must state that rdf:value is a property
that no other model theory semantics is defined specificaly for it
This issue is closed

d-2002-02-25-8

The WG resolved that the Model and Syntax recommendation would be replaced by the following documents:

a primer
an RDF/XML syntax specification
a model theory specification
a test cases specification

d-2002-02-25-9

Issue: rdfms-editorial

The WG resolved, given decision d-2002-02-25-8, the editorial issues with M&S are now not relevant to the current document set and this issue be closed.

d-2002-02-25-10

Issue: rdfms-fragments

The WG resolved that RDF uses URI's with fragment ID's to identify resources. This issue is now closed.

d-2002-02-26-1

Issue: rdfms-xmllang
Issue: rdfms-literal-is-xml-structure

The WG resolved that a literal consists of three components:

A representation of the parseType, which is a single bit
A language indicator which is a string as defined in XML.
A fully normalized UNICODE string.

Further, the WG resolved that these two issues are closed.

d-2002-02-26-2

Issue: rdfms-not-id-and-resource-attr

The WG resolved that:

  <rdf:Description>
    <foo:bar rdf:ID="foo" rdf:resource="bar"/>
  </rdf:Description>

is legal. This issue is now closed.

d-2002-02-26-3

Issue: rdfs-xml-schema-datatypes

The WG resolved that:

Pat's proposal simple datatypes 23-02-2002 is the basis for our datatypes solution
We will only make changes to fix specific problems, with test cases. A problem is something the WG agrees is problem.

Further, the WG agreed that the following issues with the datatyping proposal were problems:

The lack of XML examples (Patrick Stickler to provide)
Jeremy's entailment is a problem
Octal is not an XML Schema datatype
XML Schema datatypes do not support different language representations.

d-2002-02-26-4

The WG made the following decision with respect to the documents under development:

The primer, syntax spec, model theory, schema spec and test cases documents are all recommendation track.
Datatypes will be included in the schema document.
The WG will aim for last call 1st May 2002.
Patrick Sticker will assist with editing the Schema document.

d-2002-02-26-5

The WG agreed to continue until six months after the documents become recommendations, though not necessarily meeting regularly.

d-2002-02-26-6

The WG decided to hold a face to face meeting in June to consider last call comments.

Minutes

Notes from I18N/RDFCore Meeting

I18N recommend that literals (strings) in the RDF graph be fully normalised UNICODE and should not start with a combining character.
I18N suggests that IRI's be used to name resources in the graph and that comparison of IRI's behaves as if they are UNICODE normalised, but not does require that such normalization is performed.
I18N agree that RDFCore requires a transitive string comparison algorithm and requests that the specs do not mislead application developers into thinking they are not permitted to implement a more flexible string matching algorithm, e.g. on queries.
I18N note that the strings defining languages occasionally change and suggests that RDFCore may choose to use URI's to name languages. RDFCore agree to consider.
I18N found the proposed solution of literals being a pair of a string and a language tag acceptable.
I18N agree that n-triples is an internal tool for the WG and developers and is not subject to the same internationization concerns of more public syntaxes. I18N request that the specs make this limited role for n-triples clear.
There was some dicussion of RDFCore concerns that I18N specs would not be final and of a lack of implementations delaying completion of RDFCore.

IRC Logs and other Notes

rdfcore/i18n breakout session

Monday, 25th February 2002

Tuesday, 26th February 2002

Monday Morning

Monday Afternoon

Tuesday Afternoon

Venue

Royal Hotel Casino, Cannes-Mandelieu, France

605 avenue du G诩ral de Gaulle
06210 Mandelieu la Napoule
T誮 : +33 (0)4 92 97 70 00
Fax : +33 (0)4 93 49 51 50

Reading

None required.

Brian McBride <brian_mcbride@hp.com> $Id: Overview.html,v 1.21 2002/05/24 09:14:57 bmcbride Exp $