RDFcore 2003-03-28 draft minutes

Attached, draft of minutes from last week. Sorry for the lag.

Forwarded message 1

  • From: Dan Brickley <danbri@fireball.danbri.org>
  • Date: Fri, 04 Apr 2003 14:43:12 +0100
  • Subject: RDFcore 2003-03-28 draft minutes
  • To: danbri@w3.org
  • Message-Id: <E191RTM-0001SR-00@fireball>
RDFCore telecon 2003-03-28 (2 hrs)


Summary of NEW actions:

ACTION: danbri review OWL Reference when published as Last Call
ACTION: Brian review OWL Reference when published as Last Call
ACTION: JanG review OWL Semantics when published as Last Call
ACTION: Brian to review CC/PP spec for next friday
ACTION: brian to put pointers on web page to Shadow TR
ACTION: brian to update shadow TR for Semantics
ACTION: Dave Beckett to get edits for Syntax from XML Schema review
ACTION: Jeremy to update Concepts in light of XML Schema LC review
ACTION: Frank to check for editorial actions on Primer from XML Schema LC reviewACTION: DanBri to check for editorial actions on Schema from XML Schema LC review
ACTION: gk send rseponse on macgregor-02
ACTION: jjc send rseponse on danc-01
ACTION: dave to check on whether status of 'notes' should be informative vs normative
ACTION: jeremy to add anchors to Concepts for def of canonicalistion, and for Impl Note
ACTION: jermemy update proposal on tex-01
ACTION jermemy to email i18n to get written confirmation re advice they give on
normal form C for IRIs (affirm, withdraw?)


Agenda:
	http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0163.html


Transcript:
	http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0177.html

1.Volunteer Scribe:
===================
	  
Jan Grant volunteered. Thanks Jan!


2: Roll Call
============

Regrets: Dan Connolly, Eric Miller

Participants:
	 DanBri (scribe)
	 Frank Manola
	 Graham Klyne 
	 Mike Dean 
	 Jeremy Carroll
	 Brian McBride
         Dave Beckett
         Jan Grant


3: Review Agenda
================

AOB: some discussion of the document that you get if 
     you HTTP GET on the RDF uri ref

danbri: is that M+S only or RDFS too?
dave: mostly former



4: Next telecon
===============

	04 Apr 2003 1000 Boston Time

Volunteer Scribe:
	  Jan Grant volunteered. Thanks Jan!


5: Minutes of 21 Mar 2003 telecon
=================================

	http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0138.html
	http://www.w3.org/2003/03/21-rdfcore-irc.html

Approved.

6: Confirm Status of Completed Actions
======================================

All recorded as DONE.


7: Last call review of webont docs
==================================

log:
	http://ilrt.org/discovery/chatlogs/rdfcore/2003-03-28#T15-08-43

Congratulations to WebOnt on their decision to go to LC

(except Test Cases, which will follow)

Jeremy: TC didn't go LC, as Tests are incorrect

brian: need to find reviewers for the reference document
they asked for Reference, Semantics, TestCases
...latter not appropriate (yet)
volunteers?

timescale: 24st april, draft review to list

see also:
    http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0137.html
    Reviewing Owl documents
    From: Brian McBride (bwm@hplb.hpl.hp.com)
    Date: Mon, Mar 24 2003 

danbri: I intend to review Reference, but am overloaded so is at risk.
Jang: I'll review semantics
brian: anyone else got time to commit to doing Reference?
jjc: since January, S+AS has improved in a linear fashion; Ref change was non-linear

brian: I'll review OWL Reference

ACTION: danbri review OWL Reference when published as Last Call
ACTION: Brian review OWL Reference when published as Last Call
ACTION: JanG review OWL Semantics when published as Last Call




8: Last call review of cc/pp doc
================================

Call for reviewers.

see http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0162.html

log:
	http://ilrt.org/discovery/chatlogs/rdfcore/2003-03-28#T15-16-19

ACTION: Brian to review CC/PP spec for next friday


9: Editors Drafts
=================

see:	http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0148.html
log:	http://ilrt.org/discovery/chatlogs/rdfcore/2003-03-28#T15-22-16

ACTION: brian to put pointers on web page to Shadow TR
ACTION: brian to update shadow TR for Semantics

brian:
...when we are sending responses, we should point to things that are in 
the archived record. 
...i'll point to current docs in shadow TR as editors WDs
...Pat, can I copy yours, put it up there/



10: Status on Incoming Last Call Comments
=========================================

log:	http://ilrt.org/discovery/chatlogs/rdfcore/2003-03-28#T15-40-21

discussion of 2003-03-14#1 (jang): take a final pass of the comments list
	to identify remaining items that have not been dealt with
	or been assigned process numbers

Jang: will get to this early next week


-Have editors reviewed the editorial comments from the xml schema WG?


ACTION: Dave Beckett to get edits for Syntax from XML Schema review
ACTION: Jeremy to update Concepts in light of XML Schema LC review
ACTION: Frank to check for editorial actions on Primer from XML Schema LC review
ACTION: DanBri to check for editorial actions on Schema from XML Schema LC review



Handling late comments.


brian: better shape than we were
...but new stuff just come in, SVG

dave: 4 weeks after LC ended, still ack'ing comments
...should say 'sorry but no'

danbri: obligation is for the period announced, but
...we can say 'we don't plan to treat this as a formal LC comment, 
but we nevertheless value it and will bring to attention of the WG and 
editors'

dave: do we have to handle comments forever? I close two, two more arrive...


jjc: i think it depends on the substance of the comment
...if it results in an improvement, we should look at it
...some are a waste of time



frank: two dimensions here
...if someone points out a ghastly error, we have to be responsible
...but process perspective is w.r.t. formal comments during LC
...we have to have a cut off date


dave: we're now two months after start of LC
...date should be today

brian: i disagree with this
...with having a fixed date

dave: hard to progress

brian: we have to balance a number of things ...have to get finished
...formally, process is that the Call ended
...WG has discretion to accept comments after the deadline
...in a sense dave is right, deadline was when set
...anything since, accepted out of goodness of our hearts
...as frank says, if glaring error we shouldn't ignore
...also politcal aspect, if we get comments from other WGs, 
we should err on the accepting side
...as need to build relationships with other WGs

see logs for full discussion: 
    http://ilrt.org/discovery/chatlogs/rdfcore/2003-03-28#T15-52-56

brian: is there support for taking view... be polite... receptive... 
with comments from now on, we'll decide collectively whether to feed them 
into LC process

danbri: what do you mean collectively? use telecon time to discuss whether to disucss...

pat: discuss on list, editors should be primary filter

brian: yup. default is still to accept
...editors can say 'this is a waste of time..., we can ignore this time'

danbri:
we have a standing invitation to the world to send RDF comments to www-rdf-comments.  they're just trying to be helpful...


11: Status on responses to Last Call Comments
=============================================

log:	http://ilrt.org/discovery/chatlogs/rdfcore/2003-03-28#T15-58-35

Awaiting responses to:
  http://www.w3.org/2001/sw/RDFCore/20030123-issues/#macgregor-01
  http://www.w3.org/2001/sw/RDFCore/20030123-issues/#macgregor-02
  http://www.w3.org/2001/sw/RDFCore/20030123-issues/#qu-01
  http://www.w3.org/2001/sw/RDFCore/20030123-issues/#qu-02
  http://www.w3.org/2001/sw/RDFCore/20030123-issues/#danc-01 ??
  http://www.w3.org/2001/sw/RDFCore/20030123-issues/#williams-01


[missed detail of discussion]

ACTION: gk send rseponse on macgregor-02
ACTION: jjc send rseponse on danc-01

12: Schedule
============

log:	http://ilrt.org/discovery/chatlogs/rdfcore/2003-03-28#T15-59-52

brian:
current microschedule "now seems a tad optimistic :("

http://www.w3.org/2001/sw/RDFCore/#microschedule

Noting www 2003:  http://www2003.org/ 20-24th May 2003

brian: WWW2003 looms 
...a good thing to aim for, to stand up and say something 'good'
...such as 'we processed and integrated all LC review comments'

May 24th?


pat: that's reasonable
danbri: a reasoonable goal
jjc: as a goal, couldn't guarantee
gk: too long!
jjc: i can only spent 1.5-2 days / week on RDFCore currnetly
pat: have we more substnative things to do? dotting is and crossing ts and fixing little bugs

[...]



13: Issues reagle-01, reagle-02
===============================

log:	http://ilrt.org/discovery/chatlogs/rdfcore/2003-03-28#T16-07-20


  http://www.w3.org/2001/sw/RDFCore/20030123-issues/#reagle-01
  http://www.w3.org/2001/sw/RDFCore/20030123-issues/#reagle-02

Proposal from Jeremy:

  http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0151.html

See:
  http://www.w3.org/2001/sw/RDFCore/20030123-issues/


jjc: we discussed this previously
essense of change... to present in the specs model in whuch 
syntax spec does all the work + then add impl notes that 
say one can be a valid implementation by not doing 
the canonicalisation (?bad scribe notes?)

dave: general outline is good

jjc: implementation note in concepts is key point.
...impl are free to represent things in other forms (...)

jjc: ref to non-exclusive canonicalisaion gets dropped

ACTION: dave to check on whether status of 'notes' should be informative vs normative

dave: re reagle-01/2, is it 2 faced to say 'do it this way, but you don't have to'

ACTION: jeremy to add anchors to Concepts for def of canonicalistion, and for Impl Note

RESOLVED: accept jjc's proposal
(unanimously)

14: Issue pfps-04,pfps-05,pfps-06,pfps-07,pfps-10
=================================================

  http://www.w3.org/2001/sw/RDFCore/20030123-issues/#pfps-04
  http://www.w3.org/2001/sw/RDFCore/20030123-issues/#pfps-05
  http://www.w3.org/2001/sw/RDFCore/20030123-issues/#pfps-06
  http://www.w3.org/2001/sw/RDFCore/20030123-issues/#pfps-07
  http://www.w3.org/2001/sw/RDFCore/20030123-issues/#pfps-10

log:	http://ilrt.org/discovery/chatlogs/rdfcore/2003-03-28#T16-20-23

brian: who else could pick it up?
...jang, your review?

jang: basically positive. could be more explicit in a few places
context: review semantics editor's draft wrt
changed arising out of pfps-04 -05 -06 -07 -08 -10

brian: pat, you're off email until mid next week anyway
...may as well put it off another week. gk if you manage the 
review, that'd be great. if not, we will have to go with the 1 we have.




15: Issue horrocks-01
=====================

log:	http://ilrt.org/discovery/chatlogs/rdfcore/2003-03-28#T16-26-26

  http://www.w3.org/2001/sw/RDFCore/20030123-issues/#horrocks-01

See:
  http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0081.html

discussed action 2003-03-14#11 (path) produce words for a resolution to horrocks-01

pat: prob do thurs next week


16: Issue tex-01
================

log:	http://ilrt.org/discovery/chatlogs/rdfcore/2003-03-28#T16-26-49


 http://www.w3.org/2001/sw/RDFCore/20030123-issues/#tex-01

Jeremy's proposal to resolve:
 http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0159.html
jjc: in a way quite sim to canon'n one ...sol'n is similar
...move effort into an impl note
...on other hand opposite
...should lang tag be normalised to lowercase?
...ab syntax suggests should lowercase
...an i18n reviewer (not a Group comment) was concerned that
 lowercase norm'n was incorrect
...so should add a note saying 'only for purposes of abstract syntax that it is lc'

brian: why nromalise this way?

jjc: we wanted graph to be tidy with low effort

..if literals equal, need to be identical
...rfc3066 on lang tags, they're case insensitive
...tidyness violated by = instead of == test
...so we normalise


pat: we could normalise to uppercase
jjc: why not, we don't care
b: why not say isn't a case insensitive compare?
p: graphs become untidy
d: i don't like that
p: just say that component of the literal is a 'language tag', and 
   identity for those defined elsewhere
p: some doc somewhere should draw readers attention to lc caselesness
b: we were told not to normalise to lowercase
...why that not work for you, j?
j: we could say in 'lit equality, two lits compare equal if...'
j: I'll come back with that proposal

ACTION: jermemy update proposal on tex-01

17: Issue tex-02
================

irc: http://ilrt.org/discovery/chatlogs/rdfcore/2003-03-28#T16-36-28-1

 http://www.w3.org/2001/sw/RDFCore/20030123-issues/#tex-02

Jeremy's proposal to postpone:
  http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0155.html  

[discussion]

danbri: i rec postponment, but being careful to track this and related 
	issues for drafting of future WG charters
brian: i think we should postpone

jang notes on a design:
     http://ilrt.org/discovery/chatlogs/rdfcore/2003-03-28#T16-50-03-1


RESOLVED: accept http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0155.html
 - amended to empaphasize that the proposed equals operation is inappropriate for the graph mechanism
(2 abstained)






18: Issue williams-02
=====================


log:	http://ilrt.org/discovery/chatlogs/rdfcore/2003-03-28#T16-52-47

 http://www.w3.org/2001/sw/RDFCore/20030123-issues/#wiliams-02

Jeremy's proposal:
  http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0154.html 


jjc: I heard from MartinD, that i18n now no longer think normal form C good for IRIs
...they like it for Literals though
...but for IRIs, not mandatory or even suggested

jjc: we need an email from them saying that

...if we get that, and amend Concepts accordingly
...then our def of a URIref is same as Namespaces1.1's notion of an IRIref
jjc: message ref'd from agenda should use IRIref, not IRI
jjc: duplication not as good as reference...

ACTION jermemy to email i18n to get written confirmation re advice they give on normal form C for IRIs (affirm, withdraw?)

gk: stablitiy of IRI spec... it's still up for discussion
...how to deal w/ how chars such as SPACE, '>', '<' in an IRI are to 
       be encoded in (say) RDF/XML. Also note some discussion about 
       additional chars in IRIs noted at IETF URI BOF
..also gen concerns about stability of cited txt
...as it is a piece of text that says it'll itself be replaced at some point


danbri: do we know whether test cases need changes?

jjc: 3 or 4 testcases would have to change

jos: commenting on gk's....
...namespaces 1.1 is in CR, just point to that, whatever they do...
...i support idea of IRIrefs as used in CR

frank: i oppose notion of changing to say 'IRIs'
...i believe it might be appropriate to have a note that our syntax 
     for rdf urirefs is in fact an iri as defined in this doc
...but i prefer to have our pointer for such a basic concept not to be 
       to an XML spec, but to a web arch spec of some kind

pat: me too
danbri: q+ to express concern about URI2IRI

b: there isn't one that defines IRIs
as chair, mindful to avoid change unless well motivated

ADJOURNED.

Received on Friday, 4 April 2003 08:45:02 UTC