W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > April 2002

(draft) minutes from 2002-03-22 teleconf

From: Dan Brickley <danbri@w3.org>
Date: Tue, 2 Apr 2002 21:49:05 -0500 (EST)
To: <w3c-rdfcore-wg@w3.org>
Message-ID: <Pine.LNX.4.30.0204022146500.27218-100000@tux.w3.org>


Folks, sorry for the delay and remaining sketchyness. The main points
(actions) were captured by Jema and circulated, and the IRC log
transcripts (crossref'd per issue) were as near verbatim as I could get.

Dan

===================================================================================

RDFCore WG minutes for the telecon 2002-03-22

IRC Transcript: http://ilrt.org/discovery/chatlogs/rdfcore/2002-03-22#T14-52-58

Agenda: http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Mar/0304.html

Jema action log: http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Mar/0326.html (copied below)

ACTIONS and DECISIONS (noted by Jema tool):

 Jema noted the following actions during RDFCore WG Telecon 2002-03-22

 2002-03-22#1:  daveb make the requested updates to the syntax wd
 2002-03-22#2:  ericm organise publication
 2002-03-22#3:  bwm update home page to point to test cases
 2002-03-22#4:  jeremy Send mailto:uri@w3.org with appropriate tests and
		explanation, explaining how we interpreted RFC2396 and xml base
 2002-03-22#5:  jjc send message to uri list telling folks how we
		interpreted rfc 2396 and xml:base specs with appropriate text
 2002-03-22#6:  pats publish DT asWD

Jema noted the following decisions during RDFCore WG Telecon 2002-03-22

    1:  publish syntax wd
    2:  dlex stays
    3:  there is no problem to fix by removing dlex



    Roll call - Participants:

    PatrickS
    AaronSw
    EricM
    JanG
    DaveB
    Manola
    JosD
    Jeremy
    AndyS (guest),
    Brian
    RonD
    DanBri
    DanC
    MDean

Regrets / Absent:
	@@everyone else


Review agenda - no amendments

Next telecon - 10am Boston time, 05 April 2002 (=4pm in UK, for this week)

Previous week's minutes -
 http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Mar/0235.html (2002-03-15)
   APPROVED


Confirm status of completed actions -
   ALL CONFIRMED.


==== topics


8: Status of Primer WD
2002-03-15#1  EricM  drive publication process for primer WD

PUBLISHED:
	  http://www.w3.org/TR/rdf-primer/
	  http://www.w3.org/TR/2002/WD-rdf-primer-20020319/

[15:08:42] DanC
	   at this point, the WG throws a small party for the primer editors. :-)
[15:08:43] Zakim
+MDean
[15:08:44] danb_scri
Well done Frank!
[15:08:52] danb_scri
Thanks to EricM et al for getting it done

Frank: note that we're geting comments already

Dave: I sent a summary out a few hours ago. Accepted changes from Graham, Jeremy.


9: Syntax WD
Propose authorize publication of new WD See:
  http://lists.w3.org/Archives/Public/www-archive/2002Mar/att-0053/01-rdf_syntax.html

Brian: Proposal to authorize publication
Frank: One minor problem, ie with the new drawings.

Some detailed discussion of the (now published) doc:
http://ilrt.org/discovery/chatlogs/rdfcore/2002-03-22#T15-12-42 for details.

Brian: Are there any showstoppers that'd prevent publication?
Brian: PROPOSE publish, with image fixed.

[15:21:35] bwm
-action daveb/ make the requested updates to the syntax wd
[15:21:36] Jema
Jema notes action 2002-03-22#1
[15:21:50] bwm
-action ericm / organise publication
[15:21:52] Jema
Jema notes action 2002-03-22#2









10: XML Base Test Cases
IRC log: http://ilrt.org/discovery/chatlogs/rdfcore/2002-03-22#T15-22-10

Brian: some of our test cases are testing our interpretation of (other specs), not RDF.
...a suggestion that this risks our tests being invalidated by other groups work

DanC: that's a reason to have tests!

graham: I second DanC:  if our interpretation is wrong, it helps reviewers to know it!


Jeremy: there are about eight cases in total

these linked from http://www.w3.org/2000/10/rdf-tests/rdfcore/

Jeremy: test cases are on website under xmlbase

DanC: can I get to the test case repository from group home page?

brian: no, will fix

action brian / update wg page to point to test case repository
Jema notes action 2002-03-22#3

* DanC wishes for a README in the directory

Jeremy: the relevant tests are numbered 7 and upwards, excluding number 14 and negative cases.

jeremy: they're summarised in a msg no 301 from March
 http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Mar/0301

jeremy: the point is that the tests describe approx a dozen cases of uris and their resolution

...only 3 going fwd. involve some degree of interpretation from me

DanC: I recently coded this up (2000/10/swap/uripath.py); I could test these, but I have not.

brian: issue is that the lack of clarity isn't RDF's fault

Jeremy: it makes meaning of an rdf/xml doc that uses it unclear

DanC: it matters for our spec more than some others

Jeremy: more pertinent for us than others as we use URIs as identifiers not for retrieval

Brian: current status is that they're all approved
error 01 isn't approved, test nn isn't

bwm: in general folk seem ok with having them in here, though my view is that we needn't

danc: would be fair to notify uri@w3.org of our issue/concerns

jeremy: at least one tag member worried (?)
brian: OK, we'll keep them.

graham: Hmmm... I'd not asay "at fault" so much as we're stating our interpretation

-action jeremy/ Send mailto:uri@w3.org with appropriate tests and explanation, explaining how we interpreted RFC2396 and xml base
[15:32:44] Jema
Jema notes action 2002-03-22#4


dave: in 3.6 (2nd note)
http://ilrt.org/discovery/2001/07/rdf-syntax-grammar/#section-baseURIs





11: Datatypes
The only outstanding issue against datatypes at present (did I really say that?) is the whether we need rdfs:dlex.  For discussion and decision. See:
  http://www.coginst.uwf.edu/users/phayes/simpledatatype23-02-2002.html
  http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Mar/0292.html

IRC log context: http://ilrt.org/discovery/chatlogs/rdfcore/2002-03-22#T15-36-04

http://www.coginst.uwf.edu/users/phayes/simpledatatype23-02-2002.html
http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Mar/0292.html


brian: as jeremy points out, we have two outstanding actions against datatypes.


1st: do we need rdfs:dlex

JosD:
dlex example http://www.agfa.com/w3c/n3/p7.nt

(discussion)

RESOLVED: dlex stays

DanC
I think the decision is that there's no problem to discuss.







12: Dark Triples
The need for this, and the timing, should be on the agenda for the next swcg telecon.
Early indications from the chairs of WEBONT are that that this issue may go away and that they
do not feel that a solution is needed in the next set of RDF specs to go to rec.
The chairs need a briefing from WG members also in WEBONT on their assessment of
the current situation and recommendations on the process for resolving it.

IRC context: http://ilrt.org/discovery/chatlogs/rdfcore/2002-03-22#T15-43-05

Brian: good stuff from PatH.
...have exchanged mail w/ jimhendler. No sense of urgency in "this round of specs".

Brian: expecting this to come up on monday's SW CG call
..webont members of rdfcore... coudl they brief us?

jeremy: there is a WebOnt f2f mid april

...a suggestion that rdfcore would be unwilling to change things. we might want to counter that...

danc: we currently have no plans to change anything

pat'sstuff starts here:
http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Mar/0253.html


some detailed discussion, partly captured in logs:
http://ilrt.org/discovery/chatlogs/rdfcore/2002-03-22#T15-45-17


path: hoping to go to webont f2f saying "hey, we solved your problem"
...so wanted to get it done in a hurry

Jang: Pat suggested two things. Dark triples and unasserted contexts.
...we want this for the provenance case that people have been trying to do with reification


...there are simple syntax proposals around...is there a reason to prefer dark triples than contexts?

Path: simpler than using contexts.

...contexts would be of great utility
...but would be a bigger change

JanG: many implementations have notion of a context
...often people code to quads not triples

...not a big step for people who've built real rdf systems

PatH: I thought contexts would be too much forthis wg

Jeremy: We aren't chartered to extend RDF like this.
...if webont demand such changes, them could investigate

...am ok with PAt going to webont meeting and saying 'rdfcore would support adding this if asked'

Brian: Thanks, I have plenty of input for SW CG meeting.





13: Internationalization

Brian: Proposal is that string components of literals should be in unicode normal form C.

IRC context: http://ilrt.org/discovery/chatlogs/rdfcore/2002-03-22#T15-55-25-1

(some detailed discussion, no actions/decisions recorded)





DanC: propose to extend meeting to discuss last call schedule

(meeting extended by 15 mins)

Brian: I put out a schedule, do we believe it is achievable?

Agenda item 16: Schedule Review
http://ilrt.org/discovery/chatlogs/rdfcore/2002-03-22#T16-05-50

Brian: I don't thing this is doable
PAth: could do monday after that.

danc: we can get a lot of review at May web conf. Go to last call shortly afterwards.

graham: can we get folk to read this at web conf

em: www2002 is great opportunity, but inclined to agree w/ brian. we're pretty tight on the schedule

...need to think about how datatypes relate to schema

dave: when we say 'going to last call', are we assuming everything is going on same day

brian: syntax, schema, model theory at same time

dave: in that case I vote for a MT asap

brian: other three are interdependent
danc: a lot will want primer to read
jeremey: can't separate test cases either
patricks: even if primer and test cases aren't going to last call......people will need to read them when reviewing the others

danc: if we had drafts of everything, and one or two at last call (Esp. Syntax). that'd be good.

patr: what dates?
7-8 may

pat: model theory / datatyping by them

danc: too agresssive re datatyping

* em corrects himself...

www2002 -  7-11 May 2002

brian: I persoanlly don't believe datatyping should be going out as agressively as we're doing it.

...at f2f was decided we wanted to integrate it into schema

DanC: best guess right now is a (datatypes?) WD in 3 weeks.

yes, datatypes WD

Frank: we can adopt as aggressive a schedule as we like, but if we're serious about listening to comments, we must accept that the users community may not be for it.

* danb_scri nods

patricks: not that datatyping is a shoe-in, but we should in 2-3 weeks put together a reasonably complete spec that can be reviewed.

Brian: datatyping puts rest of specs at risk?
...inclined it to maybe go out as a Note?
...the decision to include datatypes into the current Schema doc.

danc, danbri, daveb all happy with pullig DT out of schema
brian: suggestions that we do datatypes as a separate WD

patricks: I'm happy with that. prefer REC not Note.
...would be more useful to community with weight behind it of REC

patricks: if we're looking at schema going out...

split schema,dtas separate WDs

danc: the community haven't seen RDFS for ages, datatypes ever. We put them out shortly as WDs, and see what happens.

-action pats / publish DT asWD
[16:17:44] Jema
Jema notes action 2002-03-22#6

ADJOURNED.
Received on Tuesday, 2 April 2002 21:49:06 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:47:20 EDT