W3C RDF Core Working Group IRC Chat Logs for 2003-03-11

16:03:08 Users on #rdfcore: logger DaveB GK RRSAgent em jjc @bwm

16:05:06 <GK> GK is now known as gk-scribe

16:05:20 <jjc> jeremy, frank

16:05:23 <jjc> brian, daveb

16:05:25 <jjc> graham

16:05:28 <jjc> ericm

16:05:48 <em> regrets from ...

16:05:54 <em> Jan, MikeD, Jos

16:06:00 <gk-scribe> Regrets: Jan, MikeD, Jos

16:06:33 <gk-scribe> Present, PatH

16:07:14 <gk-scribe> [Brian wonder's if we're quorate]

16:07:32 <gk-scribe> jjc: suggests we make decisions, and then ask the group to ratify or challenge

16:09:11 <DaveB> +1 617 252 7000

16:10:43 <gk-scribe> Brian proposes: we progress as WG meeting, and send announcements of decisions after next meeting

16:11:32 <gk-scribe> Item 1. XML literal equality

16:11:35 <gk-scribe> [issue tag?]

16:12:00 <DaveB> reagle-01 XMLLiteral equality

16:12:08 * gk-scribe thanks dave

16:12:44 <DaveB>http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0025.html

16:12:46 <jjc>http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0025.html

16:13:52 <gk-scribe> WebOnt's problem: given XML literal, compare with self, answer depends on parser implementation.

16:16:19 <gk-scribe> Joseph's point: if decision is not definitive, why try at all (?)

16:16:54 <bwm>http://www.w3.org/TR/rdf-concepts/#section-XMLLiteral

16:17:18 <gk-scribe> ... value space of XML literal is canonical XML

16:17:53 <jjc>http://lists.w3.org/Archives/Public/www-webont-wg/2003Jan/att-0533/01-msg

16:18:27 <gk-scribe> Syntax talks about XML->Graph transformation

16:19:57 <DaveB>http://www.w3.org/TR/2003/WD-rdf-syntax-grammar-20030123/#parseTypeLiteralPropertyElt

16:20:11 <gk-scribe> Parser variability may result in variability in inference as defined by Concepts

16:20:41 <gk-scribe> We allow some freedom concerning comments and not-visibly-used namepsaces

16:21:00 <gk-scribe> Those choices are preserved by lexical->value mapping in concepts

16:23:08 <gk-scribe> Proposed that the lex->value mapping discards comments and not-visible-used namespaces, with commentary text that resoning systems should canonicalize on input.

16:24:34 <gk-scribe> Also has advantage that text is easier to understand.

16:25:13 <gk-scribe> Downside is that minimum RDF implementation is more complex, by introducing requirement to canonicalize.

16:25:36 <gk-scribe> Maybe, "SHOULD canonicalize"?

16:26:26 <gk-scribe> E.g. RSS tool that uses XHTML markup, but doesn't need to reason over XML literals

16:27:58 <gk-scribe> Brian: we're getting too worried about implementation. A non-canonical form is a perfectly good representation of a canonicalized form.

16:28:51 <gk-scribe> Q+

16:30:10 <gk-scribe> PatH prefers abstract syntax to be canonicalized

16:31:40 * em wonders if we can use zakim as floor control even if we cant take advantage of the rest of zakim/telecon features

16:31:59 <em> q+ gk-scribe

16:31:59 * Zakim sees gk-scribe on the speaker queue

16:32:28 <gk-scribe> Q+ to ask is the issue here that6 XML datatype is hardwired into the RDF core, other datatypes are not?

16:32:29 * Zakim sees gk-scribe on the speaker queue

16:32:34 * em wonders if bwm wants zakim to help with floor control ?

16:32:55 * bwm didn't know zakim was here

16:33:09 * em just invited him... "em wonders if we can use zakim as floor control even if we cant take advantage of the rest of zakim/telecon features"

16:34:59 <gk-scribe> q-

16:34:59 * Zakim sees no one on the speaker queue

16:36:14 <gk-scribe> Summing up... people want a new proposal in which syntax requires canonicalization, so c14n doesn't appear in L2V mapping and reasoning is sound.

16:36:50 <gk-scribe> Brian: abstract syntax requires C14N without comments, implememtations are free to use non-C14N as representation of C14N form.

16:37:56 <gk-scribe> (I think we can require N-triples to use C14N form, and still use Brian's suggestion.)

16:38:24 <gk-scribe> (... but Brian points out that breaks test cases.)

16:39:41 <em> +Mike_Dean

16:39:42 * Zakim wonders where Mike_Dean is

16:39:42 <gk-scribe> What implementations C14-ize? Jena, RDFLib, ....

16:41:04 <gk-scribe> Suggest: Require C14N, but add note in concepts that implementations can use non-C14N form internally to represent the C14N form

16:42:38 <gk-scribe> PatH: Then XML datatype is less of a datatype, more like string. Minimal change suggests keep XML literals as-are.

16:44:20 <gk-scribe> Patrick's preference is to keep comments. But this could lead to problems in pipelined implementations.

16:45:15 <DaveB>http://www.w3.org/TR/2003/WD-rdf-syntax-grammar-20030123/#parseTypeLiteralPropertyElt

16:45:16 <DaveB> 7.2.17

16:45:48 <gk-scribe> ACTION: jjc, to come back with a proposal that abstract syntax is C14N, with comments.

16:45:49 * RRSAgent records action 1

16:46:06 <bwm> thats abstract syntax for xml literal

16:46:40 <gk-scribe> Item -- social meaning

16:46:52 <gk-scribe> pfps-14 Social Meaning

16:47:08 <gk-scribe> Jeremy reports from meeting at Plenary.

16:47:25 <DaveB>http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0027.html

16:47:45 <gk-scribe> Unanimous recomendation to strike social meaning, and attendant issues.

16:50:18 <gk-scribe> Subsequently, GK has proposed: http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0029.html

16:50:44 <gk-scribe> DanBri seconded

16:51:47 <gk-scribe> GK proposes, EM seconds, no objections or abstentions

16:51:51 <gk-scribe> carried

16:52:27 <gk-scribe> Actions arising:

16:52:46 <gk-scribe> Check other sections of concepts doc

16:53:20 <DaveB> (rdfms-assertion appears in no WD)

16:53:31 <DaveB> (but words arising do)

16:53:45 <gk-scribe> Primer has example of people making statements in RDF. But the intended use hasn't chjamged, just that we're not trying to say what that is.

16:54:17 * gk-scribe notes that no LC comments about this came up against any other docs

16:54:43 <gk-scribe> ACTION: GK, delete section 4 of concepts

16:54:43 * RRSAgent records action 2

16:54:58 <gk-scribe> ACTION BWM: move rdfms-assertion to proposed

16:54:58 * RRSAgent records action 3

16:56:02 <gk-scribe> ACTION EDITORS: review docs for consequential changes

16:56:02 * RRSAgent records action 4

16:57:51 <gk-scribe> ACTION EM: to raise issue with CG "to prioritize further discussion, co-ordinated with the TAG over URI denotation"

16:57:51 * RRSAgent records action 5

16:58:26 <gk-scribe> >>>>>>

16:58:28 <gk-scribe> hendler-01 literals in parsetype collection

16:58:51 <jjc> redialing

16:58:52 <gk-scribe> DaveB proposes: Sorry, too late ;-)

17:02:18 * gk-scribe is the fact that encoding literals in lists suffciient cause to make a change?

17:02:31 * gk-scribe ... is ugly ...

17:03:02 <gk-scribe> Possible form of response:

17:03:31 <gk-scribe> Resolve to reject, on grounds... difficult change to syntax... late stage of affairs

17:03:46 <gk-scribe> s/Reject/Postpone/ ??

17:05:02 <DaveB>http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0034.html

17:05:06 <DaveB> ^- my summary so far

17:05:34 <gk-scribe> Propose to postpone ... would require extyensive change to current spec ... not critical as resulting graph is still describable in RDF/XML ...

17:06:02 <bwm> propose: RDFCore resolves to postpone this issue on the grounds:

17:06:11 <DaveB> would require extensive changes to current spec

17:06:11 <DaveB> not a critical requirement

17:06:14 <DaveB> for webont

17:06:39 <DaveB> that it would involve considering several different approaches, taking time

17:06:58 <DaveB> and consequent changes to syntax draft, test cases, implementations

17:07:03 <DaveB> and primer

17:08:11 <gk-scribe> Proposed bwm, seconder Jeremy

17:08:22 <gk-scribe> None against, none abstain

17:08:23 <jjc> s/bwm/DaveB

17:08:55 <gk-scribe> >>>>>>

17:09:02 <gk-scribe> hodder-01 RDF in HTML

17:09:44 <DaveB>http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0032.html

17:10:57 <gk-scribe> Second: GK

17:11:06 <gk-scribe> None against, none abstain

17:11:14 <gk-scribe> CARRIED

17:11:14 <gk-scribe> >>>>>

17:11:23 <gk-scribe> krech-01 emptyPropertyElt ambiguity

17:12:26 <gk-scribe> ACTION DaveB: respond to WEBONT on Hendler-01

17:12:27 * RRSAgent records action 6

17:12:33 <DaveB> (& to hendler)

17:12:47 <gk-scribe> ACTION GK: respond to (various people) on pfps-14

17:12:47 * RRSAgent records action 7

17:13:05 <gk-scribe> NOTE: hold off responses until after Friday

17:13:55 <gk-scribe> ACTION DaveB: respond to commentator on hodder-o1

17:13:55 * RRSAgent records action 8

17:14:21 <DaveB> krech-01,jjc-01 prposal to accept http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0033.html

17:14:59 * em mutes, be back in 3 min

17:15:13 <gk-scribe> DaveB proposes, per URI above

17:15:33 <gk-scribe> Seconded: GK

17:15:46 <gk-scribe> No onjections, no abstentions

17:15:53 <gk-scribe> CARRIED

17:15:56 <gk-scribe> Actions...

17:16:07 <gk-scribe> ACTION DaveB: to respond

17:16:07 * RRSAgent records action 9

17:16:32 <gk-scribe> ACTION DaveB: to update docs as outlined in message

17:16:32 * RRSAgent records action 10

17:16:45 <gk-scribe> >>>>>>

17:16:59 <gk-scribe> pfps-15 Say anything quote

17:17:19 <DaveB>http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0054.html

17:17:37 <DaveB> ^-frankM's proposal for pfps-15 etc.

17:23:23 <gk-scribe> Suggest s/information/statements/ in Franks revised wording???

17:24:37 <bwm> These examples also illustrate one of the basic architectural

17:24:37 <bwm> principles of the Web, which is that anyone should be able to freely add rdf statements about a resource, using any rdf vocabulary they please

17:24:37 <bwm> [BERNERS-LEE98].

17:24:43 <gk-scribe> Another suggestion: s/information/RDF statements/ in Franks revised wording???

17:27:47 <bwm> These examples also illustrate that anyone should be able to freely add rdf statements about a resource, using any rdf vocabulary they please

17:28:45 <jjc> These examples show that anyone should be able to freely add rdf statements about a resource, using any rdf

17:28:45 <jjc> vocabulary they please

17:29:01 <gk-scribe> Brian, I like that... maybe follow with "This is in keeping with a basic architectiral principle of the web ... [BERNERS-LEE98]

17:29:34 <jjc> illustrating the basic architectural principles of the Web "anyone [is] (technically) allowed to say anything about anything"

17:29:53 <jjc> [BERNERS-LEE98].

17:30:23 <gk-scribe> (jjc, that works for me)

17:30:31 <jjc> q+

17:30:31 * Zakim sees jjc on the speaker queue

17:32:20 <bwm> ack jjc

17:32:21 * Zakim sees no one on the speaker queue

17:34:50 <gk-scribe> Proposed: Frank's edirorial discretion to amend wording in light of the comment

17:35:15 <gk-scribe> Proposed: PatH

17:35:43 <gk-scribe> Seconded: Eric

17:35:45 <gk-scribe> None opposed, jjc abstains

17:36:11 <gk-scribe> ACTION FrankM: make change as discussed

17:36:11 * RRSAgent records action 11

17:36:24 <gk-scribe> ACTION FrankM: to respond

17:36:24 * RRSAgent records action 12

17:37:04 <gk-scribe> >>>>>>

17:37:13 <gk-scribe> danc-03 rdf/xml reification examples

17:40:21 <gk-scribe> Frank's response: http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0054.html

17:44:42 <gk-scribe> Proposed: Frank's editorial discretion to amend wording in light of the comment

17:44:48 <gk-scribe> Proposed: Frank

17:44:54 <gk-scribe> Seconded PatH

17:45:07 <bwm>http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Mar/0054.html

17:45:33 <gk-scribe> Proposal is discretion to make changes per message cited.

17:46:06 <gk-scribe> No objections, no abstentions

17:46:10 <gk-scribe> >>>>>>>>

17:46:21 <gk-scribe> pfps-17 namespace v vocabulary

17:47:09 <gk-scribe> Frank proposes to s/namespace/vocabulary/, except when sopecifically talking about XML namespaces

17:48:40 <DaveB> DaveB: comment on section "The RDF Vocabulary" with rdf namespace URI reference ...

17:50:40 <gk-scribe> Technically, "namespace name" is the technically correct term for "namespace URI [ref]" (per [XML-names]) ...???

17:52:47 <gk-scribe> DaveB seconds

17:52:53 <gk-scribe> None against, none abstain

17:52:57 <jjc> rdf core wg can say nothing about nothing without offending someone ...

17:53:05 <gk-scribe> ACTION FrankM: respond to Peter

17:53:05 * RRSAgent records action 13

17:53:26 <gk-scribe> ACTION FrankM: update the document

17:53:26 * RRSAgent records action 14

17:55:04 <gk-scribe> ACTION BWM: re-post biolerplate for response to comment

17:55:04 * RRSAgent records action 15

17:58:06 <DaveB> pat - horrocks-01 pfps-01 on friday

17:58:09 <DaveB> pat - qu-03

17:58:21 <DaveB> jjc - "pfps-23 is easy"

17:58:27 <DaveB> gk - williams-0?

17:58:34 <gk-scribe> williams-01

18:00:09 <gk-scribe> I think there are possible proposals for pfps-16, -18, -22, -23. I'd need to think about them

18:00:10 <DaveB> daveb - timbl01

18:00:41 <gk-scribe> Meeting closed.

18:01:06 <DaveB> logger, pointer?

18:01:06 <DaveB> See http://ilrt.org/discovery/chatlogs/rdfcore/2003-03-11#T18-01-06