W3C RDF Core Working Group IRC Chat Logs for 2003-07-11

13:00:01 Users on #rdfcore: @logger

13:59:12 <danbri> zakim, this will be RDF_Core

13:59:12 <Zakim> I do not see a conference matching that name scheduled near this time, danbri

13:59:16 <danbri> hi brian

13:59:21 <danbri> zakim, this will be RDFCore

13:59:21 <Zakim> ok, danbri, I see SW_RDFCore()10:00AM already started

13:59:28 <danbri> zakim, who is here?

13:59:28 <Zakim> On the phone I see ??P11

13:59:29 <Zakim> On IRC I see bwm, Zakim, danbri, Ralph, ps, logger

13:59:37 <bwm> G'day

13:59:44 <bwm> Are we having fun yet?

13:59:59 <bwm> Good to see you Ralph.

14:00:39 <Zakim> +Ralph

14:00:46 <Ralph> Hi, Brian

14:01:07 <Ralph> zakim, ??p11 is Patrick

14:01:07 <Zakim> +Patrick; got it

14:01:18 <Zakim> +PatH

14:01:36 <Zakim> +DanBri

14:01:43 <ps> ps is now known as PStickler

14:01:55 <danbri> Ralph, thanks for joining us :)

14:01:57 <Zakim> +FrankM

14:02:28 <Ralph> I'm substitute EM for this telecon

14:02:32 <Zakim> +??P17

14:03:38 <Ralph> zakim, ??p17 is probably Jan

14:03:38 <Zakim> +Jan?; got it

14:04:28 * bwm dialing

14:04:31 * DanC Zakim, call DanC-BOS

14:04:31 * Zakim ok, DanC; the call is being made

14:04:32 <Zakim> +DanC

14:04:52 <Zakim> +??P16

14:05:00 <bwm> Zakim, ??p16 is bwm

14:05:01 <Zakim> +bwm; got it

14:05:14 * DanC suffers vonage wierdness...

14:05:18 <Ralph> the noise is from Patrick

14:05:22 <Ralph> zakim, who's muted?

14:05:23 <Zakim> I see Patrick muted

14:05:35 * Ralph zakim, unmute patrick temporarily

14:05:35 * Zakim Patrick should no longer be muted

14:05:41 * DanC Zakim, call DanC-BOS

14:05:41 * Zakim ok, DanC; the call is being made

14:05:44 <Zakim> +GrahamKlyne

14:05:50 * Zakim Patrick should now be muted again

14:05:50 * Ralph zakim, unmute patrick

14:05:51 * Zakim Patrick should no longer be muted

14:05:51 * DanC often has to do the 80# dance on the 1st call of the day

14:05:54 <danbri> OK, RDFS should reflect the iff->if changes now (see mail)

14:05:57 <Ralph> zakim, who's muted?

14:05:57 <Zakim> I see no one muted

14:06:12 * DanC Zakim, drop DanC

14:06:12 * Zakim DanC is being disconnected

14:06:13 <Zakim> -DanC

14:06:15 * DanC Zakim, call DanC-BOS

14:06:15 * Zakim ok, DanC; the call is being made

14:06:16 <Zakim> +DanC

14:06:33 <bwm> Zakim, who is on the phone?

14:06:33 <Zakim> On the phone I see Patrick, Ralph, PatH, DanBri, FrankM, Jan?, DanC, bwm, GrahamKlyne

14:06:52 <DanC> DanC has changed the topic to: 11Jul http://www.w3.org/2001/sw/RDFCore/

14:07:00 <jang> ralph is substitute EM for this call

14:07:15 <DanC> agenda + 11Jul http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Jul/0132.html

14:07:15 * Zakim notes agendum 1 added

14:07:30 <jang> regrets jjc, jos, daveb

14:07:40 <jang> comments on agenda, aob?

14:07:47 <jang> DanC: two hours?

14:07:54 <jang> martin asked to turn up at 11

14:08:11 <jang> pats: would like to see straw poll on alternatives I posted

14:08:22 <DanC> DanC has changed the topic to: 11Jul http://www.w3.org/2001/sw/RDFCore/ chair: BrianM; scribe: JanG

14:09:05 <jang> http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Jul/0152.html pats' straw poll message

14:09:27 <jang> bwm pencils in 10a to prep discussion

14:09:49 <danbri> danbri has changed the topic to: 11Jul http://www.w3.org/2001/sw/RDFCore/ chair: BrianM; scribe: JanG agenda: http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Jul/0132.html

14:09:53 <jang> telecon next week same time

14:09:59 * DanC wonders about doing this literal stuff without jeremy... ah... gk is here, at least

14:10:21 <DanC> Zakim, pick a scribe

14:10:21 <Zakim> Not knowing who is chairing or who scribed recently, I propose DanC

14:10:25 <DanC> bad bot

14:10:26 <DanC> Zakim, pick a scribe

14:10:26 <Zakim> Not knowing who is chairing or who scribed recently, I propose FrankM

14:10:35 <DanC> Zakim, pick a scribe

14:10:35 <Zakim> Not knowing who is chairing or who scribed recently, I propose PatH

14:10:38 <jang> scribe... (gk pencilled in)

14:10:47 <jang> DanC, do we do this until it picks the person we're after?

14:10:53 <DanC> :)

14:10:59 <jang> minutes of telecon, jos sent regrets

14:11:18 <jang> actions done... approved

14:11:20 <jang> misc actions.

14:11:24 * Ralph zakim, mute patrick temporarily

14:11:24 * Zakim Patrick should now be muted

14:11:39 * Zakim Patrick should now be unmuted again

14:11:45 <PStickler> I'll try to figure it out...

14:11:58 <jang> jang: to respond to webont.

14:12:25 <jang> path: I told webont yesterday to the effect that there'd be no show-stoppers

14:12:33 <jang> bwm: I hope you were right!

14:12:43 <jang> danc has DONE action 20030627#7

14:12:46 <jang> item 8

14:12:58 <jang> closing semantics comments

14:13:12 <jang> item 13 scratched, it's about xml literals

14:13:39 <PStickler> Looks like the noise was my two cell phones annoying one another (too close)...

14:13:39 * DanC is getting lost; I just see a bunch of message numbers in the agenda

14:13:40 <jang> reviews of these responses?

14:13:55 <jang> gk: I've read, not reviewed

14:14:01 <jang> bwm asks for summary of pan-01

14:14:18 <bwm> pan-01 is msg 16

14:14:20 <jang> path: the comment was "do it this way"

14:14:27 <gk> I remember reading that now, and thought it was an excellent response

14:14:33 <jang>http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Jul/0016.html

14:14:39 <jang> pan was after layered languages

14:14:53 <jang> path's point was that this was a general as possible design decision

14:14:59 <jang> it's good theoretical basis for this

14:15:13 <jang> aczel is a prof at manc, where pan is

14:15:31 * DanC q+ to express discomfort with post-hoc decision rationale and ask for at least a pointer to the section in concepts that records some of the relevant design principles

14:15:31 * Zakim sees DanC on the speaker queue

14:15:32 <jang> also pointed out that a "layered" mode of operation is possible within our deisgn

14:15:45 <jang> eg, jjc's webont work

14:16:12 <jang> on recognition of syntacic structuring

14:16:15 * DanC realizes this isn't post-hoc... it's proposed to decide this here and now

14:16:24 <jang> this was less of a response as an explanation of why we did it this way

14:16:30 <jang> and how he might be happier with this design

14:16:33 <bwm> ack danc

14:16:33 <Zakim> DanC, you wanted to express discomfort with post-hoc decision rationale and ask for at least a pointer to the section in concepts that records some of the relevant design

14:16:36 <Zakim> ... principles

14:16:37 * Zakim sees no one on the speaker queue

14:16:37 <jang> so ultimately, a polite rejection of the comment

14:16:39 <jang> danc:

14:16:54 <jang> I heard path say "an explicit design decision".

14:17:00 <jang> path: "motivation" might have been better

14:17:10 <jang> DanC: can we DECIDE today to use these words then?

14:17:13 <jang> also,

14:17:22 <jang> concepts has words about this. Good to point to that.

14:17:28 <Ralph> zakim, mute patrick

14:17:28 <Zakim> Patrick should now be muted

14:17:32 <jang> path: jjc has also suggested referrences

14:17:52 <DanC> "I support this." -- jjc

14:17:53 <jang> bwm: gk liked this; jjc supports this response too

14:18:28 <jang> bwm: sounds like path proposes this as resolution for pan-01

14:18:40 <jang> DanC: with amendment to point to concepts

14:18:58 <danbri> I support it.

14:19:09 <jang> (danc happy with this decision using these words)

14:19:18 <jang> path prop; jang seconds. no abstain, no agin

14:19:19 <jang> resolved.

14:19:22 <jang> NEXT:

14:19:32 <jang> http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Jul/0015.html pfps-02

14:19:37 <danbri> http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Jul/0015.html pfps-02

14:19:48 <jang> path: lbase isn't available in a form stable enough for pfps to check

14:20:01 * danbri snap,sorry -- was trying to help w/ url pasting...

14:20:01 <jang> we need to get a more up to date and stable version of this document in public space

14:20:21 <jang> [cheers danbri, keep doing that in case I miss them]

14:20:38 <jang> path: pfps didn't liek the big red text saying "this is not final"

14:20:43 * Ralph q+ to ask PatH what he needs to be able to update LBase on www.w3.org Web space

14:20:44 * Zakim sees Ralph on the speaker queue

14:20:46 <DanC> I see http://www.w3.org/TR/2003/NOTE-lbase-20030123/

14:21:11 <danbri>http://www.w3.org/TR/2003/NOTE-lbase-20030123/

14:21:21 <DanC> perhaps this http://www.w3.org/2002/06/lbase/ should point there

14:21:24 <jang> (path: can links point to new version, will help here)

14:21:37 <jang> DanC: cite the 23 jan link

14:21:44 <jang> path: yes, I'll go back to pfps with this

14:22:02 * Ralph zakim, who's muted?

14:22:02 * Zakim sees no one muted

14:22:06 <Ralph> zakim, mute patrick

14:22:06 <Zakim> Patrick should now be muted

14:22:07 <jang> path: I'll have to double-check that this is sufficient to cite

14:22:10 * danbri thanks pstickler; oh noise came back.

14:22:22 <jang> bwm: does it include lbase translation of rdfs semantics?

14:22:32 <jang> path: I'll bring that into line with the semantics doc

14:22:37 <jang> it's an appendix/comment however

14:22:52 <jang> bwm: so semantics would cite out of date lbase rdfs translation

14:23:01 <jang> ... is this kosher or should it be fixed?

14:23:04 <jang> DanC: it's kosher

14:23:47 <jang> this isn't a technical matter that affects testable features of rdfs

14:23:52 <jang> path: true

14:23:57 <bwm> q?

14:23:57 * Zakim sees Ralph on the speaker queue

14:24:53 <danbri> I have updated http://www.w3.org/2002/06/lbase/ to point to the W3C Note in TR space.

14:24:53 <jang> ralph on the queue...

14:24:56 <bwm> ack ralph

14:24:56 <Zakim> Ralph, you wanted to ask PatH what he needs to be able to update LBase on www.w3.org Web space

14:24:56 <jang> ralph:

14:24:58 * Zakim sees no one on the speaker queue

14:25:04 <jang> does path need help in getting this updated?

14:25:25 <jang> I can take action to talk with eric as to why that lbase note isn't cited from WG page

14:25:36 <jang> bwm: I can fix that, it should be listed

14:25:54 <jang> ACTION bwm - to update rdfcore home page with link to lbase

14:26:10 <jang> (danbri does bwm's action right now!)

14:26:20 <jang> action path - to update link in semantics to new lbase doc

14:26:25 * DanC is still looking for what pfps-02 is

14:26:32 <jang> bwm: also hearing general consensus that this note is sufficient to cite

14:26:35 <jang> ok, then:

14:26:43 <Ralph> I believe /RDF/Overview.html should also cite the LBase Note

14:26:48 <jang> what is pfps-02 and what resolution are you proposing?

14:26:55 <jang> path: proposing to reject the comment

14:27:00 <bwm>http://www.w3.org/2001/sw/RDFCore/20030123-issues/#pfps-02

14:27:01 <jang> that lbase appendix should be trashed

14:27:05 <jang> rejected on the grounds

14:27:05 <DanC> I'm confused. the msg says "we accept your comment."

14:27:10 <jang> (a) it's informative, not normative

14:27:15 <jang> oops

14:27:23 <jang> this is simpler. this is that there's a mistake that there's an erorr

14:27:34 <jang> disposition: accpet this, point out this is fixed

14:27:35 <DanC> pfps-02 is summarized as "Lbase translation of XML Literals"

14:27:44 <jang> so I'll go back to pfps with the new accurate citation of lbase doc

14:28:42 <jang> should we hang fire on this because of xmlliteral?

14:28:43 <gk>http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Jul/0012.html

14:28:44 <jang> apparently so.

14:28:47 <DanC> "pfps-04RDF closure rules"

14:28:52 <gk> Link above

14:28:55 <jang> thanks gk

14:29:02 <jang> path summarises:

14:29:20 <jang> there are / were bugs in closure rules. the response is: accepted, yes there was, that's been fixed

14:29:45 <jang> bwm: and pfps likes this, right?

14:29:48 <jang> path: yes, he agreed.

14:29:58 <jang> ok, path prop to close pfps-04 second danc

14:30:01 <jang> no agin no abst

14:30:03 <jang> CLOSED

14:30:15 <jang> actions arising:

14:30:24 <jang> ACTION path respond to pfps (he may have already done that)

14:30:53 <jang> path: ok, I'll respond to -comments

14:31:13 <jang> pfps-09

14:31:14 <gk> pfps-09: link: http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Jul/0011.html

14:31:22 <bwm>http://www.w3.org/2001/sw/RDFCore/20030123-issues/#pfps-09

14:31:23 <DanC> "pfps-09datatypes"

14:31:35 <jang> path: I've run it past pfps, he says, yes, it's ok now

14:31:42 <jang> DanC: is this oservable from a test?

14:32:01 <jang> path: only in owl

14:32:18 <jang> only if you say DTname1 same as DTname2

14:32:29 <jang> originally it wouldn't have worked right, now it does

14:32:34 <danbri> Aside: OK, http://www.w3.org/2001/sw/RDFCore/#documents now links to LBase. re <jang> (danbri does bwm's action right now!)

14:32:35 <jang> it was basically a matter of wording.

14:33:18 <gk> q+ to suggest we might give "offcial" blessing to an OWL test case?

14:33:18 * Zakim sees gk on the speaker queue

14:33:28 <jang> DanC: I'm bothered that you can't tell the difference from the test, but, "ok".

14:33:51 <bwm> ack gk

14:33:51 <Zakim> gk, you wanted to suggest we might give "offcial" blessing to an OWL test case?

14:33:53 * Zakim sees no one on the speaker queue

14:33:57 * Ralph seconds gk's proposal

14:35:00 <jang> DanC: I'm willing to try to get webont wg to make such a test case

14:35:15 <jang> ACTION danc check with jjc, get a test case into owl documents for this

14:35:22 <jang> this = pfps-09

14:35:30 <jang> path prop, danc seconds, no agin, no abst

14:35:32 <jang> RESOVLED

14:35:39 <jang> action path - to respond to pfps on pfps-09

14:35:44 <jang> ITEM 9

14:35:49 <jang> pfps-22 pfps-23

14:35:52 <gk>http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Jul/0024.html

14:35:54 <DanC> "pfps-22reserved names in abstract syntax"

14:36:56 <jang> gk: first, do we believe that we SHOULD clarify the text here?

14:37:04 <DanC> what text?

14:37:13 <jang> second, if we DO believe that we should clarify this, then is there a possibility to discuss

14:37:24 <jang> something along the lines of the message pointed to above?

14:37:29 <jang> bwm: jjc suggested this is good enough

14:37:35 <DanC> could we do a little more by-value discussion and less by-reference?

14:38:18 <jang> DanC: I'm looking from a message from peter, saying that rdf reserves some urirefs

14:38:26 <DanC> "Certain URI references are reserved for use by RDF and should not be used

14:38:26 <DanC> in ways not supported by the RDF specficiations."

14:38:33 <jang> frankm: are there test csaes that deal with this?

14:38:51 <jang> jang: rdf-names-use, syntactical restrictions

14:39:02 <jang> those are restrictions on rdf/xml

14:39:45 <jang> q+

14:39:45 * Zakim sees jang on the speaker queue

14:39:57 <PStickler> q+

14:39:57 * Zakim sees jang, PStickler on the speaker queue

14:39:59 <jang> path: concepts section 4 apparently

14:40:12 <bwm> ack jang

14:40:12 * Zakim sees PStickler on the speaker queue

14:40:23 <DanC> section 4 concepts, editor's draft (I think) http://www.w3.org/2001/sw/RDFCore/TR/WD-rdf-concepts-20030117/#section-URIspaces

14:40:47 <bwm> ack pstickler

14:40:47 * Zakim sees no one on the speaker queue

14:40:49 * Ralph ack patrick

14:40:50 * Zakim unmutes Patrick

14:40:50 * Zakim sees no one on the speaker queue

14:41:13 <jang> jang: the rdf-names use are about an RDF/XML parser having interpretations for bits of the syntax such that RDF/XML parser can't put some things into the abs syntax; not that they can't be there.

14:41:14 * DanC thinks patrick started before he was unmuted

14:41:29 <danbri> I would strike "Certain URI references are reserved for use by RDF, and should not be used in ways not supported by the RDF specifications." as it adds nothing.

14:41:43 <jang> pats: maybe just a warning that you say, "if you do this, genericity across rdf reasoners is lost"?

14:41:51 <jang> DanC: ed draft concepts 1 july section 4

14:42:06 <jang> strike the sentence ... danc can you post the striking bit?

14:42:25 <DanC>http://www.w3.org/2001/sw/RDFCore/TR/WD-rdf-concepts-20030117/#section-URIspaces

14:42:28 <jang> section 4 para 1, sentence 2

14:42:55 <jang> danc suggests striking:

14:43:07 <jang> [

14:43:08 <jang> Certain URI references are reserved for use by RDF, and should not be used in ways not supported by the RDF specifications. Specifically,

14:43:09 <jang> ]

14:43:23 <jang> gk:

14:43:34 <jang> a little further down, there's a para that says this definitively

14:43:42 <jang> [[

14:43:51 <jang> Vocabulary terms in the rdf: namespace are listed in section 5.1 of the RDF syntax specification [RDF-SYNTAX]. Some of these terms are defined by the RDF specifications to denote specific concepts. Others have purely syntactic purpose (e.g. rdf:ID is part of the RDF/XML syntax) and should not be used in RDF to denote any kind of resource.

14:43:52 <jang> ]]

14:43:58 <jang> I was going to propose adding to that

14:44:01 <jang> ...

14:44:41 <jang> danbri: I think this is saying too much

14:45:06 <jang> if we have this wording in the text, we can't use rdf to talk about rdf syntactic structures

14:45:16 <danbri> propose strike "... should not be used in RDF to denote any kind of resource."

14:45:53 <jang> frankm: we're not claiming it makes sense to use rdf:id to name a resource, but we're not saying it can't be done

14:46:03 * Ralph zakim, mute patrick

14:46:03 * Zakim Patrick should now be muted

14:46:06 <jang> gk: I'm trying to get over the idea that...

14:46:18 <jang> anyone can come along and make additional assertions

14:46:26 * DanC can't hear gk

14:46:27 <jang> want to post a warning, this can lead to trouble

14:46:39 <jang> danbri: where do you stop?

14:46:54 <jang> you might claim, http://www.w3.org/ rdf:type fruit:Raspberries

14:46:58 <jang> if you do, well, good luck.

14:47:15 <jang> bwm: pfps asked, ok, what does rdf:id denote?

14:48:08 <jang> DanC: let's focus on specific wording changes

14:48:24 <jang> gk; I ahve no great attachment to these phrases,

14:48:40 <jang> some of the context is swapped out now, looking at it piecemeal, removing those problematic comments

14:48:46 <jang> doesn't seem to do any irreprable damage

14:48:51 <danbri> I propose striking the following from section 4 of concepts, "Some of these terms are defined by the RDF specifications to denote specific concepts. Others have purely syntactic purpose (e.g. rdf:ID is part of the RDF/XML syntax) and should not be used in RDF to denote any kind of resource." as well as " Certain URI references are reserved for use by RDF, and should not be used in ways not supported by the RDF specifications."

14:49:03 <PStickler> The restriction appears contrary to the open nature of RDF. The RDF specs say certain things about rdf: terms. Folks are free to disregard any statements not authoritative from the specs. But an explicit exclusion is unnecessary and likely will prevent

14:49:12 <PStickler> useful statements by others about RDF terms.

14:50:02 <jang> gk: the bit that starts "certain uri refs", ok, if it doesn't help it should go

14:50:21 <jang> gk: when I look at this, the intent I think is clear

14:50:30 <DanC> concepts can't cite schema normatively. (nor should semantics)

14:50:38 <jang> it could be put better, but I find it hard to work out any other reasonable interpretation

14:50:39 <DanC> RDF is defined independent of RDFS.

14:50:56 <PStickler> Systems concerned with additional statements mucking up the metamachinery can simply exclude statements about resources denoted by RDF terms from other sources than the official schemas.

14:51:06 <jang> ^^^ what he said

14:51:18 <DanC> this section (4) is the only normative reference to [RDF-VOCABULARY]

14:51:26 <jang> [scribe's having trouble keeping up with this]

14:51:30 <bwm> q?

14:51:30 * Zakim sees no one on the speaker queue

14:51:55 <jang> frankm: if you have a definite prohibition, that is, you cannot use these resource names in certain ways, then there are going to have to be

14:52:07 <jang> test cases, eg: here is unacceptable behaviour for an redf processor

14:52:23 <jang> frankm: we don't say anything about you constructing triples with these resources referenced in them

14:52:25 <danbri> Test case: <http://www.w3.org/1999/02/22-rdf-syntax-ns#ID> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://www.w3.org/2001/04/infoset#Attribute> .

14:52:30 <danbri> q+ to propose a test case

14:52:30 * Zakim sees danbri on the speaker queue

14:52:31 <jang> however, you may bve able to say peculiar things about this

14:52:37 <jang> so be careful about this

14:52:51 <jang> ie, say here be dragons, making absolute prohibitions needs test cases

14:52:52 <bwm> ack danbri

14:52:52 <Zakim> danbri, you wanted to propose a test case

14:52:53 * Zakim sees no one on the speaker queue

14:53:15 <jang> danbri talks about claim above

14:53:30 <jang> danbri: IO'd like to tell an rdf processor about this, without it throwing an excpetion.

14:53:47 <PStickler> RDF should allow anyone to say anything about anything, even resources denoted by RDF terms. Whether one chooses to trust or include such statements is a separate issue.

14:53:56 <jang> bwm: so are we sending gk to delete those bits of text?

14:54:13 <jang> bwm: gk wants time to review this

14:54:39 * jang wonders if martin is going to have trouble dialling in at this late stage..?

14:54:51 * Ralph q+ to advocate the conservative approach

14:54:51 * Zakim sees Ralph on the speaker queue

14:55:25 <jang> danbri talks about the example

14:55:35 <PStickler> The RDF specs define conformant behavior. Additional statements may result in non-conformant behavior. The most we could/should do is warn folks about this.

14:55:42 * Ralph q+ to ...; drop just the two "should not ..." phrases

14:55:42 * Zakim sees Ralph on the speaker queue

14:55:48 <bwm> ack ralph

14:55:48 <Zakim> Ralph, you wanted to advocate the conservative approach and to ...; drop just the two "should not ..." phrases

14:55:50 * Zakim sees no one on the speaker queue

14:56:15 * DanC q+ ralph

14:56:16 * Zakim sees ralph on the speaker queue

14:56:17 * DanC oops

14:56:42 <DanC> danbri, you don't have to use that URI to refer to that (class of) attribute

14:57:05 * DanC thinks danbri and gk have made their positions clear... let others speak

14:57:14 <danbri> yup, i've said all i want to.

14:57:30 <jang> Ralph: unless we can find a specific problem case, we should drop the two sentences here

14:58:39 <PStickler> +q

14:58:39 * Zakim wonders where q is

14:58:39 <DanC> "conservative approach" misled me, ralph. I thought you meant 'conservative' in the sense of: warn folks in a conservative way not to shoot themselves in the foot.

14:58:49 <PStickler> q+

14:58:49 * Zakim sees ralph, PStickler on the speaker queue

14:59:04 <bwm> ack ralph

14:59:04 * Zakim sees PStickler on the speaker queue

14:59:21 <jang> Ralph: rdf/xml syntax warnings, maybe, but stronger should nots are premature I think

14:59:39 <jang> pats:

14:59:40 <bwm> ack pstickler

14:59:40 * Ralph ack pat

14:59:40 * Zakim sees no one on the speaker queue

14:59:41 * Zakim sees no one on the speaker queue

14:59:42 <Zakim> -GrahamKlyne

14:59:43 * Ralph ack patrick

14:59:43 * Zakim unmutes Patrick

14:59:44 * Zakim sees no one on the speaker queue

15:00:02 <Ralph> zakim, patrick is PStickler

15:00:02 <Zakim> +PStickler; got it

15:00:07 <jang> pats: want to partially agree with ralph (only partially)

15:00:09 <gk> Pbhone dropped off -- trying to redial

15:00:10 <jang> saying less is better

15:00:27 <jang> if we don't have clear use cases demostrating problems, there is no justification to add this constraint

15:00:54 <Zakim> +GrahamKlyne

15:01:05 * danbri q+ to ask for an AOB slot requesting a reviewer of my RDFS update regarding 'iff' -> 'if' semantics

15:01:05 * Zakim sees danbri on the speaker queue

15:01:12 <jang> secondly, we shouldn't want to prevent people making statements using rdf-spec-utilised urirefs

15:01:18 <jang> even if they're contradictory

15:01:38 <jang> i think we don't want to prevent useful things being said by people in the future

15:01:47 <jang> gk: I accept the WG's request, and I'll go for it

15:01:47 <MJDuerst> trying to dial in, getting "the conference is restricted"

15:02:00 <drachir> i am also unable to dial in, it appears

15:02:21 <jang> ACTION gk to come back to mailing list with revised text, just for a final pass

15:02:26 <drachir> drachir is now known as ishida

15:02:27 <danbri> Ralph? can you help w/ Zakim magic?

15:02:31 <Ralph> yes, I can help

15:02:34 <Zakim> +Martin

15:02:42 <danbri> thanks Ralph

15:02:49 <Ralph> the problem is that the bridge is only reserved for an hour, so it won't let callers in

15:03:22 <ishida> ralph, can you let me in too, please?

15:03:23 <jang> danbri, danc, are you happy dropping the should nots, not the whole sentences?

15:04:01 <jang> gk: to paste into irc

15:04:28 * DanC asks danbri to give gk the floor

15:04:39 <gk> proposed new para 1, section 4:

15:04:40 <gk> RDF uses URI references to identify resources and properties.

15:04:40 <gk> Certain URI references are reserved for use by RDF.

15:04:40 <gk> Specifically, URI references with the following leading substrings are

15:04:40 <gk> reserved for definition by the RDF specifications:

15:04:44 <Zakim> +Ishida

15:05:12 <jang> can someone capture the rest of the proposal?

15:05:32 <bwm> [Certain URI references are reserved for use by RDF, and should not be used in ways not supported by the RDF specifications.]

15:05:35 <bwm> becomes

15:05:38 <danbri> Martin, Richard, we are just wrapping up a decision, shouldn't take much longer...

15:05:49 <bwm> Certain URI references are reserved for use by RDF.

15:05:49 <MJDuerst> ok, danbri

15:05:53 <gk> proposed new 3rd para, section 4:

15:05:54 <gk> Vocabulary terms in the rdf: namespace are listed in section 5.1 of the

15:05:54 <gk> RDF syntax specification [RDF-SYNTAX].

15:05:54 <gk> Some of these terms are defined by the RDF specifications to denote

15:05:54 <gk> specific concepts. Others have purely syntactic purpose

15:05:54 <gk> (e.g. rdf:ID is part of the RDF/XML syntax).

15:06:47 * jang wishes to point out that scribe cannot extract this proposal to minutes. Will only be a pointer to transcipt

15:06:50 <gk> revised 1st para:

15:06:51 <gk> RDF uses URI references to identify resources and properties.

15:06:51 <gk> Certain URI references are given specific meaning by RDF.

15:06:51 <gk> Specifically, URI references with the following leading substrings are

15:06:51 <gk> reserved for definition by the RDF specifications:

15:07:15 <gk> try again:

15:07:16 <gk> RDF uses URI references to identify resources and properties.

15:07:16 <gk> Certain URI references are given specific meaning by RDF.

15:07:16 <gk> Specifically, URI references with the following leading substrings are

15:07:16 <gk> defined by the RDF specifications:

15:07:47 <ishida> zakim, mute ishida please

15:07:47 <Zakim> Ishida should now be muted

15:07:50 <jang> general "sounds good to me"s

15:07:55 <jang> AGREED.

15:07:59 <danbri> q-

15:07:59 * Zakim sees no one on the speaker queue

15:07:59 <jang> bwm: was there another para?

15:08:16 <PStickler> zakim, mute PStickler

15:08:16 <Zakim> PStickler should now be muted

15:08:16 <jang> gk: vocabulary terms in the rdf namespace...

15:08:20 <jang> up to

15:08:32 <jang> (eg, rdf:id...)

15:08:52 <jang> general "sounds good"s - path worried about the implied contrast

15:09:15 <PStickler> Some terms are relevant to the semantics of RDF/XML, others are relevant to the semantics of the graph.

15:09:20 <jang> path: delete "purely"

15:09:32 <jang> pats catches that

15:09:35 <jang> AGREED

15:09:38 <jang> and done (phew)

15:09:43 <jang> moving on:

15:09:53 <ishida> zakim, unmute me

15:09:53 <Zakim> Ishida should no longer be muted

15:10:00 <jang> martin duerst, richard ishida

15:10:21 <jang> so moving on

15:10:27 * gk (BTW, there's still some text in section 2.6 that Peter didn't like)

15:10:29 <jang> ITEM 11: post last call comment form I18N

15:10:49 <jang> for martin: agenda here:

15:10:55 <jang>http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Jul/0132.html

15:10:57 <jang> item 11

15:11:03 <jang> martin:

15:11:08 * DanC hears a (tolerable) echo

15:11:12 <jang> we see the original design in M+S as having

15:11:21 <jang> plain text litas and slightly marked up literals

15:11:25 <jang> and building up from there

15:11:40 <jang> and we want to suport that, esp. in intl context we expect this to increase

15:11:48 <jang> eg, the <title> example for books is the canonical example

15:12:14 <jang> with proviso that deep rdf semantics are not I18N's purview,

15:12:26 <jang> rdf model has moved further and further away from that position

15:12:43 <jang> it gets increasingly difficult to get a natural transition to text with markup

15:13:00 <jang> there appear to be a lot of inconsistencies in this work from our point of view and from a user's point of view

15:13:05 <jang> language is taken away from XMl literals

15:13:13 <jang> so that's our current main concern.

15:13:24 <jang> of course, we have different options to move on from here.

15:13:30 <jang> I hope we can get out of this bind.

15:13:34 <jang> richard:

15:13:45 <jang> process-related comment. A number of possible complicated appraoches here

15:13:53 <jang> it'll be good to keep our eye on end-user requirements here

15:14:03 <jang> I haven't been able to follow the discussion in detail, I fear

15:14:21 <jang> but the end-user requirement is that information has language information assigned to it at the appropriate moment

15:14:33 <jang> DanC: all of the designs I've seen meet that requirement

15:14:53 <jang> rishard: it's easy to get lost in low-level arguements, this is a yardstick

15:15:06 <jang> perhaps martin can expand on this?

15:15:07 <Zakim> +EricP

15:15:27 <jang> martin: in M+S and in the LC you could add just a little amrkup

15:15:41 <jang> then you could find away around this, but it's not at all consistent and smoothe.

15:15:43 <PStickler> zakim, unmute PStickler

15:15:43 <Zakim> PStickler should no longer be muted

15:15:45 <jang> smooth even

15:15:54 <jang> bwm: does anyone have any questions on this?

15:15:56 <PStickler> q+

15:15:56 * Zakim sees PStickler on the speaker queue

15:16:18 <jang> path: in your recent message replying to pats' proposal

15:16:26 <jang> you said something that puzzles me greatly

15:16:40 <jang> you want to distinguish between stuff that "looks like XML" and stuff that IS XML

15:16:48 <jang> that seems to be at odds with earlier stuff,

15:17:02 <jang> that is, a piece of XML that doesn't have XML markup should be considered plain text

15:17:02 * DanC q+ to object to the constraint on the discussion that we don't talk about solutions; it's in the discussion of solutions that we find out what the problems are, in my experience. I want to know if losing the semantic distinction between XML structures and strings that look like XML is ok by I18N

15:17:03 * Zakim sees PStickler, DanC on the speaker queue

15:17:08 <jang> I can't reconcile those positions in my head

15:17:14 <jang> can you elaborate?

15:17:17 * DanC Zakim, mute me

15:17:17 * Zakim DanC should now be muted

15:17:31 <jang> martin: if you have plain text and you add markup to it,

15:17:42 <jang> then that should be different thing and adding & or < characters to it

15:17:58 <bwm> q?

15:17:58 * Zakim sees PStickler, DanC on the speaker queue

15:17:59 <jang> that doesn't create a conflict in my view

15:18:50 <jang> martin (paraphrasing I think that &lt; foobar / &gt; isn't the same as <foobar />

15:19:10 <jang> the distinction between plain and MXL literals (at least up until the LC )

15:19:17 <jang> was a technical distinction for technical reasons

15:19:21 <ishida> zakim, mute me

15:19:21 <Zakim> Ishida should now be muted

15:19:28 <jang> it turned out for tech reasons to be more convenient to make such a distinction

15:19:51 <jang> if plain text just happens to be labelled as XML literal or not, those should not be completely different things.

15:19:57 <jang> pats:

15:19:58 <bwm> ack patrick

15:19:58 * Zakim sees PStickler, DanC on the speaker queue

15:20:04 <bwm> ack Pstickler

15:20:04 * Zakim sees DanC on the speaker queue

15:20:18 <jang> like many people I've been trying very hard to understand what you want to see happen in rdf

15:20:25 <jang> because we want to see I18N done right

15:20:35 <jang> but your concerns don't seem to fit with ANY of the proposals on the table.

15:20:43 <jang> I think maybe what you'd like to see happen is that

15:20:52 <jang> if there's language context in the XML,

15:21:12 <jang> that if someone adds markup in an xml literal, that contextual information will "flow down" into the XML literal

15:21:29 * DanC q+ to ask about preservation of namespaces

15:21:29 * Zakim sees DanC on the speaker queue

15:21:33 <jang> the problem seems to be that you don't want a semantic distinction between

15:21:48 <jang> serialisation model for the RDF graph and the encapsulated XML markup

15:22:02 <jang> ie, the X view rather than the G view

15:22:23 <jang> martin:

15:22:33 <jang> I'm sorry I didn't manage to answer all your emails,

15:22:40 <jang> you've got a lot of reasonable questions and comments

15:22:59 <jang> I originally focussed on the XML view, perhaps too much for your taste,

15:23:16 <jang> what happens currently is xml:lang is inherited by plain literals

15:23:36 <jang> exc-can is used for xml literals

15:23:48 <jang> but the question I see is: why not xml:lang on XML Literals>

15:23:52 * danbri zakim, who is on the phone?

15:23:52 * Zakim sees on the phone: PStickler, Ralph, PatH, DanBri, FrankM, Jan?, DanC (muted), bwm, GrahamKlyne, Martin, Ishida (muted), EricP

15:24:22 <jang> your main applications seem to be "closed" pieces of XMl fragment

15:24:46 <jang> and in those cases, yes, we might think that that can carry its own language tag

15:24:56 <jang> in other cases, what we might call "open" fragments

15:25:02 <jang> it's not as easy as that.

15:25:18 <jang> pats: I agree that XML itself has a problem with dealing with mixed-content fragments

15:25:32 <jang> but I can't see how RDF is supposed to fix that shortcoming of XML while retaining some

15:25:41 <bwm> q?

15:25:41 * Zakim sees DanC on the speaker queue

15:25:41 <jang> way of dealing with XML fragments as literals

15:25:43 <DanC> ack danc

15:25:43 * Zakim unmutes DanC

15:25:44 <Zakim> DanC, you wanted to object to the constraint on the discussion that we don't talk about solutions; it's in the discussion of solutions that we find out what the problems are, in my

15:25:46 <jang> danc:L

15:25:48 <Zakim> ... experience. I want to know if losing the semantic distinction between XML structures and strings that look like XML is ok by I18N and to ask about preservation of namespaces

15:25:50 <PStickler> zakim, mute PStickler

15:25:50 * Zakim sees no one on the speaker queue

15:25:51 <Zakim> PStickler should now be muted

15:26:29 * gk I'm all at sea ... it would be good to have an example of something that the current solution doesn't adequately address

15:26:29 <jang> DanC: there's this example from feb 1999 M+s spec

15:26:39 <jang> the NS decl for mathml is outside the title element

15:26:54 <jang> by the time you get it into an rdf graph, the namespace decl will have gone away

15:27:07 <jang> so that's one way you can tell if the rdf graph has captured this

15:27:18 <jang> I saw implementations doing this

15:27:31 <jang> so I raised this, the WG addressed this

15:27:32 <Ralph> DanC is referring to example 7.5 in M&S

15:27:53 <jang> however, I've been unable to find concrete examples of problems

15:28:01 <jang> so if this goes away again I wouldn't scream too hard.

15:28:14 <gk> I thought we'd already undone the namespace import

15:28:16 <jang> path: jjc had the strongest opinions, alas he's not here today,.

15:28:33 * Ralph q+ to reflect on M&S intent

15:28:33 * Zakim sees Ralph on the speaker queue

15:28:49 <jang> martin: I'm not a namespace expert, I'm not asking to throw out NS, I'm asking to reintroduce the lang tag

15:29:03 <jang> DanC: one proposal is partype=literal is like a cdata section.

15:29:17 <jang> (danc explains analogy)

15:29:26 <Ralph>http://www.w3.org/TR/1999/REC-rdf-syntax-19990222/#examples

15:29:28 <jang> ie, after the parser's done with it, it just looks like a string

15:29:33 <jang> containing angle brackets

15:29:48 <jang> and if it's just a string, then any extant NS decls disappear, right?

15:30:01 <PStickler> zakim, unmute PStickler

15:30:01 <Zakim> PStickler should no longer be muted

15:30:04 <jang> bwm: don't think that's right. if it's a string, the string would be produced according to exc-can

15:30:11 <jang> pats: that's correct

15:30:25 <jang> ie, paretype=literal introduces a string according to exc-can

15:30:26 <Ralph>http://www.w3.org/TR/1999/REC-rdf-syntax-19990222/#ex-Literal

15:30:30 <jang> ralph:

15:30:48 <jang> I wish I could cite ancient email archive messages, but:

15:30:48 <bwm> q frank

15:30:54 <bwm> Zakim, q frank

15:30:54 <Zakim> I don't understand 'q frank', bwm

15:30:54 <jang> q+ frank

15:30:55 * Zakim sees Ralph, frank on the speaker queue

15:30:59 <jang> ack ralph

15:30:59 <Zakim> Ralph, you wanted to reflect on M&S intent

15:31:00 * Zakim sees frank on the speaker queue

15:31:01 * bwm thanks jang

15:31:13 <jang> rdf M&S tries hard to be xml friendly

15:31:38 <jang> that is, that the RDF serialisatioin (... incl of XML) doesn't look overly unusual to an "XML user"

15:31:47 <jang> so we did discuss whether things should be escaped

15:32:05 <jang> the group took a strong position that that kind of thing would separate RDF too far from the xml community

15:32:10 <jang> and that would hamper uptake too much

15:32:10 <tex> I couldn't phone in on 7332

15:32:28 <ishida> hold on tex

15:32:36 * bwm thanks ishida

15:32:46 <ishida> ralph will help in a moment

15:32:46 <jang> the idea was to rely on an rdf processor to have done full XML processing

15:33:00 <danbri> tex, sorry, let's try to fix that for you... hmm can you try again and go thru to the human operator?

15:33:01 <jang> ie, the context that you found an xmol fragment in would have in-scope namespaces, language, etc.

15:33:11 <DanC> oh the irony! the RSS community tends to write &lt;p&gt; in their RSS description fields. 1/2 ;-)

15:33:20 <jang> :-)

15:33:40 <jang> we wanted to leave some of this decision to the xml community

15:33:44 <jang> anyway

15:33:56 <Zakim> +Tex

15:34:13 <jang> we expected that the context in the XML would be expressed in XMl literals

15:34:15 <bwm> welcome Tex

15:34:23 <tex> thanx

15:34:23 <jang> so if we could survey all the deployed apps

15:34:27 <jang> and find that nobody's done that

15:34:31 <jang> we'd have a bit more freedom

15:34:36 <jang> but my position is that

15:34:49 <jang> it's a change to the expectations of M&S

15:34:51 <PStickler> q+

15:34:51 * Zakim sees frank, PStickler on the speaker queue

15:35:00 <PStickler> zakim, mute PStickler

15:35:00 <Zakim> PStickler was already muted, PStickler

15:35:15 <danbri> there are several ways that folk do it in RSS -- that was one concern the echo folks are trying to address. http://intertwingly.net/wiki/pie/Motivation

15:35:17 <jang> to say that literals may not have, or are precluded of having namespace or language context associated with them

15:35:25 <PStickler> zakin, unmute PStickler

15:35:26 <jang> I think that's part of where i18n is coming from

15:35:33 <jang> frank:

15:35:38 <bwm> ack frank

15:35:38 * Zakim sees PStickler on the speaker queue

15:35:48 <jang> people are coming from several differnet directions.

15:36:01 <jang> let me drag us back to the inheritanec of context direction.

15:36:24 <jang> are we saying that: the literals of whatever kind they are, or pt=L things or lits associated with DTs

15:36:38 <jang> do they mandatorially have lang tags that they might have inherited?

15:36:50 <jang> is that information required to be carried in the rdf graph?

15:36:52 <jang> in addition

15:37:06 <jang> are rdf processors free to ignore the presence of those language tags in doing whatever tehy want?

15:37:16 <jang> one of the reasons I raised the NS issue

15:37:21 * DanC wishes frank would leave "rdf processor" out of the discussion... hmm... he doesn't seem to be in IRC

15:37:32 <gk> q+ to say that the namespace context of XML literals is absent in the LC version, and the latest editors' version

15:37:32 * Zakim sees PStickler, gk on the speaker queue

15:37:35 <jang> if the use of XML requires rdf processors to carry all of that machinery

15:37:44 <jang> then I'm worried waht the impact on DT literals

15:38:04 <jang> do we need the graph to carry all this information

15:38:11 <jang> [scribe wonders, and any future context?]

15:38:19 <jang> so we can reconstruct that context in the future?

15:38:19 <PStickler> zakim, unmute PStickler

15:38:19 <Zakim> PStickler should no longer be muted

15:38:24 <jang> pats:

15:38:26 <bwm> ack pstickler

15:38:26 * Zakim sees gk on the speaker queue

15:38:40 <jang> the background information ralph provided was very informative

15:39:01 <jang> I'd view it that M&S approach supports/assumed one particular mode of use of XML literals

15:39:20 <jang> I'd like to refine my first proposal a little to support the needs of most groups:

15:39:22 <DanC> "1st proposal"?

15:39:31 <jang> it's be essentially identical to the first proposal:

15:39:43 <DanC> pls let's discuss by-value, not by reference. don't assume folks have read all recent mail

15:39:46 <jang> that DT literals, INCLUDING xml literlas, would not inherit contextual information

15:39:52 <jang> whereas non-types literals would

15:40:03 <jang> non-typed even

15:40:26 <jang> ah

15:40:29 <bwm> q?

15:40:29 * Zakim sees gk on the speaker queue

15:40:30 <jang> pats expands

15:40:36 <jang> difference between parsetype literal

15:40:40 <jang> and pt=L, dt=foo

15:41:10 <jang> martin: I'm not sure I understand your proposal fully,

15:41:16 <jang> so I'd like to think it through first

15:41:34 <jang> one thing: would that mean that there'd be two ways of putting XML into rdf?

15:41:46 <jang> one to say, it's just a plain literal

15:41:51 <jang> the other to say that it's datatyped?

15:42:06 <jang> pats: already there's more than one way to put xml into an rdf graph, with escaping etc

15:42:29 <jang> there'd be a difference between contextual versus encapsulated xml

15:43:03 <jang> martin: given I'm careful to write these things nicely,

15:43:16 <jang> and use the contextual inside the non-contextual

15:43:22 <jang> would the two things in the graph end up the same?

15:43:28 <bwm> q?

15:43:28 * Zakim sees gk on the speaker queue

15:43:28 <jang> pats: no, they'd look entirely different

15:43:33 <jang> q+

15:43:33 * Zakim sees gk, jang on the speaker queue

15:44:09 <jang> pats: [expands a little on the graph difference]

15:44:25 <jang> pats: I can rewrite the proposal in those terms

15:44:27 <jang> ack gk

15:44:27 <Zakim> gk, you wanted to say that the namespace context of XML literals is absent in the LC version, and the latest editors' version

15:44:28 <bwm> ack gk

15:44:29 * Zakim sees jang on the speaker queue

15:44:30 * Zakim sees jang on the speaker queue

15:44:34 <PStickler> zakim, mute PStickler

15:44:34 <Zakim> PStickler should now be muted

15:44:34 <jang> gk:

15:44:45 <jang> I was a little bit concerned by what I heard ralph saying

15:44:58 <jang> that XML contestual information should be available or a "part of" an xml literal

15:45:07 <jang> I've been looking at the LC and ed's latest draft

15:45:08 <PStickler> Contextual literals: traditional M&S literals, xml:lang xml:base, namespaces, etc. apply

15:45:23 <jang> as far as I can tell, there's no namespace context carried into xml literals

15:45:30 <PStickler> Non-contextual literals: datatyped literals, no xml:lang, xml:base, namespaces, etc. apply

15:45:40 <jang> martin: exc-can it doesn't include all the namespaces, but

15:46:01 <jang> it includes namespace definitions for stuff included inside the xml literal

15:46:24 <jang> gk: the lexical space of xml literals corresponds to the exc-can + comments

15:46:34 <jang> and with empty includive namespaces prefix list

15:46:41 <jang> martin: we'd have to check with the spec

15:46:55 <jang> I think that just means don't include irrelevant namespaces

15:47:12 <jang> bwm: what that does, it includes all VISIBLE namespaces

15:47:16 * DanC wonders if there are tests about namespace visibility

15:47:19 <jang> a qname inside an attribute isn't visible

15:47:26 <jang> an element or attribute name makes it visible

15:47:47 * DanC q?

15:47:47 * Zakim sees jang on the speaker queue

15:47:49 <jang> the current position basically means only visible namespaces are included in the context

15:47:50 <DanC> ack jang

15:47:50 * Zakim sees no one on the speaker queue

15:48:15 <PStickler> zakim, unmute PStickler

15:48:15 <Zakim> PStickler should no longer be muted

15:48:21 <jang> jang: what about future bits and pieces?

15:48:28 <jang> do we include future bits of context?

15:48:31 <jang> DanC: no

15:48:38 <jang> path; they's why we use a dated documetn

15:48:38 <DanC> that was PatH, but I concur

15:48:51 <jang> [sorry, hard to place a single syllable[

15:48:53 <DanC> oh... you mean no, there are no tests.

15:49:06 <jang> path: the answer to that question has to beno.

15:49:13 <PStickler> zakim, mute PStickler

15:49:13 <Zakim> PStickler should now be muted

15:49:23 <jang> jang: yes, that's what I'd have thought, was just checking.

15:49:47 * gk thanks for clarifying NS thing that brian, I'm less concerned now

15:50:25 <jang> martin: jjc made request from rdfcore to i18n

15:50:31 <gk> path, XML context applies to typed literals?

15:50:36 <jang> to ask, what about integers, do we need lang information for integers?

15:50:44 <jang> we replied that in our discussion with xml schema

15:50:55 <jang> is that the intent of these DTs and the way they are designed

15:51:05 <Ralph> I concur with PatH; the dependencies in the RDF specs are to the then-current versions of the other specs. If the other specs evolve, the RDF specs do not implicitly change, however they may be explicitly changed.

15:51:06 <jang> is to be independent of any locale, presentation, etc, considerations

15:51:54 <jang> so rdfcore decided that language shouldn't apply to DTed things

15:51:58 <PStickler> zakim, unmute PStickler

15:51:59 <Zakim> PStickler should no longer be muted

15:52:03 <jang> bwm: any other questions?

15:52:26 * gk so what's wrong with the current design?

15:52:34 <jang> pats: martin, do you have any significant objection to there being a treatment of literals containing xml markup that are noncontextual

15:52:48 <jang> ie, a non-contaminating treatment of XML literals

15:52:58 <jang> martin: I'd have to think about the consequnces of that again.

15:53:09 <jang> I'd have to consider if it'd confuse or hepl people.

15:53:23 * DanC wonders which design gk means by "current"; the most recent WG decision made xml:lang not inherit into parseType="Literal", to which I18N objects

15:53:27 <jang> Ralph: I'd have a hard time trying to imagine such a thing being useful in practice.

15:53:42 <jang> ex, markup having no namespaces, etc.

15:53:47 <jang> pats: that's not what I said/meant

15:54:17 * gk I meant the most recent (no lang on XML literals) ... the question is in search of understanding, not rhetorical

15:54:20 <jang> I mean, and XML fragment where they can tell, say from rdf-expressed information, what the intended context is

15:54:35 <jang> [pats alludes to expresing language in the rdf graph]

15:54:50 <jang> ^^ in explicit triples, even

15:54:55 <danbri> does xmlns:lang="" switch off lang tagging now, in generic XML?

15:55:01 <danbri> (for which version of xml?)

15:55:27 * gk my phone battery's nearly dead :-(

15:55:34 <jang> pedal faster gk

15:55:44 * gk :-))

15:57:12 <jang> martin: xml:lang="" or some new syntax to express "don't include these namespaces" wuold be on=k to exclude context

15:57:29 <jang> bwm: have we identified a showstopper problem with current editor's draft

15:57:49 * DanC Zakim, mute me

15:57:49 * Zakim DanC should now be muted

15:57:57 <jang> path: I can't see what the objection to the current design; all I get is the sense that the current design isn't "nice"

15:58:08 <jang> gk, jang echo that.

15:58:18 * DanC unmute me

15:58:22 <danbri> I think the objection is that our usage is dramatically counter-intuitive to people who have learnt about xml:lang in an XML context.

15:58:26 <DanC> Zakim, unmute me

15:58:26 <Zakim> DanC should no longer be muted

15:58:27 <Zakim> -GrahamKlyne

15:58:29 <Zakim> -Tex

15:58:29 <Zakim> -DanC

15:58:30 <Zakim> -EricP

15:58:32 <jang> pats: I believe I do understand martin's opinion

15:58:33 <Zakim> -Ishida

15:58:47 <jang> I don't see it as a showstopper, but I am sympathetic

15:58:52 <DanC> Zakim decided we're done, I think. Well done, Zakim, 2 hours is plenty.

15:58:54 <danbri> zakim, who are here?

15:58:54 <Zakim> I don't understand your question, danbri.

15:59:00 <danbri> zakim, who is on the phone?

15:59:00 <Zakim> On the phone I see PStickler, Ralph, PatH, DanBri, FrankM, Jan?, bwm, Martin

15:59:06 <tex> did the call get terminated?

15:59:14 <gk> I got kicked off

15:59:14 <ishida> zakim, i was terminated

15:59:14 <Zakim> I don't understand 'i was terminated', ishida

15:59:23 <danbri> ralph 'we seem to have lost one of our T1s'

15:59:33 <MJDuerst> sorry, some technical problems with Zakim

15:59:42 <jang> bwm: martin, sorry to cut you off

15:59:47 <jang> my current position as chair

15:59:52 <Ralph> we still have free ports, however the conf code expired an hour ago

15:59:52 <danbri> tex, we are just wrapping up...

15:59:55 <jang> I don't think you've persuaded people there's a showstoppe

16:00:07 <jang> that's where I see things standing at the moment, would invite last comment from you

16:00:11 <jang> martin:

16:00:23 <jang> it's a question of degree, you've moved incrementally from the original solution

16:00:33 <jang> I agree that the last tiny step isn't the showstopper

16:00:43 <jang> this is changed post-LC and I don't see the motivation for that change

16:00:51 <jang> trhere are alternatives that wouldn't have hurt that much.

16:00:55 <jang> bwm: thanks for those remarks.

16:01:06 <jang> I've tabled discussion on this for the next SW coord meeting

16:01:09 <jang> this also affects webont

16:01:10 * danbri mute pstickler

16:01:22 * danbri about to do: /me zakim, mute pstickler

16:01:24 <jang> what I'll say there is that we've not heard a showstopper, but i18n are clearly not happy with the present situation

16:01:25 <danbri> ...but noise stopped

16:01:28 <jang> would that be fair?

16:01:29 <PStickler> zakim, mute PStickler

16:01:29 <Zakim> PStickler was already muted, PStickler

16:01:36 <jang> martin: that'd be a close description, thanks

16:01:41 <jang> bwm: then we're done.

16:01:48 <Zakim> -FrankM

16:01:48 <jang> thanks for taking a constructive approach to this.

16:01:49 <Zakim> -bwm

16:01:50 <Zakim> -PatH

16:01:52 <jang> meeting closes.

16:01:52 <Zakim> -Martin

16:01:53 <Zakim> -Ralph

16:01:55 <PStickler> zakim, unmute PStickler

16:01:55 <Zakim> PStickler should no longer be muted

16:02:00 <DanC> FYI, re bwm's last question, I18N's objection is clear to me: it's unacceptable to not have xml:lang 'inherit' into parseType="Literal"

16:02:01 <Zakim> -DanBri

16:02:03 <ishida> martin, can we talk?

16:02:06 <jang> Zakim, who is on th ephone?

16:02:06 <Zakim> I don't understand your question, jang.

16:02:11 <jang> Zakim, who is on the phone?

16:02:12 <Zakim> On the phone I see PStickler, Jan?

16:06:23 <Zakim> -Jan?

16:06:24 <Zakim> -PStickler

16:07:27 <danbri> logger, bookmark?

16:07:27 <danbri> See http://ilrt.org/discovery/chatlogs/rdfcore/2003-07-11#T16-07-27

16:07:49 <danbri> thx everyone; bye.

16:08:09 <Zakim> SW_RDFCore()10:00AM has ended