W3C RDF Core Working Group IRC Chat Logs for 2001-12-14

14:01:04 Topic now RDF Core WG Meeting Friday 15:00-16:00 UTC - Agenda http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Dec/0094.html

14:01:04 Users on #rdfcore: logger_2 @dajobe

15:00:15 <jan_g> jan, dajobe, bwm, jeremy at the ilrt all here

15:02:34 <mdean> mdean is now known as mdean_

15:03:16 <bwm> -open

15:03:17 <Jema> RDFCore WG Telecon 2001-12-14 is now open

15:03:17 <Jema> The agenda can be found at http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Dec/0094.html

15:03:17 <Jema> Agenda item 1: Allocate scribe

15:03:59 * jan_g is (unfortunately) leaving on a jet plane this weekend, can't minute

15:04:32 <jan_g> jan / jeremy minuting

15:04:37 <jan_g> -next

15:04:39 <Jema> command not recognised

15:04:54 <jan_g> danbri abs

15:04:54 <jan_g> bwm here

15:04:57 <jan_g> em here

15:05:00 <jan_g> dave b here

15:05:07 <jan_g> frank b abs, jeremy here, danc reg

15:05:20 <jan_g> ron d abs, bill d abs, jos here, rael abs

15:05:33 * DanC waves, debating, as usual on friday mornings, which of my obligations takes priority

15:05:34 <jan_g> jan here, martin h abs, yoshi abs, graham reg

15:05:46 <jan_g> mich kop abs kwon abs ora abs,

15:05:50 <jan_g> frank m here

15:06:00 <jan_g> satoshi abs, steve p here, pierre abs, patrick here

15:06:11 <jan_g> aaron reg, mike dean here, guha abs, pat hayes here

15:06:17 <jan_g> come on in dan, the water's lovely

15:06:21 <jan_g> #sergei en route

15:06:36 <jan_g> aob?

15:06:50 <jan_g> -item 3

15:06:51 <Jema> command not recognised

15:07:16 <jan_g> item 4 (next telecon)

15:07:20 <jan_g> review minutes:

15:07:31 <jan_g> any other commetns? no, approved

15:07:38 <jan_g> completed actions.

15:07:42 <jan_g> no objections, all done

15:07:53 <jan_g> status of withdrawn actions: ok

15:08:14 <dajobe> bwm noted f2f registration - better do soon see http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Dec/0081.html

15:08:43 * DanC thinks Jema has gotten to the "I want that" stage; would appreciate pointer to source.

15:09:40 <jan_g> Patrick: P writeup not in sergei's document?

15:09:57 <jan_g> pat h: what was the decision?

15:10:45 <jan_g> recap last week: pat had concerns about union datatypes

15:10:58 <jan_g> patrick: I think this is a non-issue

15:11:21 <jan_g> bwm: tech discussion delayed to item 15

15:11:48 <jan_g> ok item 7: not confirmed, open sytate

15:12:02 <jan_g> item 8: continues (danbri's got in touch with technical people)

15:12:06 <jan_g> item 9: primer status:

15:12:10 <jan_g> telecon yesterday

15:12:24 <jan_g> frank: currently there are new submissions

15:12:32 <jan_g> we're trying to get a common agreement on the outline

15:12:46 <jan_g> we feel onece that's straight, the rest "follows easily"

15:13:09 <jan_g> bwm: is there a plan on how to move forward from here?

15:13:17 <jan_g> frank: we're going to iterate over the outline

15:13:28 <jan_g> there's some good material to go into the doc

15:13:45 <jan_g> once the outline is clear then we'll have a much better idea of what material we need, how to get it, etc.

15:13:53 <jan_g> expect significant progress in the next few weeks.

15:14:13 <jan_g> bwm: can we get a timescale sense when you've got a reasonable idea of this?

15:14:17 <jan_g> frank: sure

15:14:24 <jan_g> em: feel this is quite urgent

15:14:44 <jan_g> people have earmarked time over the next few weeks for this

15:15:17 <jan_g> I anticipate documents out within weeks

15:15:29 <jan_g> jeremy: how much feedback are you looking for from within the community?

15:15:46 <jan_g> em: we're looking for that, obviously, but the primer's not at the maturity of the techie documents

15:16:10 <jan_g> bwm: any other comments?

15:16:11 <jan_g> moving on

15:16:17 <jan_g> item 10 mt wd

15:16:30 <jan_g> pat: I've got lots of feedback saying why it needs rewriting :-/

15:16:46 <jan_g> bwm: was the plan to publish this as wd?

15:16:57 <jan_g> ph: once the group has approved it, yes; next draft, basically

15:17:14 <jan_g> i want to publish the entaiolment notions quickly

15:17:32 <jan_g> ph: some small changes needed, but that's just a part of the process

15:17:48 <jan_g> jos: I'm going to have a very close look in the coming days

15:17:59 * DanC thinks the updated MT is worth publishing

15:18:07 <jan_g> bwm: can we get people to review this for the next telecon?

15:18:09 <jan_g> patrick, jeremy

15:18:09 * DanC is kinda surprised to see PatH in attendance; thought he was on holiday break

15:18:49 <jan_g> danc: so noted in telecon, just discussing the changes you suggested

15:19:22 <DanC> just to clarify: I don't think any of my review comments were of the "stop the press! don't publish in this form!" variety.

15:19:38 <jan_g> bwm: review the version put out for review, let pat integrate comments he gets back

15:19:57 <jan_g> danc - yes, we all know your stock answer to "should we publish?" :-)

15:20:37 <jan_g> comments within the group until jan 11th

15:20:47 <jan_g> jos: I've no basic objections to this, I just want to run my tests

15:20:57 <jan_g> I can't see any fundamental problems

15:21:07 <jan_g> ph: only fundamental math changes are deletions of vocab

15:21:47 <jan_g> jeremy: tech question: in webont, peter's saying theres a problem with the mt that doesn't work with daml

15:21:55 <jan_g> ph: we've been discussing this for some time

15:22:05 <jan_g> the basic problem with any rdf mt working with model

15:22:10 <dajobe> see differences in path's drafts at http://cvs.ilrt.org/cvsweb/redland/rdfcore/mt/

15:22:15 <jan_g> to do with the way daml encodes itself inside rdf

15:22:22 <jan_g> this breaks conformity with the mts

15:22:37 <jan_g> there's a "discussion" on webont about this at the moment

15:22:51 <jan_g> jeremy: so this is a daml/webont problem, not a core one?

15:23:01 <jan_g> ph: prbably a webont problem: webont = daml'

15:23:17 <jan_g> we're trying to relate daml to rdf in a more rational way, for webont

15:23:26 <jan_g> bwm: is there a proposal to publish now?

15:23:27 <jan_g> jos: yes

15:23:36 <jan_g> pat: seconded modulo namespace problem

15:23:44 <jan_g> ^^ pat=patrick

15:23:54 <jan_g> ph: I'd like to make the changes danc asked

15:24:13 <jan_g> daveb: apparently w3c are ok to publish tuesday

15:24:28 <jan_g> bwm: I'd still like this to be reviewed properly...

15:24:38 <jan_g> path: we're not under a great deal of time pressure, agreed

15:24:49 <jan_g> em, daveb: webont are waiting for this

15:25:00 <jan_g> ph: webont f2f on jan 14th

15:25:13 <jan_g> if this comes out now some people might be able to read it before then

15:25:36 <jan_g> em: if dan's comments are purely editorial in nature, then

15:25:45 <jan_g> ...don't pause to reconsider as a group.

15:26:07 <DanC> I'm not sure they're purely editorial. they're just not critical

15:26:08 <jan_g> bwm: it's not danc's comments i'm concerned about, I'd like a more rigorous review first, is all

15:26:29 <jan_g> ph: this hasn't really changed very much from the doc available for the better part of a month now

15:26:56 <jan_g> em: if we're just converging, then I'd say hold off and look for last call

15:27:08 <jan_g> ph: we're still waiting on dts, containers, reification

15:27:34 <jan_g> ps: is there a need to publish as a draft; can't we jsut give it to webont?

15:27:36 <jan_g> bwm: good idea

15:27:51 <jan_g> so how about: don't formally publish now, so flag the editor's version to webont

15:27:59 <jan_g> daveb: ... and copy to www-archive

15:28:09 <jan_g> bwm: is that decision clear?

15:28:32 <jan_g> jos, pats, jeremy reviewed

15:28:37 <jan_g> aim for a decision january telecon

15:28:43 <jan_g> (sergei arrives)

15:28:49 <jan_g> item 11: syntax wd

15:29:09 <jan_g> daveb: in summary, some typos, no showstoppers.

15:29:16 <jan_g> if the reviewers are on call: what do you think?

15:29:27 <jan_g> ps: a few points of clarification would help

15:29:39 <jan_g> daveb: I've sent a summary of my reply to your comments to the list

15:29:50 <jan_g> title change to rdf/xml syntax revised

15:29:55 <jan_g> that seem ok to everyone?

15:30:13 <jan_g> the otehr things are: need namespace clarification, and examples

15:30:18 <jan_g> ...this document contains no xml!

15:30:29 <jan_g> I expect to put out another draft in january

15:30:39 <jan_g> ps: we need the defeinition of "bnode" in the document

15:30:52 <jan_g> daveb: I use "blank node" when talking about the graph

15:31:05 <jan_g> that's to do with ntriples output, defined in the ntriples docu,menmt

15:31:08 <jan_g> I can say more about this

15:31:16 <jan_g> ps: at least, add a pointer to it

15:31:31 <jan_g> daveb: ok, uri qname and id generation/scope: I'm happy to add pointers to this.

15:31:56 <jan_g> bwm: have you reviewd this properly ps?

15:32:06 <jan_g> ps: unfortunately, not yet - not really until this evening

15:32:16 <jan_g> bwm: bearing in mind that a WD is a draft...

15:32:21 <jan_g> are there any showstoppers?

15:32:39 <jan_g> ps: no: most of my comments are trying to avoid too much discussion on the lists

15:32:46 <jan_g> I'm trying to avoid generating too many questions

15:32:58 <jan_g> bwm: I'm proposing daveb to make those changes, then publish

15:33:02 <jan_g> any objections?

15:33:05 <jan_g> no, no abstain

15:33:15 <jan_g> decided: daveb to make editorial changes, then publish

15:33:18 <jan_g> well done, dave!

15:33:22 <em> woohoo!

15:33:26 <jan_g> dasveb; summarises changes:

15:33:28 <jan_g> add an example

15:33:36 <jan_g> add uri id example

15:33:48 <jan_g> add a "changes" section

15:33:59 <jan_g> ...although it'll just say "everything in this document has changed"

15:34:07 <jan_g> em to act as team contact

15:34:28 <jan_g> em: need it between now and (at the latest) ... well, the sooner the better

15:34:33 <jan_g> I can do it as late as monday

15:34:44 <jan_g> daveb: i'll get it to you monday, the changes are smnall enough for that

15:34:55 <jan_g> action: daveb to get doc to em by monday

15:35:01 <jan_g> action: em: to drive publishing process

15:35:07 <jan_g> item 12, issues.

15:35:15 <jan_g> difference between id and about

15:35:31 <jan_g> proposal: that we agree that the wd resolves this issue

15:35:39 <jan_g> jeremy: i18n is orthogonal issue

15:35:49 <jan_g> we have test cases

15:36:07 <jan_g> do we resolve that the new wd resolve that id is (basically) shorthand for about

15:36:16 <jan_g> ^^ bwm putting the issue

15:36:34 <jan_g> dissent? no abstentions? no

15:36:37 <jan_g> resolved unanimously

15:37:24 <jan_g> jeremy: on test cases, internationalisation is a different issue, test1 error1 were approved

15:37:29 <jan_g> test2 waits for i18n issues

15:37:37 <jan_g> everyone ok with that? no dissent

15:37:48 <jan_g> item 13: not-id-and-resource-attr

15:38:02 <jan_g> daveb: jang produced these for the empty property element

15:38:12 <jan_g> (test cases already present)

15:38:19 <lilo> [Global Notice] Hi all. This is a semi-automated reminder. We are approaching our next release upgrade, to 1.0.31, a bug fix release. Please take some time today and through the weekend to use the test bed extensively. It's on irc.openprojects.net, port 9001. Use the channels, have conversations, do everything you do here only more so. It'll help ensure a smooth upgrade. bugs@openprojects.net for problems. This message will repeat. Thanks!

15:38:22 <jan_g> we can change this, but this is a "take a view" case

15:38:41 <jan_g> we can change the idea, but the WD gives a view and it never previously had a reading

15:38:47 <jan_g> s/reading/meaning

15:39:25 <jan_g> jeremy: production 4 of 6.12 of the old grammar, rdf id or resource and para 232 had label on object semantics, not label on reification

15:39:34 <jan_g> bwm: to clarify....

15:39:52 <jan_g> on a property element, you can stick a rdf:id attr, which says "reify this statement, here's the id of the reified statement"

15:40:06 <jan_g> daveb: there is a clarification in the syntax document

15:40:46 <jan_g> bwm: volounteers to look at this issue specificalyl wrt wd

15:40:51 <jan_g> to close this next time around>

15:40:58 <jan_g> jeremy, jang to do this.

15:41:10 <jan_g> action: jeremy, jang to look at this issue closely for next telecon

15:41:13 <jan_g> item 14

15:41:20 <jan_g> review test cases for para 196

15:41:33 <jan_g> daveb: test cases created

15:41:39 <jan_g> jos: I've reviewed this.

15:42:00 <jan_g> action: jeremy to review this for next time too

15:42:08 <jan_g> item 15: datatypes

15:42:16 <jan_g> sergei: we're doing fine.

15:42:24 <jan_g> ...I've the second draft posted to the list

15:42:33 <jan_g> bwm, ps gave very good comments I've still to address

15:42:41 <jan_g> there are still a couple of issues I'd like to be resolved

15:43:01 <jan_g> in particularly, I'd like to ensure that people who raised relevant issues gets them resolved as they like them

15:43:09 <jan_g> I'd like mike's input on that

15:43:18 <dajobe> my para 196 message: http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Dec/0009.html

15:43:33 <jan_g> miked: I've an email to send out on this, next two working days

15:43:52 <jan_g> sergei: we're looking for community feedback - what's the best way to proceed?

15:44:07 <jan_g> bwm: my take: once the doc is in a state we're happy with, either with a single or multiple proposals

15:44:12 <jan_g> we'll circulate and seek feedback

15:44:25 <jan_g> I'd like to add pressure here to get this done as early as poss in the new year

15:44:35 <jan_g> which mneans pressure on anyone who should be helping with this

15:44:58 <jan_g> sergei: I'd like to ask anyone who posted questions on this to the list to ensure that the document addresses this

15:45:10 <jan_g> sergei: are P, U going to make it into this document?

15:45:29 <jan_g> ph: wrt the P writeup - it's not going to appear anywhere else. If I get it done, you'll get it

15:45:36 <jan_g> I'll try to get it done for next tuesday or so

15:45:44 <jan_g> sergei: then I'll work it into the document

15:46:11 <jan_g> ps: I'm also going to help out here

15:46:57 <jan_g> I've tried to show that P/S/DAML/U are effectively just idiomatic versions of the same thing

15:47:13 <jan_g> sergei: so I'm asking ph, ps for help.

15:47:22 <jan_g> I'm asking for a "how to" for each of those proposals

15:47:44 <jan_g> ph: should we be considering literal-is-xml-structure?

15:48:02 <jan_g> jeremy: no, don't worry about that. I own the issue, it's really a separate issue

15:48:17 <jan_g> and I think we looked at it, we decided to look at datatyping first

15:48:21 <jan_g> then come back to it

15:48:31 <jan_g> ie, DT influences it, not the other way around

15:48:39 <jan_g> ph: as long as you're sure it's not going to be a landmine

15:49:04 <jan_g> jeremy: no, we used canonical XMl spec for this, it looks like it might come together

15:49:22 <jan_g> jos: sergei, you seem to have V and R disjoint

15:49:33 <jan_g> is this a conflict with the mt? is that an issue?

15:49:47 <jan_g> ph: the mt says they don't have to be disjoint

15:50:05 <jan_g> sergei: if you think that's a problem... how can that be a problem?

15:50:28 <jan_g> sergei, v and R are not disjoint.

15:50:36 <jan_g> in the u scheme, that's exactly what you use

15:50:49 <jan_g> jos: peter says they're clearly disjoint, that's a problem

15:50:57 <jan_g> ph: daml strictly segregates them too

15:51:06 <jan_g> jos: so you don't see any issues here then?

15:51:22 <jan_g> ph: don't see issues for rdf; rdf is a bit more semantically liberal than daml

15:51:48 <jan_g> that's ok; you expect an extension language to impose more restrictions than the language it extends

15:52:06 <jan_g> bwm: what about this notion that a literal be identified by a pair

15:52:13 <jan_g> (dt, representation)

15:52:33 <jan_g> ps: I'm not proposing this exactly: the idea is that a given value is identified uniquely by

15:52:39 <jan_g> (lexical form, dt identifier)

15:52:52 <jan_g> and in the graph there are various idioms to say what that pairing is

15:53:18 <jan_g> (elabourates on how daml, p, s do this)

15:53:36 <jan_g> ph: this is a purely conceptual sense of pairing?

15:53:45 <jan_g> this is the idea behind what was called p++

15:54:14 <jan_g> ps: we're not trying to express the mapping itself; instead to infer it

15:54:52 <jan_g> bwm: so then the question to you and pat:

15:54:59 <jan_g> why not just make that pairt explicit in the graph?

15:55:07 <jan_g> ps: then you have to adopt a single idiom

15:55:13 <jan_g> bwm: no you don't!

15:55:36 <jan_g> ph: there are people who are interested in dting a literal by imposing constraints on something else, eg the range of a property

15:56:09 <jan_g> bwm: it struck me that we can unify all the proposals...

15:56:17 <dajobe> bwm's message http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Dec/0080.html

15:56:29 <jan_g> ps: we don't want to adopt a single rep in the graph because there are pros and cons to all proposals

15:57:07 <jan_g> bwm: I agree re the pros and cons,

15:57:13 <jan_g> (we take this to email)

15:57:43 <jan_g> sergei: i refer you to my section "type theory"

15:58:10 <dajobe> sergei's doc http://www-db.stanford.edu/~melnik/rdf/datatyping/

15:59:08 <jan_g> ph: are we looking at the need to bring together lexical form and datatype

15:59:19 <jan_g> all the propsals seem to do the same thing

15:59:50 <jan_g> ps: the biggest distinction is : does the graph contain the value itself or something that lets us infer it

15:59:56 <jan_g> bwm: closing meeting...

16:00:01 <jan_g> with "jolly holidays" all round

16:00:21 <jan_g> thanks for efforts from everyone , look forward to seeing you in the new year

16:00:34 <jan_g> action: aaron to sing the golden triples for us all

16:00:38 <jan_g> frank: a triple for you all

16:00:43 <jan_g> "ho" "ho" "ho" .

16:01:40 <jan_g> meeting closes; we natter for a while.

16:01:51 <jan_g> xml2001 highlight: doing topic maps in rdf

16:02:14 <dajobe> see http://www.xmlconference.net/xmlusa/