Semantic Web Interest Group IRC Chat Logs for 2002-10-21

02:07:01 <DanConn> B:ah; it's already been chumped.

02:07:02 <dc_rdfig> Added comment B5.

03:14:19 <mnot> anyone aware of a SF Bay area RDF/SW SIG?

03:44:38 <_Maniac> _Maniac is now known as Maniac

06:14:10 <Grantbow> hi mont, I know some people here but no groups off the top of my head.

06:14:24 <Grantbow> mnot :-)

06:55:29 <q|gone> q|gone is now known as qmacro

12:30:11 <JibberJim> C: Something like [http://jibbering.com/rdfsvg/scene.1?Cat;Person;Dog;llama;rock_hyrax|this] danbri ??

12:30:12 <dc_rdfig> Added comment C3.

13:51:37 * danbri fiddles with coldsync (Palm/Visor sync tool) under cygwin; loses again

13:52:00 <dajobe> DanConn: you around?

13:52:14 <DanConn> depends.

13:52:33 <dajobe> want to ask about your ntriples datatype uri preferences

13:52:48 <DanConn> hmm...

13:53:05 <dajobe> you said you are changing n3

13:53:15 <DanConn> I meant to say TimBL has changed N3

13:53:20 <dajobe> ok

13:53:24 <dajobe> so "foo"<uri>

13:53:29 <dajobe> or "foo"^^<uri>

13:53:39 <dajobe> where '^^' might be some other token(s)

13:53:54 <dajobe> but I want to make sure that when langstrings are used, it works

13:54:04 <dajobe> i.e. "foo"-en-US<uri>

13:54:11 <dajobe> or change to "foo"^en-US^<uri> ?

13:54:39 <DanConn> what's new in N3 is path syntax: :fred^:brother , i.e. fred's brother, and :bob!:father, i.e. somebody bob is father to.

13:55:10 <DanConn> TimBL's idea of how languages should work is "chat"!:fr , i.e. something that in french is "chat"

13:55:27 <dajobe> is that an abbreviation for more triples?

13:55:44 <DanConn> yes, it's short for [ :fr "chat" ]

13:56:19 <dajobe> that's not how langstrings work, oh well

13:57:56 <DanConn> I think "foo"^^<int> was TimBL's way of sorta giving in and saying OK, if we're going to compress triples, cwm can kludge that.

13:58:18 <dajobe> that syntax is fine

13:58:52 <DanConn> so I could maybe sell him on "foo"!!en-US^^<uri> analagously.

13:59:24 <dajobe> ok

14:00:06 <dajobe> I actually never liked "foo"-en-US; the '-' was used for two things

14:00:33 <dajobe> I'm going to change to the ^^ form

14:00:47 <DanConn> cool.

14:00:50 <dajobe> but I'd like at least some positive noises about !! before doing that

14:01:15 <DanConn> what other delims are you considering? - , I guess/

14:01:16 <DanConn> ?

14:01:44 <dajobe> hard to pick one n3 doesn't already use in some other way

14:02:31 <DanConn> quite.

14:02:33 <DanConn> hence !!

14:05:04 <dajobe> another parsing thing

14:05:11 <dajobe> is it more important to see the ^^<uri> first

14:05:20 <dajobe> i.e. "foo"^^<uri>!!en-US

14:05:30 <dajobe> rather than "foo"!!en-US^^<uri>

14:05:35 <dajobe> or don't care?

14:06:51 <DanConn> depends on which one you mean...

14:07:02 <dajobe> well, you prefered not to have <uri>"foo"

14:07:04 <DanConn> I suppose "foo"^^<uri>!!en-US , but I'm not sure.

14:07:41 <dajobe> I'll go with "foo"!!en-US^^<uri> since it makes the grammar simpler

14:08:01 <Snakker> Snakker is now known as libby

14:08:29 <DanConn> hmm... I think I'm actually confused... I think I've been misusing !!

14:08:50 <dajobe> looking at http://www.w3.org/TR/rdf-testcases/#ntrip_grammar

14:08:55 <DanConn> "10"^^<int> means [ :int "10"]

14:09:03 <dajobe> er, no

14:09:08 <dajobe> not for n-triples

14:09:12 <DanConn> sorry...

14:09:44 <dajobe> each line is one triple (doh) and the !! and ^^ being used here as separating parts of the literals

14:09:51 <DanConn> "10"^<int> means [ :int "10"]... I think languages sorta go the same way (which suggests ^^ for them too :-{) not the opposite way (which would suggest !!)

14:11:23 <dajobe> that's a bit tricky to parse

14:11:32 <dajobe> so "foo"^^en-US^^<uri>

14:11:54 <dajobe> or "foo"^^<uri> with no lang

14:12:02 <dajobe> possible I guess

14:12:39 * DanConn tentatively suggests...

14:12:45 <DanConn> "foo"^^^en-US^^<uri>.

14:12:52 <dajobe> lol

14:13:24 <dajobe> suddenly "foo"-en-US^^<uri> looks more appealing

14:13:39 * dajobe thinks

14:14:08 * DanConn wishes he could think clearly.

14:14:21 <dajobe> does n3 allow @ outside the first column?

14:14:34 <dajobe> i.e would "foo"@en-US^^<uri> be possible

14:15:04 <DanConn> columns don't matter... hmm... yes, I suppose that's as possible as any other.

14:15:32 <DanConn> the actual use case for using both a datatype and a language suggests "frog"^^<animal>-en-US.

14:15:40 <DanConn> hmm... no...

14:15:50 <DanConn> gee, I can't get my head around the use case for both.

14:16:27 <dajobe> I'm assuming there is one, just trying to fashion a "neat"-enough syntax

14:33:41 <dajobe> logger, here

14:33:41 <dajobe> See http://ilrt.org/discovery/chatlogs/rdfig/2002-10-21#T14-33-41

14:44:02 <danbri> dajobe, got a sec right now?

14:44:14 <danbri> Am being asked about mandatoryness of <rdf:RDF> wrapper...

14:44:15 <dajobe> yes

14:44:43 <dajobe> that's answering - I've got a sec, and yes it is manditory

14:45:01 <danbri> what's the current status of that business? I recall you saying it mandatory. Got a pointer into spec (/TR/-published ideally...)

14:45:14 <dajobe> yes

14:45:27 <DanConn> it's not mandatory in the existing REC, right dajobe? do you remember when the RDFCore WG decided to make it mandatory?

14:45:29 <dajobe> see http://www.w3.org/TR/rdf-syntax-grammar/#start

14:45:36 <dajobe> [[f the RDF/XML is a standalone XML content, then the grammar starts with Root Node doc]]

14:45:44 <dajobe> and doc is rdf:RDF

14:46:00 <DanConn> hmm... that's not an unqualified "yes, it's mandatory"

14:46:03 <danbri> logger, pointer?

14:46:03 <danbri> See http://ilrt.org/discovery/chatlogs/rdfig/2002-10-21#T14-46-03

14:46:05 <dajobe> (that sentence needs better grammar, changed in)

14:46:13 <dajobe> (editors' draft)

14:46:55 <dajobe> if the doc is rdf/xml, it is mandatory

14:47:16 <danbri> Do we have text explaining why this change was made?

14:47:24 <dajobe> "doc is rdf/xml" to me == mime type application/rdf+xml == .rdf file

14:47:34 <dajobe> this isn't a change

14:47:34 <danbri> ah, so embedded in SVG or whatever, our syntax can be used without RDF:RDF wrapper?

14:47:37 <dajobe> yes

14:47:40 <dajobe> but not SVG

14:47:46 <danbri> ok, thanks. SVG was an example...

14:47:47 <dajobe> since it allows any metadata content

14:48:01 <dajobe> you better form a proper question

14:48:13 <dajobe> is this only about SVG?

14:48:27 <JibberJim> in SVG can I have <metadata><rdf:Description> ... </></> ?

14:48:37 <dajobe> no

14:48:48 <dajobe> since you can't know it is rdf content by the context

14:48:58 <dajobe> .. the context being the <metadata> containing element

14:49:20 <dajobe> see http://www.w3.org/TR/SVG/metadata.html

14:49:27 <dajobe> [[Other XML-compatible metadata languages, including ones not based on RDF, can be used also.)]]

14:49:36 <dajobe> and they would have a different root element

14:50:32 <JibberJim> So I could do:

14:50:46 <JibberJim> <metadata><jims:stuff><rdf:Description ....

14:51:05 <dajobe> I would have to read the spec for jims:stuff to tell that

14:51:35 * danbri nods

14:52:18 <JibberJim> hmm... /me thinks it would be best to require it everywhere rather than try to make agents intelligent enough to read my spec...

14:52:32 <dajobe> that's a possiblity

14:52:44 <dajobe> and if somebody said that to the rdf-comments list, that would be quotable

14:55:05 <JibberJim> trying to keep up with rdf-ig is hard enough...

14:58:24 <sbp> logger, pointer

14:58:24 <sbp> See http://ilrt.org/discovery/chatlogs/rdfig/2002-10-21#T14-58-24

14:59:10 <dajobe> JibberJim: I said rdf-comments, not rdf-core

15:06:58 <JibberJim> it's still more mail to keep up with!

15:07:33 <dajobe> you don't need to do that

15:07:46 <dajobe> it is the official rdf comments list; if you want to say something to rdf core, use it

15:07:55 <bijan> I.e., you don't have to subscribe.

15:08:05 <bijan> I've sent comments to it.

15:08:09 <dajobe> see the status section ofa ll the Working drafts

15:08:22 * JibberJim always feels obligated to subscribe to a list before posting... RFC 1855 and all that...

15:08:25 <danbri> dc_rdfig:view

15:08:26 <dc_rdfig> A: Par's pictures from the SWAD-E calendar workshop (http://lannero.com/p/bristol2002/index.html)

15:08:27 <dc_rdfig> B: PIM project using RDF (http://blogs.osafoundation.org/mitch/000007.html)

15:08:28 <dc_rdfig> C: Parallax - Cel Animation using SVG, by Max Froumentin (http://people.w3.org/maxf/Parallax/)

15:08:50 <DanConn> ew... there's an RFC that says you have to subscribe before posting? Who wrote it? what's its status?

15:09:08 <danbri> C:Yes, something very like that! :)

15:09:09 <dc_rdfig> Added comment C4.

15:09:16 <bijan> Seems totally inapplicable

15:09:47 <bijan> E.g., the first thing it recommends for one-many communication is to "read the list/group for one to two months before posting"

15:09:49 <bijan> That's silly.

15:09:59 <bijan> rdf-comments *solicits* your comments.

15:10:22 <JibberJim> true, but I still think you should check recent archives to check you're not just repeating comments...

15:10:30 <bijan> Sure.

15:10:31 <dajobe> lol

15:11:00 <bijan> *But* it's not the same sort of annoyance that it would be in a "regular" mailing list.

15:11:49 <bijan> I.e., I check so I don't have the bother of writing up something someone else did, not because of fear of annoying the inhabitents. Repeated information, in a public comments list, does add some marginal value.

15:12:33 <dajobe> I don't think we care too much about repeats

15:12:44 <dajobe> if multiple people give the same comment, or ask for same thing, that's useful

15:13:09 <JibberJim> danbri - did you see how the svg doc worked?

15:13:19 <bijan> Hmm. And anyway, I'm not spying where it says "subscribe before you send".

15:13:40 * JibberJim thinks read before you send == subscribe before you send.

15:13:43 <danbri> I saw sorta how Max's version worked; not view-src'd yours yet

15:13:47 <bijan> Why?

15:13:50 <bijan> When you have archives?

15:14:00 <JibberJim> I meant just my scene - how to pick the objects it shows.

15:14:02 <bijan> Plus it's to "help get an understanding of the culture of the group"

15:14:13 <bijan> Here you have a direct solicitation from a member of the group.

15:14:20 <danbri> No; I guessed it was url hacking

15:14:34 <bijan> Which suggests strongly that your post *would* be in tune with the culture of the group :)

15:14:44 <danbri> could do with setting an opacity on the fence/hedge thing so can see what's lurking...?

15:15:04 <JibberJim> archives==subscribe - invest time in reading the group, but yeah I might post my comment.

15:15:25 <bijan> Indeed, as it's the main way of officially giving comments about wg matters, I'd say that the need to subscribe/read up is almost completely nullified.

15:15:26 <JibberJim> I got rid of MaxF's purple bunny, I hope he doesn't mind...

15:15:29 * bijan stops arguing :)

15:16:31 <libby> aw, I liked the bunny

15:17:03 <JibberJim> if you take a photo of it, then you could annotate the image as containing a wordnet:Bunny...

15:17:40 <danbri> its a tease of a demo, can't really see the things so well

15:18:19 <JibberJim> reload it might pick a different person!

15:18:30 <JibberJim> or rock_hyrax

15:32:45 * maxf waves

15:33:09 <JibberJim> hi maxf - nice parallax thingy!

15:33:15 <maxf> nice app for Parallax JibberJim!

15:34:08 <JibberJim> well, I'm not so sure, it's just encouraging danbri to search for penises again...

15:34:35 <maxf> :-)

15:35:24 <libby> hey maxf!

15:36:27 * maxf looks for the dog in http://jibbering.com/rdfsvg/scene.1?Cat;Person;Dog;llama;rock_hyrax

15:36:39 <maxf> hi libby.

15:37:46 <JibberJim> it's there maxf, but I think it's lower than the others. scene.2.1 might be better as you can see through the scenery.

15:38:27 <maxf> scene.2.1?

15:39:21 <JibberJim> - http://jibbering.com/rdfsvg/scene.2.1?Cat;Person;Dog;llama;rock_hyrax

15:39:23 <maxf> Ah, I see.

15:39:39 <JibberJim> (my odd versioning method...)

16:36:48 * danbri waves to maxf; nice work on the parallax thingumy

16:37:38 <maxf> hi danbri. Thanks. You gave me the idea.

16:38:24 * maxf thinks he should go to Bristol more often ;).

16:38:48 <danbri> I gave you the idea by clicking on goofy Flash stuff (but yes, come to Bristol more often :)

16:40:17 <danbri> Jim, looking at 2.1 version... nice that we can see thru the hedge, wondering if you might position the creatures higher slightly in the scene, as still a bit hard to see them

16:40:36 * JibberJim doesn't position the creatures in the scene... that's tough.

16:40:50 <danbri> yeah, its the iamges they're embedded in.

16:41:09 <JibberJim> I could do it client-side I guess...

16:41:21 <danbri> we'd need to do some preprocessing i think to zoom in on the box around each path within its larger context

16:41:24 <danbri> if that makese sense

16:41:54 <libby> it's v cool, maxf, jim; yeah, come to bristol more often maxf

16:42:35 <maxf> heh, thanks.

16:43:11 <JibberJim> I've done that with http://jibbering.com/rdf/foafwhoss4.1 but that's client-side.

16:45:35 <DanConn> libby, re calendaring workshop report, any idea when?

16:46:06 * DanConn has low expecatations on presentations, but high hopes for timeliness

16:46:24 <libby> I'm hoping to draft something this week, but have to see how it goes a bit

16:46:26 <DanConn> i.e. a front-side-of-one-page deely this week would be fine by me.

16:46:53 <libby> yeah, be nice to get the actions and main points summarised

16:47:05 <DanConn> are there any particular constraints on the report from the funders?

16:47:20 <DanConn> hmm... actions? did anybody accept actions?

16:47:32 <libby> yeah, month 8 which is december end, but i want to keep up momentum

16:47:55 <libby> perhaps not actions....things to do maybe? this is why I need to do this soon

16:48:11 <DanConn> any constraints from the funders on what the report should look like?

16:48:54 * DanConn did do a tiny bit on http://www.w3.org/2002/10/calendar , btw.

16:49:09 <libby> not very strong ones I don;t think, bar some metadata type stuff. Our EC project officer always said he'd much rather it was useful topeople than anything else

16:49:22 <DanConn> good to know.

16:49:30 * DanConn wishes libby luck in finding time to assemble it soon

16:49:34 <libby> yeah, he's an excellent bloke

16:49:41 <libby> cheers DanC

16:50:25 <libby> hm http://www.w3.org/2002/10/calendar is forbidden - is that right DanC?

16:50:33 <DanConn> whoa! somebody else in the team used http://www.w3.org/2002/10/calendar.html for something else!

16:50:55 <danbri> bummer!

16:51:01 <DanConn> I was just gonna chacl http://www.w3.org/2002/10/calendar.rdf world-readable, but I see http://www.w3.org/2002/10/calendar is already in use!

16:51:06 <danbri> lets swipe /11/

16:51:31 <DanConn> (this is why namespace names shouldn't 404; by making sure folks get a 200 response, you discover collissions like this)

16:52:36 <DanConn> in the tiny bit of work on calendar.rdf, I discovered I really need a directory.

16:53:09 <DanConn> I'd prefer 2002/10cal/vocab along with 2002/10cal/docs and 2002/10cal/Makefile and 2002/10cal/test and such

16:53:42 <DanConn> er... and a tutorial/howto is a pretty high priority.

16:54:42 <libby> how to what exactly?

16:54:44 <danbri> I want to make something that generates iCal files for mac and moz fans...

16:59:18 <DanConn> how to use the schema in a few known and useful ways.

16:59:44 <libby> ok, yep

22:41:32 <DanConn>http://www.w3.org/Talks/2002/10/16-sw/slide1-0.html

22:41:32 <dc_rdfig> A: http://www.w3.org/Talks/2002/10/16-sw/slide1-0.html from DanConn

22:41:40 <DanConn> A:|Weaving Meaning: The Semantic Web

22:41:40 <dc_rdfig> Titled item A.

22:41:49 <DanConn> A:Eric Miller, W3C Semantic Web Activity Lead

22:41:49 <dc_rdfig> Added comment A1.

22:44:09 <DanConn> A:[DC-2002|http://www.bncf.net/dc2002/index.html] October 13-17, 2002 Florence, Italy

22:44:09 <dc_rdfig> Added comment A2.

23:33:51 <em_lap>http://cio-asia.com/pcio.nsf/unidlookup/2A3AA4D4EFA7D03448256C460027658A?OpenDocument

23:33:51 <dc_rdfig> B: http://cio-asia.com/pcio.nsf/unidlookup/2A3AA4D4EFA7D03448256C460027658A?OpenDocument from em_lap

23:34:13 <em_lap> B:|Doing It with Meaning

23:34:13 <dc_rdfig> Titled item B.

23:34:37 <em_lap> B:Doing It with Meaning

23:34:38 <dc_rdfig> Added comment B1.

23:34:38 <em_lap> : Semantics tools promise a world of universally compatible data,

23:34:38 <em_lap> By John Edwards

23:35:02 <em_lap> B:Doing It with Meaning - Semantics tools promise a world of universally compatible data, by John Edwards

23:35:02 <dc_rdfig> Added comment B2.

23:36:17 * danbri wonders what 'universally compatible data' amounts to; waits for page to load

23:37:05 <danbri> not loading, mystery can wait 'til another day

23:37:19 * em_lap thinks the web is a tease sometimes...

23:38:51 <danbri>http://www-kasm.nii.ac.jp/~hamasaki/acksh/develop/index.html

23:38:51 <dc_rdfig> C: http://www-kasm.nii.ac.jp/~hamasaki/acksh/develop/index.html from danbri

23:39:07 <danbri> C:|Acknowledge Shake Hands System (Acksh)

23:39:07 <dc_rdfig> Titled item C.

23:39:31 <danbri> C:See also [http://www-kasm.nii.ac.jp/~hamasaki/acksh/develop/list.html|user list]

23:39:32 <dc_rdfig> Added comment C1.

23:41:04 * em_lap thinks 802.11 connection is failing.... heads back to desktop and dedicated connection to continue work...