Semantic Web Interest Group IRC Chat Logs for 2004-01-13

00:31:58 <GregElin> hey bryce

00:32:29 <bryce> hi. my wife has the phone with her right now, so for the time being, it's chat only.

00:32:51 <GregElin> I'm on phone, too.

02:21:56 <DanCon> hmm... "IBM vs. Content Chaos" on slashdot...

02:27:02 <DanCon>http://www.spectrum.ieee.org/WEBONLY/publicfeature/jan04/0104comp1.html

02:27:03 <dc_rdfig> A: http://www.spectrum.ieee.org/WEBONLY/publicfeature/jan04/0104comp1.html from DanCon

02:27:13 <DanCon> A:|A Fountain of Knowledge

02:27:13 <dc_rdfig> Titled item A.

02:27:25 <DanCon> A:By Stephen Cass in IEEE spectrum

02:27:25 <dc_rdfig> Added comment A1.

02:28:19 <DanCon> A:mentions XML and TAP by name, but not RDF

02:28:19 <dc_rdfig> Added comment A2.

02:28:31 <DanCon> A:(and expands XML incorrectly)

02:28:31 <dc_rdfig> Added comment A3.

02:29:05 <DanCon> A:as seen in [http://slashdot.org/articles/04/01/12/142247.shtml?tid=126&tid=136&tid=187&tid=95|a slashdot article]

02:29:05 <dc_rdfig> Added comment A4.

06:19:19 <dmwaters> {global notice} hi all! it appears that we lost one of our main us hubs. we're currently looking into the situation.

06:38:23 <esigler> esigler is now known as esigler_away

07:37:30 <tav> tav is now known as tav|sleep

08:24:18 <arnarl|away> arnarl|away is now known as arnarl

10:33:44 <deelan> redland sub-models make me mind to blow

10:34:06 <dajobe> you mean contexts

10:34:11 <dajobe> sub-models aren't implemented

10:34:58 <deelan> ah. i've read the RDF paper, they are still in the TODO list then

10:35:37 <deelan> i was looking to that document to get some inspiration

10:40:00 <dajobe> that paper's from 2000

10:40:20 <dajobe> more here http://www.redland.opensource.ac.uk/notes/contexts.html

10:40:43 <dajobe> the perl & python api docs have more examples, and there are all the example codes, for every language I support

10:40:52 * deelan is printing contexts.html

10:44:19 <deelan> thx

11:09:36 <deelan> quick question about RSS 1.0: it's not clear to me what's the preferred way to describe comments to a given RSS entry. mod_threading or mod_annonation?

11:10:08 <dajobe> dunno

11:11:56 <deelan> mm, threading seems the most appropriate: http://web.resource.org/rss/1.0/modules/threading/

11:12:36 <deelan> comments are "children" of a given entry

11:16:56 <deelan> ...but i don't like to have to express a rdf:Seq for those. ;(

11:17:26 <deelan> after all i'm supposed to sort by date the comments

11:17:59 <dajobe> I can't say I've seen that used much but I've not been following it

11:18:29 <deelan> k. i will do some searching then

11:19:14 <deelan> he: http://philringnalda.com/ns/blogcomments/

11:21:25 <deelan> the thing is that i see a comment more or less like another blog post, no reason to differentiate them, only one "depends" from another :)

11:26:40 * deelan is thinking that libby may known the answer...

11:27:54 <libby> sorry deelan, I'm not sure

11:28:10 <deelan> ah :(

11:28:20 <libby> maybe ask on the list, see if opne of the aggregators will tell you what's most popular? (rss-dev I mean)

11:30:47 <deelan> k. i'm reading the DC terms... there's a DC.relation, but i cannot find out if a can specify an URI or just Literal.

11:31:15 <libby> does it not say in the dc module for rss?

11:32:35 <deelan> no. it just list DC.relation as a possible term. nothing else

11:33:41 <deelan> i'm not so good to read RDF/XML: http://dublincore.org/2003/03/24/dces#

11:40:24 <dajobe> why don't you try http://www.schemaweb.info/schema/SchemaDetails.aspx?id=36

11:40:41 <dajobe> and nearby

11:41:39 <deelan> cool. didn't known about that. thx

11:45:33 <D[a]vey> is there anything stopping me from publishing my FOAF as N3 or N-Triples?

11:47:38 <dajobe> portability

11:47:42 <dajobe> internationalisation

11:47:54 <dajobe> so don't

11:48:37 <dajobe> the rdf transfer format is rdf/xml

11:49:27 * deelan finds '<dc:relation rdf:resource="http://www.twoshortplanks.com/bots/basic/"/>'

11:51:12 <deelan> N-triples doesn't indicate a declared/default encoding?

11:52:15 <dajobe> it allows only 1 encoding - ASCII

11:52:20 <deelan> argh

11:52:21 <dajobe> since i wrote the spec

11:52:23 <deelan> too bad

11:52:31 <deelan> :)

11:52:32 <dajobe> it was *not* designed as a general user format

11:52:38 <deelan> i see.

11:52:44 <dajobe> see turtle for that, my ntriples++ format

11:53:05 <dajobe> turtle http://www.ilrt.bris.ac.uk/discovery/2003/11/ntriplesplus/

11:53:06 <deelan> of course, i read something about turtle lately.

11:53:26 <dajobe> that doc needs a lot of reformatting, adding of examples. Although there are some in the paper linked

11:53:30 <deelan> wheee: Unicode strings

11:53:39 <dajobe> utf8 strings

11:55:01 <deelan> i'm out for lunch for a while. see ya later.

11:58:04 <swh_home> swh_home is now known as swh

12:15:30 * libby plays with cwm, tries http://esw.w3.org/topic/RdfCalendarSchema

12:15:42 <libby> " cwm: Unknown option: -filter"

12:25:43 <libby> d'oh, my mistake

12:49:10 <libby> heya jibberjim

12:50:48 <JibberJim> hey libby

12:52:10 <libby> BLURB: WWW2004 photos meeting, 2004-01-13, 1600 GMT, #rdfig

12:52:11 <dc_rdfig> B: WWW2004 photos meeting, 2004-01-13, 1600 GMT, #rdfig from libby

12:53:49 <libby> B:[see agenda|http://lists.w3.org/Archives/Public/www-archive/2004Jan/0021.html]

12:53:49 <dc_rdfig> Added comment B1.

13:00:25 <libby> jibberjim, what's your main image schema file - http://jibbering.com/2002/3/svg/# ?

13:01:02 <JibberJim> er yeah, but then the one talked about in discussion/ for other stuff, but that's the one in actual use

13:01:22 <libby> is that where you might have put depicts?

13:02:13 <libby> ah, depiction is in there

13:02:13 <JibberJim> nah, I don't think I put depicts anywhere, I put it in foaf...

13:02:35 * JibberJim gotta go, see privmsg in about 15 mins....

13:02:40 <libby> cool, ta

13:20:29 <karlcow>http://autrans.crao.net/albums/MousseAutrans2004-01-07/HormonesRDF02.sized.jpg

13:20:30 <dc_rdfig> C: http://autrans.crao.net/albums/MousseAutrans2004-01-07/HormonesRDF02.sized.jpg from karlcow

13:20:50 <karlcow> C:|RDF Humour.

13:20:51 <dc_rdfig> Titled item C.

13:21:15 <karlcow> C: The man: I declare my hormones in RDF for you

13:21:15 <dc_rdfig> Added comment C1.

13:21:28 <deelan> LOL!

13:21:30 <karlcow> C: The woman: (very romantic)

13:21:31 <dc_rdfig> Added comment C2.

13:48:59 * mmealling wonders if there are other examples of RDF humor (besides that XML porn one)

13:49:24 <dajobe> google for reality distortion field (re apple) for unrelated humour

13:51:07 <mmealling> so are you saying there's a reality distortion field around this channel?

13:51:11 <mmealling> ;-)

13:51:31 <dajobe> or refuse derived fuel, one of those

13:53:06 <mmealling> I'm just looking for more xml geek related shwag for my cafepress shop: http://cafeshops.com/rocketforge/223887

13:59:16 <karlcow> mmealling: I have another one in french I have already posted here in the past :)

13:59:38 <karlcow> -> http://www.fixedbyfix.com/pour/Autrans2004/WebSemantique/metadata01.jpg

14:00:11 <karlcow> It says... "It's not a graffiti (tag in french), It's Metadata"

14:00:43 <dajobe> a new wiki topic I feel

14:07:35 <dajobe> let me drag things on topic then

14:07:37 <dajobe>http://www.w3.org/2001/sw/Europe/reports/dev_workshop_report_4/

14:07:38 <dc_rdfig> D: http://www.w3.org/2001/sw/Europe/reports/dev_workshop_report_4/ from dajobe

14:07:44 <dajobe> D:|SWAD-Europe Deliverable 3.11: Developer Workshop Report 4 - Workshop on Semantic Web Storage and Retrieval

14:07:44 <dc_rdfig> Titled item D.

14:08:20 <dajobe> D:detailed report (says me, the author) on the workshop we had 13-14 Nov 2003 on semweb storage and retrieval, lots of summaries

14:08:20 <dc_rdfig> Added comment D1.

14:13:55 <mdupont> mdupont is now known as md-afk-30m

14:14:03 <nick_kew> nick_kew is now known as niq

14:14:09 <D[a]vey> D[a]vey is now known as Davey

14:24:05 <evlist> hi karlcow. back home? that was a pleasure to see the interest for the semantic web last week in Autrans!

14:27:08 <karlcow> yes I'm back

14:27:16 <karlcow> It was really cool

14:27:32 <karlcow> and there are a lot of things to do now.

14:28:26 <evlist> yes. I find it encouraging that web developers get interested in the sem web. that seems natural but has not always been the case!

14:36:39 * mmealling chuckles at the comments in the slashdot article on VeriSign and RFID....

14:46:08 <tav|sleep> tav|sleep is now known as tav

14:59:38 <libby> heya nmg

14:59:45 <nmg> hi there

15:04:46 <libby>http://kanzaki.com/docs/sw/img-annotator.html

15:04:46 <dc_rdfig> E: http://kanzaki.com/docs/sw/img-annotator.html from libby

15:05:14 <libby> E:|Image Annotator by Masahide Kanzaki

15:05:15 <dc_rdfig> Titled item E.

15:05:26 <libby> E:uses javascript

15:05:26 <dc_rdfig> Added comment E1.

15:06:03 <libby> E:see also [http://kanzaki.com/works/2003/imagedesc/yakitoriya.rdf|an rdf iamge file rendered using css/xslt]

15:06:03 <dc_rdfig> Added comment E2.

15:08:10 <Davey> libby: thats AWESOME

15:08:28 <libby> isn't it great? :)

15:08:48 <Davey> yes

15:09:55 <bryce> iamge --> image (can't remember how to correct spelling on scratchpad, sorry)

15:10:18 <libby> oops, ta

15:10:29 <libby> E2:see also [http://kanzaki.com/works/2003/imagedesc/yakitoriya.rdf|an rdf image file rendered using css/xslt]

15:10:30 <dc_rdfig> Replaced comment E2.

15:10:45 <bryce> it's similar to fotonotes by gregElin

15:10:54 <libby> yes indeed

15:10:59 <libby> works differently though

15:11:13 <libby> (and is RDF-centric)

15:14:13 <deelan> yakitoriya, very nice.

15:14:28 <mmealling> speaking of web services, I keep wondering how I can sanely use RDF with the EPC Network without scaring the bejesus out of hte likes of walmart.

15:14:59 <mmealling> anyone aware of any EDI related RDF work?

15:37:09 <md-afk-30m> md-afk-30m is now known as mdupont

15:42:39 <phantoz> Any RDF Schema experts out there?

15:43:05 <mortenf> perhaps, but ask your q, and we'll try to make up an a

15:44:20 <phantoz> How do I represent reification in RDFS?

15:44:28 <phantoz> Or, rather, can I?

15:44:36 <mortenf> erm, that's already part of RDF?

15:45:03 <mortenf> or perhaps i don't understand your question

15:45:07 <phantoz> I want my schema to indicate that a given property applies to a reified statement

15:45:23 <phantoz> so if I want to statement "Joe opened webpage"

15:45:30 <mortenf> ah, not sure, but perhaps rdfs:range rdf:Statement?

15:45:46 <phantoz> I want to be able to say also ""Joe opened webpage" at 2:00"

15:45:48 <mortenf> or domain...

15:46:20 <phantoz> hrm

15:46:21 <mortenf> in general reification is avoided, better to use another level of indirection

15:46:41 <mortenf> i.e. model the event (in this case) as a class

15:47:16 <phantoz> ok so joe does some event

15:47:21 <phantoz> that event happened at a time

15:47:27 <phantoz> that event had an action

15:47:29 <phantoz> something like that?

15:47:35 <mortenf> yep, sounds reasonable

15:48:07 <phantoz> ok, I'll give that a wing

15:48:19 <phantoz> thx for the help

15:48:22 <mortenf> np

15:56:52 <GregElin> Hey all.

15:57:19 <mortenf> hi greg

15:58:32 <GregElin> hey mortenf

15:58:38 <bryce> morning.

15:58:44 <GregElin> hey

15:59:01 <LotR> ooh, greetfest :)

16:01:53 <GregElin> .time

16:01:54 <datum> Tue, 13 Jan 2004 16:01:54 GMT

16:02:32 <GregElin> Well...look at that 16:00 GMT! Are we ready to pick up or chat on www2004 photo project?

16:02:42 <bryce> i'm ready.

16:03:04 <GregElin> Libby, how do we do the attending thing again?

16:03:54 <LotR> use BLURB: attendance meeting foo? and then everyone can add a comment to that?

16:04:14 <GregElin> can you start, LotR?

16:04:32 <mortenf> there's already a B i think

16:04:34 <mortenf> B:

16:04:34 <dc_rdfig> blurb

16:04:35 <dc_rdfig> WWW2004 photos meeting, 2004-01-13, 1600 GMT, #rdfig

16:04:36 <dc_rdfig> (1:libby) [see agenda|http://lists.w3.org/Archives/Public/www-archive/2004Jan/0021.html]

16:04:39 <LotR> I'm not attending :)

16:04:58 <GregElin> BLURB: attendance meeting GregElin

16:04:58 <dc_rdfig> F: attendance meeting GregElin from GregElin

16:05:07 <mortenf> B:attending: [http://purl.org/net/morten/|Morten]

16:05:07 <dc_rdfig> Added comment B2.

16:05:10 <libby> sorry onn phone...back soon

16:05:31 <GregElin> Please use "BLURB: attendance meeting YourName" to indicate attending.

16:05:47 <LotR> no greg, do what morten did

16:05:47 <mortenf> erh, no, please don't

16:05:51 <mortenf> :)

16:06:06 <GregElin> B:

16:06:07 <dc_rdfig> blurb

16:06:07 <dc_rdfig> WWW2004 photos meeting, 2004-01-13, 1600 GMT, #rdfig

16:06:08 <dc_rdfig> (1:libby) [see agenda|http://lists.w3.org/Archives/Public/www-archive/2004Jan/0021.html]

16:06:09 <dc_rdfig> (2:mortenf) attending: [http://purl.org/net/morten/|Morten]

16:06:24 <LotR> the blurb thing was just to make an item for the chump

16:06:28 <Jhendler> B: attending Jim H for [http://owl.mindswap.org|MINDSEWAP group]

16:06:28 <dc_rdfig> Added comment B3.

16:06:37 * GregElin enrolls in remedial IRC class.

16:06:42 <bryce> B:attending: [http://tenyearsof.us/|Bryce Benton]

16:06:43 <dc_rdfig> Added comment B4.

16:06:43 <mortenf> heh

16:06:53 <LotR> it's not really irc

16:07:02 <GregElin> B: attending GregElin

16:07:02 <dc_rdfig> Added comment B5.

16:07:02 <nmg> B: attending: [http://www.ecs.soton.ac.uk/~nmg/|Nick Gibbins]

16:07:03 <dc_rdfig> Added comment B6.

16:07:04 <Jhendler> B3: attending Jim H for [http://owl.mindswap.org|MINDSWAP group]

16:07:04 <dc_rdfig> Replaced comment B3.

16:07:45 <LotR> I thought it was supposed to be MINDSEWAH ;)

16:08:02 <GregElin> Should we wait another couple of minutes or get started? Danbri? Libby?

16:08:33 <libby> I'll be there in a coupla mins but don;t wait for me

16:08:37 * LotR should remember how to stay out of the log

16:08:53 <karlcow> hmmm does it start now?

16:09:07 <GregElin> Sure. Why not. Let's start. If you are attending, please blurb so.

16:09:08 <swh> B3: attending [Steve Harris|http://inanna.ecs.soton.ac.uk/]

16:09:08 <dc_rdfig> Replaced comment B3.

16:09:14 <swh> sorry!

16:09:23 <swh> B3: attending Jim H for [http://owl.mindswap.org|MINDSWAP group]

16:09:24 <dc_rdfig> Replaced comment B3.

16:09:27 <swh> B: attending Jim H for [http://owl.mindswap.org|MINDSWAP group]

16:09:28 <dc_rdfig> Added comment B7.

16:09:34 <GregElin> Just getting started Steve.

16:09:36 <LotR> right

16:09:38 <swh> B: attending [Steve Harris|http://inanna.ecs.soton.ac.uk/]

16:09:38 <dc_rdfig> Added comment B8.

16:10:05 <GregElin> Okay announcements. List seems to be working okay? Anyone having problems?

16:10:12 <swh> no, works for me

16:10:19 <GregElin> Discussion already getting interesting (^_^)

16:10:26 <GregElin> ...on the list.

16:10:29 <mortenf> which list...

16:10:39 <GregElin> There is an email list for the project.

16:10:54 <libby> bijan: attending [libby miller|http://ilrt.org/people/libby]

16:10:59 <libby> B:attending [libby miller|http://ilrt.org/people/libby]

16:10:59 <dc_rdfig> Added comment B9.

16:11:04 <karlcow> B: attenting [Karl Dubost, W3C|http://www.w3.org/People/karl] and [Karl Dubost, Grange|http://www.la-grange.net/], who said schizophrenia.

16:11:04 <dc_rdfig> Added comment B10.

16:11:08 <GregElin> type "subscribe" in an email to semantic-photolist-request@unitboy.com

16:11:12 <mortenf> yep, i know it was being set up, but haven't heard about it otherwise

16:11:14 <mortenf> ah, thanks.

16:11:19 <mortenf> archives anywhere?

16:11:23 <GregElin> or send me an email at greg@fotonotes.net

16:11:58 <GregElin> here temporarily

16:12:00 <GregElin>http://lists.w3.org/Archives/Public/www-archive/

16:12:12 <mortenf> thx

16:12:22 * Jhendler will be in and out for beginning of irc - apologies in advance

16:12:31 * libby back

16:12:39 <GregElin> I have to leave before 1700 myself.

16:12:56 <GregElin> There is a proposed front door web site up at:

16:13:07 <GregElin>http://www.fotonotes.net/www2004photo

16:13:07 <dc_rdfig> G: http://www.fotonotes.net/www2004photo from GregElin

16:13:39 <bryce> G: proposed front door web site

16:13:39 <dc_rdfig> Added comment G1.

16:13:41 <GregElin> People want to take a quick look right now, if they haven't/

16:14:08 <GregElin> Does it seem okay?

16:14:14 <libby> G:|Greg's proposed front door website for WWW2004 photo annotation

16:14:15 <dc_rdfig> Titled item G.

16:14:22 <mortenf> looks fine

16:14:43 <mc_> why not put an exemplar photo on the page of TBL at a conference so folks get the idea?

16:14:43 <GregElin> Is it okay to starting telling people about this site if (a) the repository itself is not onlien and (b) if we use this URL?

16:14:48 * karlcow is not sure about the design and table layout :) but can arrange that if needed

16:14:48 <swh> its a bit hard on the eyes

16:14:54 <GregElin> Or should we wait to hear from Mike and Stu about a better URL?>

16:14:58 <bryce> on the list, i volunteered to web standards-ize it.

16:15:10 <libby> could always redirect later?

16:15:13 <bryce> sans tables for layout.

16:15:24 <mc_> tables are fine

16:15:24 <teefal> B:attending [timothy falconer|http://bigfractaltangle.com]

16:15:24 <dc_rdfig> Added comment B11.

16:15:25 <GregElin> TBL, mc?

16:15:27 <libby> that would be very cool bryce

16:15:29 <bryce> the contrast may be an accessiblity issue.

16:15:34 <karlcow> Very good bryce

16:15:41 <GregElin> Definitely redirect later.

16:15:45 <mc_> there's nothing wrong with tables

16:15:54 <libby> oooh!

16:16:01 <GregElin> All things open for change. I just wanted to get the big pieces there.

16:16:03 <mc_> there's a site to check for colorblindness issues

16:16:20 <mc_> pieces are good

16:16:25 <swh> mc_: theres lots wrong with tables - what about visually impared users?

16:16:30 <swh> [ot, sorry]

16:16:57 <mc_> so? you can still use tables and be ok for visually imparied users

16:16:57 <bryce> i agree that the site is off to a good start... there are some spelling issues that i can correct also.

16:16:58 <karlcow> [Hmmm, I would encourage to have the discussion about tables layout not now ;) and on another channel]

16:17:01 <GregElin> I think we should put up an example photo/photos asap.

16:17:12 <mc_> fine

16:17:15 <GregElin> Maybe we should get an example, if just screen shots online first?

16:17:33 <libby> .google WWW2003 photos

16:17:33 <bryce> +1 gregelin

16:17:34 <datum> WWW2003 photos: http://rdfweb.org/topic/BudapestWebConfPics

16:17:44 <libby> - some from last year...

16:18:05 <GregElin> Great!

16:18:15 <mc_> i think soton has some of mr semantic web with mr grid from 2002. will check

16:18:21 <GregElin> Okay..one more question...any designers to do a logo?

16:18:52 <GregElin> Please post some starter URLs for existing photos to the list, so we can contact people and get some initial contributions. thanks!

16:18:53 <bryce> Do we have an official title for the project?

16:19:20 <mc_> i meant put a sample photo right on the main web page you have

16:19:21 <GregElin> Semantic-Photo History of IW3C2 Conferences.

16:19:26 <bryce> i mean, a logo needs an official title to start from.

16:19:26 <libby> Liz Turner is a designer who has done some stuff for faof that's very nice

16:19:30 <GregElin> I like that idea, mc_

16:19:33 <nmg> +1 libby

16:19:35 <libby> danny ayres does some cute logos too

16:19:59 <GregElin> let's bounce them some ideas. I had a couple thoughts, if you could follow up.

16:20:00 <GregElin> Thanks.

16:20:22 <libby> s/ayres/ayers/

16:20:24 <inkel> inkel is now known as inkel_lunching

16:20:30 <GregElin> And lastly, how do we maintain? Volunteers? Bryce will help fix for accesibility and stadnards.

16:20:37 <Jhendler> Wendy Hall and Ben Schneiderman both have lots of pictures - that's part of how this whole thing got started

16:21:03 <GregElin> Bryce, can you help generally maintain the site, too, with me?

16:21:09 <bryce> sure

16:21:13 <GregElin> great!

16:21:21 <GregElin> Onward...usage and privacy policy.

16:21:29 <bryce> i can also install a cms or blog, if we want.

16:21:41 <GregElin> let's discuss that separately. I think it is a good idea.

16:22:07 <GregElin> Let me suggest the two big policy questions we face:

16:22:15 <GregElin> 1) are commercial uses of archive okay?

16:22:52 <GregElin> 2) What do we do about people in photos that do not wish to be in photos? How does EU laws affect (US doesn't care about anybody's rights but celebrities...

16:22:59 <mc_> wouldn't the license say "commercial uses would have to be looked at to the original copyright holders?

16:23:01 <karlcow> 1) Commercial: I would say no, at first shot.

16:23:17 <GregElin> Is that the general sentiment?

16:23:43 <mc_> if folks say they're in a photo and don't want it online, there's not much they can do if it was taken at a "public" event

16:23:48 <GregElin> The only 'commercial' issue I don't want to make painful is the commercial use of Ken Burns doing a documentary of the Web.

16:23:59 <bryce> Since the conf is in USA, i think usa laws of being photographed in public apply.

16:24:01 * mortenf seems to be losing connectivity, may be going, going, gone...

16:24:01 <karlcow> 2) Privacy: It's a big problem, and 2) is becoming more problematic of 1) and possible commercial use.

16:24:08 <GregElin> interesting thought, Bryce.

16:24:12 <libby> I'd agree w karl, maybe because people more likely to donate photos

16:24:14 <GregElin> But past photos from other conferences?

16:24:35 <libby> 2) I would suggest taking them down if pweople ask, like amihotornot

16:24:48 <libby> there was a recent rather stringest ruling in norway

16:25:02 <mc_> what if tbl said "i don't want my image used?" would he have the right to do that?

16:25:05 <GregElin> Okay...non-commercial it is, from the perspective of the "core" collection.

16:25:13 <mortenf> i think it was sweden...

16:25:34 <GregElin> Since contributors are known, and there's meta data...commercial users contact contributing party. Keep us out of it.

16:25:43 <karlcow> For example, I would say "no" to the commercial usage of my pictures.

16:25:44 <libby> - ttp://www.theregister.co.uk/content/6/34231.html

16:25:46 <bryce> it will be easy to create a list of all pics with a certain person.... possibly for review by that person.

16:25:51 <GregElin> That's the rub, mc_.

16:26:11 <libby> I think we have to err on the cautious side

16:26:22 <libby> on the other hand most people won;t jmind, I bet

16:26:28 <karlcow> hmm. You hvae commercial usage of the pictures you have taken, and commercial usage of the pictures you are on.

16:26:29 <GregElin> We could also distinguish between (a) being in a photo and (b) being identified in a photo.

16:26:45 <GregElin> Only identify people who say it is okay to be identified.

16:26:49 <danbri> perhaps if we thought about what 'comemrcial usage' might amount to?

16:26:56 <nmg> GregElin: interesting take

16:27:00 <bryce> good point. re: identified.

16:27:00 <danbri> eg. if it appears in a newspaper?

16:27:05 <danbri> ...an online newspaper?

16:27:07 <danbri> ...a weblog?

16:27:08 <libby> yep, might object to thewording of a description too, though not to the photo, e.g. 'X getting drunk' ;)

16:27:13 <danbri> ...a weblog that has banner ads?

16:27:23 <mc_> i don't think we should shut down commercial use - but that means there has to be a process for negotiation with copyright holders

16:27:39 <nmg> libby: does this mean that our ontology needs to be able to represent states of sobriety?

16:27:41 <libby> well copyright holders will be identified, so they can get in touch

16:27:53 <libby> oh for sure nmg :)

16:27:58 <GregElin> +1 mc_ no commercial uses coming from this collection. let the person who took the photo control/decide/be responsible for that.

16:28:07 <bryce> we said last week we could offer 2 cc licences.

16:28:21 <libby> should we have a vote on a specific proposal maybe?

16:28:56 <GregElin> I think 1 cc license on what is in the core, that is what is ON our server. People can use other licenses on virtualparts of the collection they maintain on their systems.

16:29:14 <libby> +1 for simplicity

16:29:21 <bryce> do the annotations need a separate licence from the image?

16:29:36 <GregElin> I vote for the attribution, share-alike, noncommercial CC license.

16:29:38 <danbri> consider http://creativecommons.org/license/sampling if we are doing regionDepicts markup...

16:29:58 <karlcow> hmmm is it possible two different licences for the same image hosted at two different places?

16:30:08 <libby> interesting danbri

16:30:16 <deltab> karlcow: yes

16:30:23 <GregElin> As a policy, let's put preliminary decisions on the email list and wiki and let people think about. Acceptance will happen if there is no complaint.

16:30:44 <bryce> karl: share alike means "no" to your question

16:30:51 <Jhendler> nmg: I think our ontology has to allow extension to anything - sobriety definitely something someone could add - I can think of lots of others as many of these pictures are likely to have been taken at receptions

16:30:59 <GregElin> very interesting, danbri.

16:31:11 <libby> +1 on take to the list

16:31:22 <bryce> +1 to list

16:31:28 <danbri> re sampling license, they do invite comment: "The licenses (there are two) are ready for use, but we're eager to learn how we can make them better. Sampling and collage bring up some complex issues, and we look forward to hearing your feedback on how we might better address them — though that shouldn't stop the pioneers among you from diving in."

16:31:31 <danbri> ok

16:31:53 <deltab> karlcow: it's even possible to have different licences for something hosted at the same place - some software is dual-licenced like that

16:32:09 <mc_> when, in the timeline, does the copyright issue have to be resolved so a copyright agreement can be on the site?

16:32:10 <Jhendler> I thnk we need a definite target date (soon) on getting this off the list - the privacy, commercial use, copyright issues are important to get done soon as we can

16:32:26 <GregElin> Agree, Jhendler, ontology can extend, but we have to track that a contributor agreed to a specific license as a minimum. It would be cool if ontology pointed toward additional licenses and where to get image under that license.

16:32:52 <libby> greg, shoudl be strightforward as long as license is included in metadta

16:32:58 <Jhendler> greg = I was respondiong to an earlier line - I think the contract/license has to be case in cement.

16:33:04 <GregElin> Ultimately the policies must be determined when we allow the public to contribute images.

16:33:17 <mc_> which is by when, now?

16:33:37 <GregElin> Which would be Feb 5 by the timeline. But we need to decide before. I'd recommend we decide by Friday.

16:33:39 <Jhendler> it's more than that Greg -- can I submit a picture of someone else without their permission?

16:33:51 <mc_> exactly, Jim

16:33:58 <mc_> i'd say, yes, tho

16:34:01 <mc_> if you took it

16:34:11 <mc_> and it's a public event

16:34:13 <Jhendler> I have several pictures of, for example, timbl that were taken under the condition that I promised not to make them public

16:34:36 <mc_> then you'd be a dishonorable cur if you published them, eh?

16:34:39 <Jhendler> if I published those, Janet Daly would have the right to punch me out - would also suggest she has the right to challeneg me legally

16:34:39 <GregElin> If it was taken in a public setting at conference and the portrayal was appropriate and family rated.

16:34:55 <libby> I think that's ok, as long as we take it down if requested to by that person

16:35:08 <mc_> this taking down on request is a potential trap

16:35:19 <GregElin> But do we take down WHOLE photo, crop, or punt it back to contributor?

16:35:21 <Jhendler> Greg - I can't take a random picture of you and sell it without your permission (legally) - so I'm thinking of two things -- one is privacy rights, the other is commercial development

16:35:27 <karlcow> Jhendler: no, Do not take a picture of someone else

16:35:33 <teefal> once it's posted, it's out... the internet never forgets :)

16:35:38 <GregElin> Yep, mc_ how doe we know who is asking for it to be taken down?

16:35:47 <libby> I do;t think it matters

16:35:50 <danbri> Does anyone have pointer to summary of how the laws differ in different counrties?

16:36:08 * sandro wants foaf:depicts info to turn automatically into asking the subject of the picture for a waiver......

16:36:17 <GregElin> You can do what you want Jim...the project won't sell the photos.

16:36:28 <Jhendler> no. no no - teefal, that is naive and wrong -- I can post a picture of someone on my web site, copyright it, and say not for commercial use and then could sue you if you took it and published it in a magazine

16:36:29 <danbri> (eg. In France recently I was taking pictures (of parisians) in the public street, and someone told me that's illegal...)

16:36:31 <deltab>http://www.krages.com/phoright.htm

16:36:31 <dc_rdfig> H: http://www.krages.com/phoright.htm from deltab

16:36:43 <teefal> three angles here? 1) what's legal, 2) what's friendly, 3) what will encourage people to post photos

16:36:55 <mc_> re: taking down, what if all the major players said "i want any pictures depicting me taken down, please" - there's little legal basis for that

16:37:13 <libby> why do you think they would do that?

16:37:21 <deltab> H:|The Photographer's Right: Legal Handbook for Photographers

16:37:21 <dc_rdfig> Titled item H.

16:37:21 <teefal> jim, i meant that the damage is done if someone doesn't want a photo public

16:37:37 <mc_> i'm not saying they'd want to i'm saying we have to have a general policy about putting up and taking down

16:37:48 <GregElin> Okay...hold up a second.

16:37:52 <GregElin> Let's back up.

16:37:56 <Jhendler> bottom line for me is that I think if we find the appropriate way to say "No commercial use without permission of photographer and subjects" that we would be better off

16:38:04 <karlcow> for now in France, the law on images is defined by Jurisprudence, but there's a move to make a law like in Germany.

16:38:28 <GregElin> Are we suggesting, in the first case, that we look at submitted photos to remove unflattering? That could be okay. We are pursuing an editorial objective: visual history of the W3 conferences.

16:38:56 <libby> no, not unflattering, unless requested to by the depicted

16:39:11 <GregElin> if panel feels photo "does not contribute to the visual history" it doesn't go into the core that is stored on our server.

16:39:22 <mc_> what panel?

16:39:30 <karlcow> 1. Accept only submitted pictures by the owner of the picture (photographer).

16:39:33 * libby surprised...

16:39:38 <mc_> if this is large scale automated, who's adjudicating?

16:39:40 <danbri> "flattering" is a tricky concept. Each photo might be flattering in its way, but if 10 photos of a person are arranged in a timeline, they may well go "heeeey, less emphasis on my expanding waistline please!"

16:39:45 <GregElin> panel = us, or a some set of identified photo editors

16:39:52 <karlcow> 2. Licenses defined clearly (IMHO, no commercial use)

16:40:01 <mc_> then we need to get assumptions about editors cleared up

16:40:06 <karlcow> 3. Remove any pictures that people don't wish to be on.

16:40:26 <GregElin> The idea is to create a significantly sized collection...but not completely opened ended archive of photos.

16:40:31 <karlcow> In fact for 3. the better would be that 1) have the authorizations of people before.

16:40:34 <Jhendler> +1, +1, +1 to karlcow - I think that is the right way to go - the no commercial solves a lot of issues

16:40:41 <bryce> there could eventually be a rating system (possibly related to degree of inebriety)... but i'm not suggesting a filter for this project.

16:40:53 <teefal> karl, better, get a list of people that *don't* want their photos

16:40:56 <libby> agree karlcow

16:41:08 <teefal> this is a closed group, yes?

16:41:14 <GregElin> That's interesting...rating system...bubble up photos...

16:41:25 <teefal> some notice saying "photos will be taken" with some way to opt out

16:41:29 <mc_> don't think agreements in advance is fair

16:41:32 <libby> grg, are you serious about rating by a closed panel?

16:41:46 * bryce might be suggesting it after all... but perhaps on quality.

16:42:25 <GregElin> Well...it is 'closed' in the same sense any internet project is closed: you don't like the emerging agreement, go fork yourself.

16:42:26 <Jhendler> closed panels/ratings -- yecch - open if we can - rules of the archive up front - most people will be friendly and flattered as long as we assure that the photos will not be reused (as much as we can legally)

16:42:35 <mc_> shall we look at the time and find a way to close this and move on?

16:42:57 <GregElin> LOL...not talking closed panel rating system.

16:43:11 <bryce> it might be useful to have a flag for review metadata option.

16:43:45 <GregElin> Ben recommended we start small...and we do some editing on the photos. We don't just take everything. It could get out of hand, people could be wading through lots of too dark photos...

16:44:10 <mc_> what a problem to have!

16:44:20 <teefal> or have an opt-in web database beforehand.... list of participants, with notice that they can click "I'm shy" to opt out... then some lucky person gets to compare that to the annotations

16:44:35 <GregElin> So the proposal distinguishes between a CORE collection of 'official' images (official in that they are hosted together, can be mirrorred, have the same license, are focused on the theme).

16:45:07 <mc_> it takes so much time to annotate a single photo that it seems it would be a miracle to be inundated with content

16:45:08 <GregElin> And a larger Virtual collection in which people use web semantics to attach their collections of images which they maintain and are responsible for.

16:45:13 <bryce> maybe only reciprocal foaf:knows can see certain images.

16:45:26 * libby agrees mc_

16:45:44 <Jhendler> you know, there's another approach that might solve this whole issue -- what if people submit the URIs of images, rather than the images themselves? Then the images are covered by the legalism of the hosting website

16:45:55 <teefal> +1 jhendler

16:46:02 <libby> I've got a few hundred photos here: http://swordfish.rdfweb.org/discovery/2001/08/codepict/ - I've never had any complaints

16:46:08 <danbri> Oh, I didn't realise that wasn't the intent :(

16:46:10 <libby> yeah that's how that site works

16:46:14 <GregElin> Problems that have to be figured out sooner or later, mc_. Question is, what part of the problem do WE want to figure out now, and what part to leave for others?

16:46:14 <mc_> greg and jim: seems a good idea to have a seed collection

16:46:17 <Jhendler> we have shown on the mindswap page that it is possible to do svg overlays of images that live elsewhere in web space

16:46:37 <mc_> what's the speed on that, Jim

16:46:53 <Jhendler> danbri - think there is a big difference between hosting the web page and pointing as I understand the law -- but I'm not a legal expert

16:46:58 <bryce> svg censoring might be possible... but ugly.

16:47:04 <GregElin> I need to call a break for a moment.

16:47:07 <danbri> svg overlays, yes that's common knowledge for a while I think, Jim.

16:47:09 <GregElin> BREAK!

16:47:22 <GregElin> It is 15 minutes before the hour. This discussion is great.

16:47:26 <Jhendler> mc - speed is web speeds - not great for lots of images, works well for a couple

16:47:49 <GregElin> I have to get going...how do we want to proceed with this discussion? With the rest of this chat?

16:48:23 <GregElin> I know I'd like to get a quick update from nmg on soton's work.

16:48:26 <bryce> task list update?

16:48:46 * libby wonders if we coudl schedule a meeting about tech aspects, esp image vocab

16:48:57 <GregElin> Can someone summarize the privacy discussion and links for the list? Try to nail the big questions?

16:49:09 <nmg> most of my efforts in this last week have related to the ontology

16:49:21 <GregElin> (Wed works for me libby)

16:49:25 * Jhendler suggests someone take an action to do a proposal on licensing to the mailing list, discuss it there, and then see if we can get a consensus - I think we also need to run it past someone who knows the law on this stuff

16:49:25 <nmg> I've (finally) got around to posting a review of jen's oconference ontology to the mailing list

16:49:41 <nmg> will do likewise for the image description ontology in due course

16:49:41 <GregElin> I like Jim's suggestion.

16:49:50 <libby> I can do wed too GregElin

16:49:58 <danbri>http://www.w3.org/2001/08/rdfweb/svg-foaf.html

16:49:58 <dc_rdfig> I: http://www.w3.org/2001/08/rdfweb/svg-foaf.html from danbri

16:50:13 <mc_> we need the ontology sussed so that we can get the annotation tool for the photos speaking the right metadata, no?

16:50:14 <danbri> I:|Using SVG overlays for arbitrary image annotation

16:50:15 <dc_rdfig> Titled item I.

16:50:49 <libby> yeah mc_, athough we maybe greg could hack something rough up inthe meanbtime, tweak it later

16:51:03 <GregElin> I think it is going to be pretty easy to glue the annotation tool and the ontology together. I'll do a screen mock up and we discuss how that back end works.

16:51:13 <nmg> indeed we do, mc

16:51:24 <GregElin> Libby, let's try to do that tomorrow?

16:51:27 <mc_> ok, who would like to be in the UI/annotation contingent?

16:51:31 <Jhendler> danbri: I didn't mean to claim that we invented the idea - I meant we showed how this could be integrated with image markup on the regions (i.e. combining jibberjim's stuff, foaf and some of our homebrew

16:51:40 <mc_> i'd like to be involved in the mock up discussions

16:51:41 <GregElin> nmg, great stuff. Thanks for sending to list.

16:51:44 <GregElin> Anything more?

16:51:52 <libby> ok gregelin

16:51:54 * Jhendler nmg - thanks for the review - Jen and I will bang on it

16:51:56 <danbri> I:Some experiments with Amaya SVG, plus [http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/|conversion from HTML imagemaps] using XSLT. Largely obsoleted by JimLey's pure SVG version.

16:51:56 <dc_rdfig> Added comment I1.

16:52:03 <mc_> yes, shall we again have a separate UI meeting on this mockup?

16:52:09 * GregElin Starts to pack up computer...encourages evryone to continue...and even discuss privacy more and ontology

16:52:51 * GregElin attends to wake up early tomorrow...get on list...coordinate with libby and set up a time to discuss UI tomorrow or Thursday.

16:52:59 <libby> :)

16:53:19 <bryce> gregelin, please get in with me touch re: website

16:53:51 <teefal> some questions about event... 1) how many in attendence? 2) how many register with email address? 3) how much time will be spent by you folks on annotation?

16:54:32 <mc_> global question: conference info: does the w3c have the schedules from all conferences? does anyone?

16:54:48 * GregElin departs with a smile

16:55:22 <Jhendler> I: [http://swint.mindswap.org/images|MINDSWAP site] playing with Jim Ley's stuff and foaf and content markup -- click on photo to see/browse the metadata

16:55:22 <dc_rdfig> Added comment I2.

16:55:50 <danbri> w3c doesn't; we don't run the conference. people often think we do. it does get used as w3c's annual reporting-to-public forum, and w3c folk are on the appropriate committee...

16:55:59 <danbri> doesn't... have the schedules, i mean.

16:56:05 <mc_> gotcha

16:56:11 <mc_> suggestions where to poke for this?

16:56:31 <mc_> iw3c perhaps?

16:56:43 <swh> swh is now known as swh_afk

16:56:44 <danbri> the committee whose acronym escapes me. yep, them. I doubt they'll have it though.

16:56:52 <teefal> one scenerio... if we have people's email addresses, send out notice of private site with photos.. this contains photos to conference participants... perhaps encourage THEM to annotate.... annotation is akin to permission, with option to request removal

16:56:58 <danbri> except in html or worse.

16:57:00 <teefal> people will be most motivated to find pictures of themselves

16:57:16 <libby> I tried to bug them (iwwwconf people) last year but no dice...maybe they'll listen to you

16:57:40 <mc_> ok

16:57:46 <libby> yeah, though might take them a while to find photos of themselves

16:57:56 <Jhendler> danbri - I think Janet represents W3C on IW3C2 and would know the answer - be worth buzzing her

16:58:17 <teefal> how many photos?

16:58:28 <mc_> Jim: who's Janet and i'll email

16:58:58 <mc_> it should be easy in any ui to find all the photos with annotations about yourself

16:59:04 <danbri> comms lead at w3c, see http://www.w3.org/People/Janet/

16:59:10 <libby> I dunno, but the ones danbri collated for w3c amounted to quite a few. one issue is that unless you know the person who took them, you probably won;t be oin the photo

16:59:11 <Jhendler> mc - was an indirect speech act on my part hinting to danbri that he would know and might take it on himself to ask her in a higher bandwidth mode

16:59:26 <mc_> oh. ok. tnx

17:00:04 <danbri> I don't have many high bandwidth moment w/ janet in my schedule, and wary of being a bottleneck... suggest someone else gets in touch directly (tho pls cc: me).

17:00:51 <mc_> other query: let's say we have the ontology worked out tomorrow, what work will need to be done either on fotonotes or svg thing to have it up to speed in two weeks for public use?

17:00:56 <teefal> i think the photo permission issue is instructive in itself...having a demonstration of a solution would be cool, especially at this event

17:01:06 <danbri> aha, http://www.iw3c2.org/Coordinates.html invites email to iw3c2@cern.ch

17:01:36 <teefal> the central philosophy is: do we have a right to see who see us?

17:01:47 <teefal> first see=choose

17:02:14 <libby> mc_ greg seems to think fairly simple

17:02:49 <libby> or something like http://kanzaki.com/docs/sw/img-annotator.html could be altered quite quicklyI reckon

17:02:51 <teefal> regardless of laws, is it better to let people consent to their public portrayal

17:03:02 <mc_> libby if that's the case, then we need to decide whether photos or pointers get uploaded, too, yes?

17:03:13 * danbri looks for Janet elsewhere in IRC, can't find her

17:03:20 <danbri> I'll drop a quick note to iw3c2@cern.ch

17:03:26 <mc_> thanks danbri

17:03:28 <libby> yeah I guess mc_

17:03:52 <libby> it doesn;t really matter; as time goes on there will be more broken links etc, but it works fine technically

17:03:57 <danbri> whats the best url to cite to explain this project and its relation to WWW2004?

17:04:14 <libby> jibberjim says he can also change his to new schema quickly

17:04:25 <mc_> that's great

17:04:34 <libby> as can I (though I can;t do image parts)

17:05:24 <mc_> how can we wrap up today's meeting?

17:05:43 <libby> well we need someone to make a proposal about privacy

17:05:46 <libby> or that would be good

17:06:11 <teefal> as an exercise, the heavy lifting solution: 1) we have everyone's email address, 2) we have a central server with decent database capabilities, 3) we get benevolent leprechauns to circle every recognizable face in every photo (to indicate annotation slot)

17:06:33 <mc_> that's good seeding.

17:06:54 <teefal> 4) people are emailed, 5) they click on their faces, 6) only when all slots are annotated does the photo surface to the public

17:06:54 <mc_> maybe the money Ben gets for a work study person can go to this kind of start up drone work

17:07:54 <teefal> 7) people can click NO on their face and we replace it with a bag over their head

17:07:59 <libby> yeah oit's the leprecauns bit that's the problem

17:08:05 <mc_> this is where it would be nice to have a desktop ap to go through a collection and let you annotate it, then upload the annotations and fotos

17:08:22 <Jhendler> teefal - exactly what we need to try to prevent -- suspect we need to find someone who knows someone who knows something about these issues and ask them how to do something a lot less cumbersome

17:08:25 <teefal> mc, we're working on it ;)

17:08:35 <Jhendler> mc - we're working on it.

17:09:05 <mc_> teefal and Jhendler - oh dandy! when can we beta test (on mac osx, please)

17:09:18 <teefal> mc, java... will release beta on march 21st

17:09:32 <mc_> what about pre-release??

17:09:45 <mc_> something we can work with to develop the seed base

17:09:54 <teefal> i'll talk to my people

17:10:06 <mc_> by the way, congratulations and thanks please do talk to your people!

17:10:43 <mc_> do we have other business before we wwind up?

17:10:56 <libby> well we have a lack of actions...

17:11:06 <mc_> ontology resolution is job one?

17:11:25 <libby> I meant stuff we discussed here today, but yeah

17:12:00 <mc_> action item: copyright resolution - Jim suggested mailing the list with a prototype

17:12:06 <mc_> that's an action item, yes?

17:12:18 <Jhendler> mc - but no one was assigned that action...

17:12:27 <libby> [[

17:12:29 <libby> <GregElin> Can someone summarize the privacy discussion and links for the list? Try to nail the big questions?

17:12:30 <libby> ]]

17:12:32 <mc_> Jhendler well let's do that

17:12:46 <mc_> is privacy now to be equated with copyright?

17:13:03 <bryce> action item: bryce is working with GregElin on the website.

17:13:19 <libby> B:action: bryce is working with GregElin on the website.

17:13:19 <dc_rdfig> Added comment B13.

17:13:56 <mc_> who's been looking at the creative commons stuff? could that kind soul come back with a proposal to the list?

17:14:40 * libby had a brief look, and suggested a license

17:14:45 <mc_> and action item: danbri's looking into the conference schedule's with the mysterious Janet

17:15:04 <libby> I thought he said he wasn;t, but wishes to be ccd....

17:15:20 <mc_> then he emailed iw3c2

17:15:27 <libby> ah, ok cool

17:16:13 <libby> B:action danbri look into findin out if we can get the conference schedules, via Iw3c2/Janet Daley

17:16:13 <dc_rdfig> Added comment B14.

17:16:21 <mc_> action item: libby, mc and GregElin are working on the ui for fotonotes for one by one submission of photos

17:16:32 <nmg> (need to go and catch a train - apologies for leaving before the end fo the meeting)

17:16:33 * mmealling wishes we could deprecate the use of the term 'webinar'....

17:16:39 <libby> bye nmg!

17:17:04 <danbri> B:I've sent mail to the iw3c2 contact address, will let you know if I get any answer. I expecte we will have to trawl [http://www.iw3c2.org/Conferences/index.html|previous conf websites] by hand though.

17:17:05 <dc_rdfig> Added comment B15.

17:17:12 <libby> B:action item: libby, mc and GregElin are working on the ui for fotonotes for one by one submission of photo

17:17:12 <dc_rdfig> Added comment B16.

17:17:23 <libby> oh man, that's a bit lame danbri. cheers tho...

17:17:34 <mc_> is there an action item on a sub team to resolve the ontology issue?

17:17:36 <danbri> well each conf has had different local organisers

17:17:45 * libby did that last yeah for WWW2003, was awful (my perl's not good mind you)

17:17:49 <libby> right

17:18:10 <libby> mc_ I will mail the list suggesting a meet later in the week baout it

17:18:16 <mc_> cool

17:18:25 <mc_> shall that be an action item?

17:18:32 <libby> B:action libby mail list suggesting meet later this week re ontologies/vocabs issues

17:18:32 <dc_rdfig> Added comment B17.

17:18:34 <libby> yeah

17:18:36 <mc_> thus, are we done for today?

17:19:06 <libby> B:action libby summarise today's discussion re privacy and copyright to the list

17:19:06 <dc_rdfig> Added comment B18.

17:19:08 <libby> that ok?

17:19:12 <libby> yeah I think so

17:19:23 <mc_> thanks

17:19:45 <mc_> shall we call "adjournment" then?

17:19:49 <libby> yeah

17:20:01 <mc_> meeting adjourned. thanks all

17:22:18 <libby> cheers

17:23:12 <bryce> i spoke with GregElin by phone last night. it was fun and inspiring.

17:28:18 <bryce> in the coming weeks, i would like to open up discussion of accessibility issues and how rdf can perhaps encourage new levels of accessibility (esp. regarding images).

17:30:18 <libby> that's a good idea

17:30:28 <Davey> Davey is now known as D[a]vey

17:32:53 <libby> one of the reasons I got into this was because I visited some blind people who were friends of william loughborough and I wanted to annnotate the photos I took of them in a way that they could read.

17:33:52 <bryce> wow, that's cool. i'll be discussing this project with a professor who researches elderly and their access. issues with the internet.

17:34:35 <bryce> gregelin said by far the ones most interested in fotonotes are older people. he said they want to tell stories.

17:35:35 <libby> these guys - http://swordfish.rdfweb.org/discovery/2001/08/codepict/aboutimage.jsp?uri=http://swordfish.rdfweb.org/photos/2001/08/05/000642.JPG

17:35:47 <bryce> i like the concept of remembering to include human elements (eg telling stories)... meaning the things we do better than machines.

17:36:01 <libby> heh, yeah

17:36:12 <libby> well, also, the things that we know and machines don;t

17:38:08 <bryce> right. but they could "learn" to recognize a face or an object, but probably not tell a story about it... just make some connections. (although that's also interesting... for the purposes of storytelling)

17:38:48 <libby> I was thinking of the implicit knowlege people have. sometimes it's the same thing

17:39:22 <bryce> right. it's interesting to note the overlaps.

17:45:36 <swh_afk> swh_afk is now known as swh_home

17:52:45 <D[a]vey> D[a]vey is now known as Davey

18:05:28 <libby> hey jibs

18:06:30 <JibberJim> hey libby

18:07:01 <JibberJim> Is there any discussion yet on what the vocab will look like for the image metadata?

18:07:12 <libby> there's discussion but no decision

18:07:26 <JibberJim> discussion on the list?

18:07:45 <libby> a little, and previously in chats

18:08:35 <libby> this has most links in I think: http://lists.w3.org/Archives/Public/www-archive/2004Jan/0022.html

18:09:30 <libby> I suggested might be good to discuss the depictsregion type issues in a separate chat smetime - the mindswap group have one too

18:12:08 <bryce> bryce is now known as bryce_out2lunch

18:16:57 <libby> can you post to thelist jibberjim? are you on it?

18:17:14 <libby> if we had a chat about regions of images it'd be great if you coudl make it

18:17:48 <JibberJim> I seem to be subscribed to the list libby! at least I got a load of messages.

18:17:51 <libby> ...and also masaahide, though whether americans and japanese could mae it at the same time, I dunno

18:18:11 <libby> cool, I think greg subscribed people who had participated before

18:35:15 <DanCon> which list, libby? oops... JibberJim?

18:35:36 <JibberJim> I think http://esw.w3.org/topic/ImageDescriptionList

18:39:32 <DanCon> I wonder what "The WWW2004 photo project" refers to

18:57:16 <JibberJim> I'm not completely sure DanCon, but it seems to be some project to take lots of photos and collect lots of metadata from people at www2004, I'm a little fuzzy on the details there, but I'm interested in aggreeing a common vocab for image description.

18:58:41 <dajobe> see http://fotonotes.net/www2004photo/ for the initial draft, web site

19:01:22 * DanCon wonders who's behind http://fotonotes.net/www2004photo/

19:02:07 <DanCon> Greg Elin, per DNS whois info

19:02:19 * mortenf wonders if "SourceSelector ::= QName" should be "SourceSelector ::= URI" instead, in grammar section of http://www.w3.org/Submission/2004/SUBM-RDQL-20040109/

19:02:40 <dajobe> follow the thread on http://lists.w3.org/Archives/Public/www-archive/2004Jan/

19:02:52 <DanCon> "patent pending" -- http://fotonotes.net/

19:02:59 <dajobe> and libby, bryce_out2lunch several others, jhendler, nmg, mc_

19:03:51 <DanCon> thanks for the pointers

19:03:53 * DanCon -> telcon

19:04:05 <dajobe> patents and software, whoa

19:05:34 <dajobe> mortenf: I should do a line-by-line review of that, you've caught another bug; it should be URL

19:05:49 <dajobe> er, URI

19:06:01 <mortenf> ah, had me worried there...

19:06:06 <dajobe> I'm comparing to the rasqal rdql yacc grammar

19:06:21 <mortenf> the reference implementation... :)

19:06:28 <dajobe> nonono

19:06:38 <dajobe> but I did use the BNF tokens from andy's earlier grammar

19:07:23 <dajobe> he's renamed some

19:12:07 * dajobe emails andys

19:19:06 <bryce_out2lunch> bryce_out2lunch is now known as bryce

19:55:48 <libby> dancon?

19:57:14 <libby> re list:

19:57:15 <libby> [[

19:57:16 <libby> 16:11:08 <GregElin> type "subscribe" in an email to semantic-photolist-request@unitboy.com

19:57:19 <libby> 16:11:12 <mortenf> yep, i know it was being set up, but haven't heard about it otherwise

19:57:23 <libby> 16:11:14 <mortenf> ah, thanks.

19:57:27 <libby> 16:11:19 <mortenf> archives anywhere?

19:57:31 <libby> 16:11:23 <GregElin> or send me an email at greg@fotonotes.net

19:57:35 <libby> 16:11:58 <GregElin> here temporarily

19:57:39 <libby> 16:12:00 <GregElin>http://lists.w3.org/Archives/Public/www-archive/

19:57:41 <libby> ]]

19:58:11 <libby> we're hoping to use a w3c list as soon as possible: http://esw.w3.org/topic/ImageDescriptionList

19:58:23 <libby> logger, pointer?

19:58:23 <libby> See http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-01-13#T19-58-23

19:59:02 <libby> B:there's an email list for the project, see [http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-01-13#T19-58-23|details of how to subscribe]

19:59:02 <dc_rdfig> Added comment B19.

20:07:21 <libby> <DanCon> "patent pending" -- http://fotonotes.net/ - I didn't know about that. there's a lot of similar work in this area.

20:27:29 <sky> sky is now known as sky_

21:00:00 <Davey> Davey is now known as D[a]vey

21:29:24 <ericP> what are the arguments to Redland::Parser::parse_as_stream ?

21:29:46 <mortenf> source and base uris

21:29:52 <ericP> i tried new RDF::Redland::URI("file:$in"); where $in was an RDF file in the current dir

21:30:22 <mortenf> yep, should work

21:30:33 <ericP> the first call to stream->next yields "0", which seems like neither a triple nor an error message

21:30:56 <mortenf> 0 means ok, not end of stream

21:31:05 <mortenf> use ->end instead

21:31:21 <dajobe> next means move on, not get next

21:31:37 <ericP> so this code is wrong? $model->add_statement($stream->next);

21:31:41 <dajobe> I had to split get_next into get & next

21:31:41 <dajobe> yes

21:31:55 <dajobe> ... I had to split get_next into get & next in order to allow access to contexts

21:32:01 <ericP> should be ->end or ->get?

21:32:11 <mortenf> use end to test for end

21:32:16 <mortenf> get_object to get statement

21:32:34 <mortenf> and get_context to get, well, the context

21:32:41 <dajobe> see http://www.redland.opensource.ac.uk/RELEASE.html#rel0_9_12 examples

21:33:23 <dajobe> fwiw, mortenf I added the envariables for rdfproc, so now I don't even know when I'm using the mysql store; apart from the speed :)

21:33:35 <mortenf> heh, it's that slow?

21:33:54 <dajobe> the connection pause

21:34:06 <mortenf> ah, yeah

21:34:58 <mortenf> i wanted something that scales, so it's not optimized for small interactions

21:38:13 <ericP> assuming this was the old API:

21:38:13 <ericP> my $stream=$parser->parse_as_stream($uri,$uri);

21:38:14 <ericP> while(!$stream->end) {

21:38:14 <ericP> }

21:38:34 <ericP> awe, friggin' tab problem in my irc client

21:38:58 <dajobe> move the next to the end of the loop, and use $stream->current at the start to get the item

21:39:32 <ericP> aha!

21:39:33 <ericP> tx

21:39:37 <dajobe> see the "Perl interface changes" at the URL above

21:41:49 <ericP> ok. might want to put them in the RDF::Redland::Parser perldoc

21:41:58 <ericP> something like

21:41:58 <ericP> my $stream=$parser->parse_as_stream($uri,$uri);

21:41:59 <ericP> while(!$stream->end) {

21:41:59 <ericP> $model->add_statement($stream->current);

21:41:59 <ericP> $stream->next;

21:42:01 <ericP> }

21:42:15 <dajobe> or just do it without perl

21:42:22 <dajobe> $parser->parse_into_model ...

21:42:24 <dajobe> look up the rags

21:42:26 <dajobe> args

21:42:53 <dajobe> that code is in the perl/example.pl

21:48:53 <DanCon> hmm... I need new business cards; wanna do it with RDF->SVG

21:49:27 <mortenf> your earlier request was noted, still looking to score the points...

21:53:27 <DanCon> ah; did I post a reward already?

21:53:46 * evlist has also a question about redland...

21:53:47 <mortenf> yep

21:54:40 <DanCon> ah; indeed... 750 points. http://rdfig.xmlhack.com/2003/12/07/2003-12-07.html#1070835232.976231

21:55:31 <evlist> I see that librdf_query_run returns a librdf_stream and find that strange since I wouldn't expect a RDQL query to return statements.

21:55:42 <dajobe> it's not an rdql query, yet

21:55:49 <dajobe> give me, ooh, a month or so

21:55:56 <evlist> ok!

21:56:13 <evlist> what kind of query is this then?

21:56:29 <dajobe> just a dumb triples matching one for the moment

21:56:35 <dajobe> it's skeleton code

21:56:55 * DanCon googles, finds somebody else had the same idea... er... no, forwarded mine... http://kaste.lv/~captsolo/info/blog_all.php/2003/12/09/p297

21:57:39 <evlist> btw, I think that it could be useful if a query language could return statements...

21:58:01 <dajobe> my (internal) thoughts are http://www.redland.opensource.ac.uk/notes/query-api.html

21:58:17 <dajobe> but I've not finished the detail of result sets

21:59:39 <evlist> In SQL, the result of a query is like a table and I can do interesting things such as "select into" or "insert select".

21:59:50 <DanCon> hmm... "2.2.2. Creating on-line Business cards with Java Server Pages and the DOM API" -- http://www.idealliance.org/papers/dx_xmle03/papers/02-04-05/02-04-05.html#s2.2.2

22:00:16 <evlist> IMO, that would be nice to be able to create triples as the result of a request for RDF as well.

22:00:46 <dajobe> evlist: yeah, I agree and plan to. SeRQL has that as the output of it's QL, as well as result sets

22:01:04 <evlist> great!

22:01:51 * DanCon fires up OpenOffice...

22:02:44 <DanCon> "Sample SVG Project: Business Card Creator" -- http://www.liberated.ca/software/svg.htm

22:03:30 <mortenf> hmm, seems i'll never be able to score those points, too slow...

22:08:13 <teefal>http://kensall.com/big-picture/bigpix22.html

22:08:13 <dc_rdfig> J: http://kensall.com/big-picture/bigpix22.html from teefal

22:11:21 <teefal> J:|Fun joke to play on new developers..."first, learn these technologies"

22:11:21 <dc_rdfig> Titled item J.

22:11:55 <teefal> J:Must be said with straight face

22:11:55 <dc_rdfig> Added comment J1.

22:14:20 <D[a]vey> teefal: eek

22:17:49 <teefal> daunting, yes...

22:18:25 <teefal> i look at that when i feel a need for humility

22:20:04 <teefal> does anyone know all this stuff? really know it?

22:20:17 * DanCon wonders why that question is interesting

22:21:17 <D[a]vey> teefal: I don't know, but its quite clear that whoever drew that has no clear conception of 3D art :)

22:21:42 <D[a]vey> (look at XSLT 2.0 and xmlns scheme's lines, they're all wrong :)

22:22:37 <teefal> dan, the scarey part for me is that architects need to make choices on what technologies to use.... the best architects know a little about everything, so they can choose the most cost effective, flexible approach

22:23:10 <teefal> choosing the right tech can save weeks and months of project time

22:23:20 <teefal> (or maintenance time)

22:24:02 <DanCon> ah; at that level, I think I know all of them.

22:24:08 <D[a]vey> teefal: isn't it the same when undertaking any project? right tool for the right job?

22:25:03 <teefal> once the tech is chosen, we gotta pick developers who know the tech, which means we gotta know what to ask them

22:25:20 <DanCon> I suppose the author of that book is claiming to know all fo them.

22:25:56 <mattmcc> In the sense of 'I know it exists'?

22:26:19 <teefal> i'd settle for just the red ones

22:26:36 <teefal> black ones could dry up and die in a year or two

22:27:05 <mattmcc> Some (DocBook, WAP, SAX) aren't going anywhere.

22:27:07 <D[a]vey> its actually scary looking at that just how many I know a little bit of, or know what it is, and how it works.

22:27:12 <teefal> (at least some... not javascript, web services, etc)

22:27:19 * DanCon looks up what black means... ah... not W3C... dunno... DocBook has staying power, at least

22:27:22 <mattmcc> By the same token, being a W3 REC doesn't mean it's a better choice.

22:27:25 <D[a]vey> If I spent my time focusing on a smaller group of things, I'd probably be better off :)

22:27:44 <DanCon> mattmcc, you don't think WAP is going away?

22:28:00 * D[a]vey hugs WAP

22:28:07 <DanCon> WAP has already gone away, from my radar screen.

22:28:12 <D[a]vey> well, WAP is a protocol, no?

22:28:15 <mattmcc> DanCon: With new phones still using it, at least not in the short term.

22:28:16 <D[a]vey> WML is the markup

22:28:27 <mattmcc> I certainly won't miss it, but.

22:28:35 <D[a]vey> mattmcc: newer phones can run opera, w00t :D

22:28:39 <teefal> WAP= why another protocol?

22:28:42 <DanCon> 100% of the users reports about WAP that have reached my ears/eyes have been "I hate WAP"

22:28:56 <D[a]vey> teefal: Wireless Access Protocol

22:29:00 <DanCon> or at least: "I don't even bother to use WAP"

22:29:02 <teefal> i'm joking

22:29:13 <D[a]vey> teefal: I think its a subset of HTTP? I dunno

22:29:21 <D[a]vey> DanCon: I love WAP...

22:29:35 <D[a]vey> DanCon: means I can send and recieve e-mails (through yahoo) anywhere in the world, pretty much :)

22:29:38 <DanCon> do tell, D[a]vey, what do you like WAP for?

22:30:00 * mattmcc idly ponders what JS is doing in a supposed picture of XML 'family' specs.

22:30:06 <D[a]vey> DanCon: google can be handy on the move too

22:30:20 <DanCon> you're happy to log on and get a WAP session, compose a message using a 4 line display?

22:30:32 <D[a]vey> DanCon: short ones, yes :)

22:30:52 <D[a]vey> DanCon: I have good enough memory not to need to see the entire memory, and I can "type" incredibly fast with predictive text :)

22:30:54 <DanCon> wow... you're the first satisfied user I've ever encountered.

22:31:12 <DanCon> I know lots of happy SMS users, but no WAP users.

22:31:15 <D[a]vey> s/entire memory/entire message/

22:31:16 <mattmcc> Eh, predictive text always fails me, I've long since turned it off.

22:31:16 <DanCon> er.. 1 WAP user

22:31:30 <D[a]vey> mattmcc: you need to teach it ;)

22:31:47 <D[a]vey> mattmcc: every time it fails, and you press "spell" (at least on Nokia phones) it remembers it :)

22:32:12 <D[a]vey> mattmcc: also, a lot of people don't see that you should press the star key to go through the combinations of letters :)

22:32:55 <D[a]vey> my phone lets me type most words I use now. Although I had to teach is PostgreSQL and XSL-FO today :)

22:33:13 <teefal> my wife and friends have predictive text wired into their autonomous nervous systems

22:34:43 <D[a]vey> D[a]vey is now known as Davey

22:35:03 <teefal> i am thinking about writing an article about this page tonight: http://www.w3.org/Consortium/Points/

22:35:46 <teefal> great page... very nice abstract & description of points

22:37:03 <mortenf> dajobe, ::Node says that new_from_uri can take a perl URI, but it complains: " Cannot make a Node from an object of class URI::http"

22:38:25 <dajobe> it checks with if (UNIVERSAL::isa($arg, 'URI::URL'))

22:38:34 <mortenf> ah

22:41:06 <teefal> anyone have links on efforts regarding "semantic proximity"

22:41:31 <teefal> tomato is more like an apple than a bicycle

22:41:56 <teefal> mark is closer to cindy than to julie

22:42:20 <aharth> teefal: you would either need a dictionary for that or use same similarity measure for graphs (pagerank, hits, etc)

22:42:48 <teefal> we are designing a semantic graph viewer that does this

22:43:07 <teefal> allows for both computer suggested arrangements and "manual overrides"

22:43:23 <teefal> (move uncle ernie way to the left, i don't want to see him)

22:43:38 <aharth> how do you specify proximity automatically?

22:44:03 <teefal> ultimately by use.... we're using a concept of "touches" to indicate node strength

22:44:17 <teefal> everytime the user touchs a node, it gets stronger

22:44:18 <bryce> maybe something with wordnet...

22:44:44 <aharth> teefal: yes, i use this in my system also, but it would be nice to have some initial ranking without usage data

22:45:02 <teefal> when you touch one node, then another, the two are marked as closer

22:45:12 <teefal> this is all just design stage now

22:45:21 <bryce> check out http://xml.mfd-consult.dk/foaf/explorer/?foaf=http%3A%2F%2Fxmlns.com%2Fwordnet%2F1.6%2FBeer

22:46:55 <kasei> with wordnet, you could compute the distance between points on the subClass hierarchy

22:47:28 <teefal> is there a more formal term for this than "semantic proximity"

22:47:33 <aharth> the issue with wordnet is that there is no such thing for other languages (German for example)

22:48:15 <danbri>http://www.illc.uva.nl/EuroWordNet/

22:48:15 <dc_rdfig> K: http://www.illc.uva.nl/EuroWordNet/ from danbri

22:48:20 <danbri> K:|EuroWordnet

22:48:20 <dc_rdfig> Titled item K.

22:48:52 <danbri> K:EU project 1996-1999. Don't think the dataset's available, at least under terms as liberal as core Wordnet's.

22:48:52 <dc_rdfig> Added comment K1.

22:49:05 <danbri> K:Interesting project though. There's a paper somewhere on the approach they took.

22:49:05 <dc_rdfig> Added comment K2.

22:50:05 <kasei> there's also kanzaki's j-e ontology: http://www.kanzaki.com/ns/jword/

22:50:05 <danbri> K:"The design of the database, the defined relations, the top-ontology and the Inter-Lingual-Index are now frozen. Nevertheless, many other institutes and research groups are developing similar wordnets in other languages (European and non-European) using the EuroWordNet specification. "

22:50:05 <dc_rdfig> Added comment K3.

22:50:32 <danbri> K:"The cooperative framework of EuroWordNet is continued through the Global WordNet Association. This is a free and public association that builds on EuroWordNet and Princeton WordNet."

22:50:32 <dc_rdfig> Added comment K4.

22:51:00 <tav> tav is now known as tav|away

22:51:18 <danbri> K4:"The cooperative framework of EuroWordNet is continued through the [http://www.globalwordnet.org/ Global WordNet Association]. This is a free and public association that builds on EuroWordNet and Princeton WordNet."

22:51:18 <dc_rdfig> Replaced comment K4.

22:52:21 <teefal> our app tags photos/movies/audio along the dimensions of who/what/where/when ... where & when have straightforward semantic proximity relationships (london near oxford, july 2003 near august 2003)

22:52:28 <teefal> who and what are more interesting

22:54:20 <teefal> what could be a wordnet thing

22:54:29 <teefal> who implies value judgements

22:57:04 <danbri> K:Oh, a conference! [http://www.fi.muni.cz/gwc2004/|Global wordnet conference, Jan 20-23 2004], Brno, Czech Republic. Anyone going?

22:57:04 <dc_rdfig> Added comment K5.

22:57:13 <DanCon> perhaps you could model semantic proximity of people by how much/often they interact with each other... so my family is close to me because we bump into each other daily... president bush is far away, but I watch him on T.V. now and again.

22:58:20 <DanCon> if you start with photos, you can use the number of photos that codepict X and Y as a rough approximation of how close X and Y should be

23:00:26 <teefal> we likely will start with an initial interview... who do you know, etc?

23:01:01 <teefal> but co-depictions is interesting....

23:01:36 <teefal> we are trying to have public & private arrangement of same nodes & links

23:01:49 <teefal> public=shared by multiple people, whole communities

23:02:34 <teefal> on one side, this is all manual, and people fight to move the same nodes around

23:02:53 <libby> does anyone know the blokie on the right here? http://swordfish.rdfweb.org/photos/2003/05/24/2003-05-24-Pages/Image0.html (scuse me for interruption)

23:02:57 <libby> bloke

23:03:06 <libby> mid-right

23:04:29 <teefal> better would be some automatic arranging, but the scheme to do that is our current question

23:04:34 <Davey> libby: nope, but if you know whose talking to him, he might know :)

23:04:56 <Davey> libby: out of interest was any thought given to how to "identify" poeple you can't identify?

23:06:03 <Davey> libby: and what about identifying where a picture was taken? so people can search on location...

23:06:12 <libby> the other gyu is brian mcbride :)

23:07:12 <libby> there's a bunch of stuff on rdfweb-dev list and on http://esw.w3.org/topic/ImageDescriptionRdfExamples about the latter and on rdfweb-dev about the former, though nothing definitive

23:07:39 <danbri> <foaf:Image rdf:about="http://swordfish.rdfweb.org/photos/2003/05/24/2003-05-24-Pages/Image0.html"><foaf:depicts><foaf:Person><foaf:name>Kim Veltman</foaf:name><foaf:page rdf:resource="http://www.mmi.unimaas.nl/people/veltman.html"/></foaf:Person></foaf:depicts></foaf:Image>

23:07:57 <libby> ahhh

23:08:08 <libby> cheers danbri

23:08:17 <danbri> in the foreground talking to brian

23:08:48 <mortenf> hmm, is that html document really an image? ;)

23:09:24 <danbri> pedant :)

23:09:44 <mortenf> thank you :)

23:09:59 <danbri> :)

23:12:48 <kasei> is anyone here going to be at sxsw this year? I'm looking to bootstrap my way into the codepiction strong set ;)

23:13:18 <libby> heh. szsw?

23:13:20 <libby> x

23:13:40 <kasei> South by Southwest

23:13:41 <kasei>http://www.sxsw.com/

23:13:42 <dc_rdfig> L: http://www.sxsw.com/ from kasei

23:15:07 <kasei> gah... I never seem to remember that that'll happen to urls on this channel! heh

23:17:23 <libby> ok to title it?

23:17:59 <libby> L:|South by Southwest

23:18:00 <dc_rdfig> Titled item L.

23:20:42 <danbri> K4:"The cooperative framework of EuroWordNet is continued through the [http://www.globalwordnet.org/|Global WordNet Association]. This is a free and public association that builds on EuroWordNet and Princeton WordNet."

23:20:42 <dc_rdfig> Replaced comment K4.

23:20:47 * danbri used wiki syntax :(

23:22:38 <bryce> i'll be at sxsw... ha! it's just down the road from here.

23:24:59 <kasei> ah, excellent

23:25:52 <bryce> kasei: between now and then, we can figure how to connect with each other.