Semantic Web Interest Group IRC Chat Logs for 2005-03-01

00:25:32 <Mkflow> what is the correct use of latex for display XML syntax? i mean that <element> gives ?!elemnt??, something like this. what do you usually use?

03:56:32 <libby> wow, snowy snow here in boston

03:56:58 * libby had a scarey trip in a taxi to hotel, sliding all over the road in a blizzard

04:12:00 <libby> brrr

11:11:39 <et_> et_ is now known as et

11:40:48 * jeen is away: ran out for a bit

12:28:13 * jeen is back (gone 00:47:25)

12:29:21 <danbri-laptop> danbri-laptop is now known as danbri

13:10:10 <JibberJim> Any RDF/SVG proposals for SVGOpen ? They've always been popular, be a shame to not have more?

13:11:06 <danbri> when are they due?

13:11:18 <JibberJim> erm right now...

13:11:42 <danbri> would be nice to see an update on sparql from someone, i think sparql and svg should play nice together

13:11:56 * JibberJim submits it for danbri...

13:11:58 <danbri> i'm minimising commitments for the summer, though i'll try to attend

14:07:01 <libby>http://www.w3.org/Talks/2005/0229-jk-rdf-sig/

14:07:01 <dc_swig> A: http://www.w3.org/Talks/2005/0229-jk-rdf-sig/ from libby

14:07:14 <libby> A:|RDF-SIG: Leveraging XML-SIG and XKMS for digitally signing RDF Graphs by Jose Kahan

14:07:14 <dc_swig> Titled item A.

14:08:02 <libby> libby has changed the topic to: Notes from SWIG f2f: agenda http://esw.w3.org/topic/SwigAtTp2005 Semantic Web Interest Group hack-n-chat - UTF-8 charset please - Weblog: http://swig.xmlhack.com/ - Logs: http://ilrt.org/discovery/chatlogs/swig/ - please identify with NickServ | SWIG meeting Mon Feb28/Tue Mar1 Boston - chat here

14:11:15 <dajobe> erika, could you join #swig, 'cos it kills the dc_swig bot each time

14:11:25 <dajobe> it's a bug, join #SWIG kills it, join #swig not :)

14:12:08 <libby> weirdy bug!

14:12:38 <dajobe> we thought it was fixed

14:12:39 * JibberJim wonders what happens if you join #SwIg

14:12:55 <dajobe> happily we have new shiny new bot code that might get installed some day

14:13:54 <libby> hurrah :)

14:13:56 <libby>http://swordfish.rdfweb.org/photos/2005/02/28/

14:13:56 <dc_swig> B: http://swordfish.rdfweb.org/photos/2005/02/28/ from libby

14:14:23 <libby> B:|photos from swig f2f Monday 28th Feb 2005

14:14:23 <dc_swig> Titled item B.

14:14:52 <bandri> seen jjc?

14:14:55 <bandri> bandri is now known as danbri

14:15:13 <crschmidt> .seen jjc

14:17:04 <libby> hm, very interesting using a sparql query as an xpointer...

14:17:18 <danbri> jose: next logical step is to add more semantics of signature inside the rdf, so we can query it

14:17:35 <RalphS_> Ralph: so, your example in http://www.w3.org/Talks/2005/0229-jk-rdf-sig/slide8-0.html is an idea for SPARQL + Xpointer use, not endorsed by any current WD?

14:17:41 <RalphS_> Jose: yes, just an idea

14:18:48 <libby> :)

14:19:07 <libby> sorry, I got distracted and couldnt report that accutaely ralph_

14:19:10 <danbri> (a v interesting idea... though xpointer always scares me somehow...)

14:19:32 <libby> (xpointer! boo!)

14:20:17 <danbri> (i think my concern is related to # vs / debate around media types and semantics of fragment IDs...)

14:26:34 <RalphS_> ChrisB: motivation for putting the sig in RDF was to be able to load the data into a repository prior to signature verification then verify the signature later if a query needs to use the signed data

14:26:41 <danbri> (swig f2f headcount: 34)

14:28:59 <libby> much giggling at jjc's unprintable example

14:35:53 <libby> jose: the important thing is that work is not duplicated

14:36:35 <danbri> what did chris b volunteer to do? :)

14:38:20 <danbri> phil archer (Icra)

14:38:38 <danbri> ...use cases from world of sex websites, and child protection

14:38:52 <danbri> internet content rating association

14:39:55 <CLoCkWeRX> pics lives!

14:40:00 <danbri> ...parental control over controv context

14:40:04 <danbri> content

14:40:12 <danbri> ...pics early w3c rec

14:40:17 <danbri> ...designed to facilitate self-labelling

14:40:20 <Mkflow> hi all

14:40:22 <danbri> ...some issues around trust etc

14:40:29 <danbri> (hi mkflow)

14:40:36 <Mkflow> hi danbri

14:40:39 <danbri> ...work in Japan

14:40:47 <danbri> (many of whom present in room :)

14:41:17 <danbri> ...working on providing a solution to a paritcular social problem in japan: mobile dating facilties in japan being abused for child prostitution

14:41:40 <danbri> in europe, european parliament wanted a system of 'quality labels' (tho didn't define what 'quality' means...)

14:41:48 <danbri> ...eg. credit card trust, medical info, ...

14:41:53 <jeen> does anybody know who I should pester for some editorial updates of www.w3.org/RDF (specifically the Developer Resources section)?

14:41:56 <danbri> ...that ended up as Quatro project

14:42:05 <danbri> ...on Quality + COntent Description

14:42:13 <danbri> various partners

14:42:16 <danbri> including danbri

14:42:16 <jeen> hi mkflow

14:42:34 <danbri> ...in Germany, a law that puts the onus onto content providers to provide youth protection systems

14:42:47 <danbri> ...if you have all-age content, you as a publisher have to do something to protect against that

14:42:49 <danbri> age verification etc

14:42:54 <danbri> law passed in april 2003

14:43:02 <Mkflow> hi jeen

14:43:21 <danbri> ...kaz is holding a workshop at www2005

14:43:40 <danbri> ...also Mobile Web Initiative (daniel applequist will talk more about that) doing "Mobile OK" labels

14:43:48 <danbri> ...building up web-of-trust, who will believe it, etc.

14:43:52 <jeen> mkflow, I always use PowerPoint (on windows) or XFig (on unix) for drawing pictures of graphs.

14:43:55 <danbri> aim: from icra point of view...

14:44:08 <danbri> we want to be able to make statements about example.org to say 'there is no nudity here

14:44:14 <danbri> ...that you can trust it w/ your credit card

14:44:22 <danbri> ...or medical info is ok

14:44:27 <danbri> *according to somebody*

14:44:32 <danbri> In graph form:

14:44:47 <danbri> "everything at www.example.org" -hasLabel-> "some rdf description"

14:45:09 <danbri> ...ie multiple uris with a single description

14:45:11 <jeen> mkflow, but there are rdf-specific editors that do graphics, like IsaViz.

14:45:19 <danbri> use case: MArtin Luther King

14:45:31 <danbri> ..www.martinlutherking.org

14:45:36 <danbri> ...historical "facts"

14:45:48 <danbri> ...most teenagers decide quality based on looks of the website

14:45:56 <danbri> ...if it looks good, gets used in homework

14:46:16 <danbri> ...turns out to be a racist front site

14:46:25 <danbri> ...would like Annotea like info about such sites

14:46:32 <danbri> -> "Organization X says this site is a racist site"

14:46:44 <danbri> ...also thekingcenter.com

14:46:49 <Mkflow> ok thanks

14:46:53 <danbri> ..."looks a bit reliigious" to UK people

14:47:04 <jeen> np :)

14:47:14 <danbri> ...shows firefox mockup

14:47:21 <danbri> ...also IE COntent Advisor

14:47:32 <danbri> shows IE (11% of market share in 96)

14:47:48 <danbri> ...icra pics labels were shown in ui

14:48:10 <danbri> ...content advisor could show / block etc

14:48:22 <danbri> I had to import a new vocab (ie a .rat file) to use icra stuff

14:48:36 <danbri> ...some possibilities re revisiting this from Microsoft

14:50:04 <danbri> ...need pics style wildcard

14:50:09 <danbri> ...in pics, gen-true

14:50:32 <danbri> ...ie. http://example.org/*

14:50:38 <danbri> ...you want a wildcard at both ends

14:50:47 <danbri> ...rdf "out of the box" doesn't quite do it either

14:50:51 <danbri> ...groups of uris

14:51:00 <RalphS_> q+ to explore Phil's point about 'aboutEachPrefix'

14:51:07 <danbri> ...we can rdfize content labels eg icra:vz="1"

14:51:16 <danbri> ..."no user generated content" etc

14:51:24 <danbri> ......need something like RULES

14:51:35 <danbri> ...plug URI in at top, put it thru some rules, get a description out at the bottom

14:51:54 <erim> q+ thinks the 'wildcard' view is perhaps not the right solution, the requirements here are very similar to archiving, describing 'sites' as an aggregate

14:52:19 <danbri> rules such as "is the site on a particlular host

14:52:28 <danbri> ....filter goes along, finds a link to the rdf...

14:52:35 <danbri> ...tess a bunch of uri regexes

14:53:03 <danbri> importnt for brand operator

14:53:21 <danbri> if you rrun a big company, eg. adobe, vodafone etc... you won't sign something unless you're super-sure it applies only to your stuff

14:53:31 <danbri> ...ie. don't accidentally label other people's stuff

14:53:56 <danbri> ...default label tricky in RDF

14:54:48 <danbri> (shows an RDF description of some defaulting rules)

14:55:37 <danbri> ...eg. checking we're on the appropriate host, and if it has 'ads' in url, then give it a certain label

14:56:06 <danbri> ...whole set of these rules in a sequence

14:56:15 <danbri> ...but some discussion yesterday re rdf sequence not being popular

14:56:24 <danbri> ...but we need it (or similar)

14:56:38 <danbri> [i think rdf:list better; should've mentioned that before]

14:56:43 <danbri> ...workflows, and things that work in reality

14:56:46 <danbri> Yahoo example

14:56:51 <danbri> they have 70 production centres around the world

14:57:02 <danbri> content labelling comes under policy group in sunnydale

14:57:22 <danbri> ...they can't deal w/ each of those 70 groups at v fine grained level

14:57:41 <danbri> ...but they can have a policy such as "every content server points to a labels.rdf file via a link tag"

14:57:47 <danbri> ...ie. get everyone to do one thing once, then forget about it

14:58:16 <danbri> ...we're going to be putting these in place RSN (next week...)

14:58:25 <danbri> ...re "Who says...

14:58:32 <danbri> ...who says what, how do i sign it, does that help, ...

14:58:40 <danbri> ...within the rdf instance that includes the rules and the labels

14:58:46 <danbri> ...talks about the rdf instance

14:58:54 <danbri> and rdf:about="" for info about the label

14:59:14 <danbri> say things like label:authorityFor, to scope the labls to the internet domain it has competence regarding

14:59:32 * danbri waves to et

14:59:44 <danbri> ....client/request/label/trust cycle as people browse...

14:59:49 <danbri> ...consult some database

15:00:10 <danbri> ...maybe use hash of the file, send hash to label authority to confirm that the label was reviewed by some labelling authority

15:00:17 <danbri> ...and see any number of 3rd party annotations

15:00:41 <danbri> ...we're promoting this model, which fits what we've heard here in swig meeting

15:00:48 <danbri> drive: self labelling hasn't taken off in a big way

15:00:51 <danbri> ...we don't pretend it has

15:00:55 <danbri> ...people need motivation to do this

15:01:01 <danbri> ...search enginges ignore this

15:01:05 <danbri> ...want to make it useful

15:01:19 <danbri> ...political pressure, in eu parliament,japanese and german govt, for progress

15:01:41 <danbri> see quatro doc on w3 website, emphasise it isn't part of w3c standards track

15:01:49 <danbri> ...there thru dan's involvement in quatro

15:01:55 <danbri> possibility of conferences, workshops etc

15:01:57 <RalphS_>http://www.w3.org/2004/12/q/doc/rdf-contentlabels.html

15:01:58 <dc_swig> C: http://www.w3.org/2004/12/q/doc/rdf-contentlabels.html from RalphS_

15:02:00 <danbri> ...see also icra.org website

15:02:09 <RalphS_> C:|RDF Content Labels: Use Cases and Draft Schemas

15:02:10 <dc_swig> Titled item C.

15:02:54 <chaalsBOS2> chaalsBOS2 is now known as chaalsBOS

15:03:04 <RalphS_> C:a work-in-progress not currently part of any W3C Working Group

15:03:04 <dc_swig> Added comment C1.

15:07:20 <RalphS_> I wanted to ask Phil about deployment expectations; specifically requiring lots of inferencing in clients versus supporting .e.g. cwm's string:contains at the server

15:07:57 <RalphS_> It seems to me that there's better hope of getting clients to recognize filtering labels sooner when they don't also have to support inferencing

15:09:05 <RalphS_> Phil: considering RDF/A, label:hasLabel can go on any tag but not on hyperlinks -- I'd like to see if label:hasLabel could be used on hyperlinks

15:09:43 <RalphS_> JJC: motivating digital signatures for labels is flawed

15:09:44 <dc_swig> Label JJC not found.

15:09:55 <RalphS_> ... a hacker can hack the content and leave the label alone

15:10:28 <RalphS_> ... if the source is insecure then the metadata is just as insecure

15:10:56 <RalphS_> CharlesM-N: the content can be signed [within the label] as well as signing the label

15:11:53 <RalphS_> Danny Weitzner: the problem here is so far away from the technology of signing labels

15:13:57 * RalphS_ supposes q+ in irc may have no effect in this meeting

15:14:29 <danbri> can someone scribe pls?

15:14:41 <danbri> q+ doesn't, sorry. handwaving does, often.

15:15:09 * RalphS_ put his comment in the record here in irc anyway

15:15:34 <danbri> fumihiro kato

15:15:58 <danbri> labeling... faster info access; personalise/customise... categorise/trust...

15:16:21 <josek>http://www.tom.sfc.keio.ac.jp/~torry/TP2005-Labeling.ppt

15:16:21 <dc_swig> D: http://www.tom.sfc.keio.ac.jp/~torry/TP2005-Labeling.ppt from josek

15:16:36 <josek> C:|Trust with Labeling slides. Fumihiro Kato, Keio University

15:16:36 <dc_swig> Titled item C.

15:17:14 <danbri> ..by blocking unwanted websites, can get better access to those that are of interest

15:17:59 <danbri> ...enable categorisation and ratings of sites, eg. good sites, or good restaurants (4 *), or movie (pg-13...)

15:18:05 <danbri> ...want to use labels

15:18:35 <danbri> ...labelling is for making trust easier

15:18:42 <danbri> ...criterion of trust

15:19:14 <danbri> ... labelling supports trust, through various criteria

15:19:30 <danbri> ...i list three: 1. Quality label 2. Accountability 3. reputation

15:19:48 <danbri> quality label is main purpose of labelling, such as R-rating for movies, or *-s for restaurants

15:20:05 <danbri> accountability is one of the most important issues in trust

15:20:26 <danbri> ...so when movie has some rating... ask Why? by Who?

15:20:35 <danbri> ..who rates it R is important

15:20:59 <danbri> ...sometimes I'd want to know the who/why/how behind content rating

15:21:34 <danbri> finally reputation, another important aspect. For example, I bought this PC...

15:21:40 <danbri> ...because my friends who i trust recommended it

15:22:05 <danbri> ...or I like Sony, so I trust Sony and buy their PCs

15:22:14 <danbri> ...an important aspect of reputation is Social Relationships

15:22:35 <danbri> ...so labelling provies a common format for all these criteria

15:22:51 <danbri> Labelling has 2 parts. Centralised Labelling and Distributed Labelling.

15:22:55 <danbri> Centralised Labelling:

15:23:19 <danbri> ...provided by some organisation, such as insurance, or restaurant raters

15:23:21 <danbri> ...movies, companies

15:24:21 <danbri> ...so who uses these labels?

15:24:26 <danbri> ...eg. i want a 4-star hotel

15:24:42 <danbri> ...I could ask to get all labels with 4 and 5 stars

15:25:11 <danbri> ...these labels have problems around accountability and transparency

15:25:27 <danbri> ...we might want to know why some hotel has 4* or why some movie is given pg-13

15:25:41 <danbri> ....like to know its evidence for accountability purposes

15:25:56 <danbri> ...when filtering contents, filtering software recommends or blocks websites

15:26:17 <danbri> ....when filtering, we'd similarly want to know 'why'

15:26:54 <danbri> ...if i know the reason for the recommendation or blocking, then i can trust the software or company because of transparency

15:27:12 <danbri> Rule Language for accountancy and transparency

15:27:21 <danbri> ... evidence -> rating

15:27:48 <danbri> ...eg if Michelin give 4 *, I might want the evidence behind that

15:28:15 <danbri> which might be "the 4 * was awarded because restaurant=4, staff=5, etc."

15:28:34 <danbri> ...ie. show the logic behind the labels

15:28:54 <danbri> ...labelling providers show reasons

15:29:31 <danbri> Distributed Labelling

15:29:43 <danbri> ...in social network services, individual users can share ratiings, categorisations...

15:30:10 <danbri> ...typical example might be Amazon's site

15:30:19 <danbri> ...which allows labelling of books, products, etc.

15:30:24 <danbri> ...share tagging and rating

15:31:05 <danbri> ...we can watch such ratings and use them to help us decide what to buy

15:31:24 <danbri> ...people trust each other thru SNS (social networking services), friends, friends of friends, ...

15:31:34 <danbri> ...my friends recommend some book, with some rating, ...

15:31:51 <danbri> ...we trust this friend so we trust their rating

15:32:28 <danbri> ...if we describe these relationships, and use them for trust, we need more complicated rules and reasoning to find similar interests/people

15:32:47 <danbri> ...because these websites provide quite differenging formats to describe interests of people

15:33:07 <danbri> example: Social Bookmark Management

15:33:10 * reto /me trusts different friends to a different dedgree depending on topic

15:33:19 <danbri> finding people... by clipping the same URL

15:33:22 <danbri> ...by tagging the same interest

15:33:36 <danbri> ...we share the url and interest

15:33:57 <danbri> (*aside: i'd like to talk about SKOS/FOAF/RSS combination for this, later...*)

15:34:15 <danbri> ...share tags, usign websites and rss

15:34:32 <danbri> so we need to describe labelling within RSS using some common format

15:34:34 <danbri> (shows and example

15:34:59 <danbri> [an rss:item with a label:hasContentLabel property pointing to a contentlabel with a category]

15:35:22 * danbri points aliman at this; alistair, i was discussing SKOS and labelling w/ Phil recently...

15:35:32 <danbri> Summary: labelling is one of the most importnat issues in trust

15:35:40 <danbri> ...common STD format needed to make trust

15:35:46 <danbri> ...this concerns, rules, query, and other SW tech

15:35:51 * aliman looks

15:36:26 <josek> (STD == standard or ...? )

15:36:30 <danbri> standard

15:36:37 <danbri> (end; thanks|)

15:37:04 <danbri> kaz: re phil's approach vs this... this approach emphasises lots of different ratings from lots of students

15:37:11 <danbri> ...question of who makes a trustmark

15:37:36 <danbri> ..bookmark sharing

15:37:43 <danbri> ...god info source to create a reputation

15:38:35 <libby> --->coffee

15:39:04 <RalphS_> Kato-san's presentation points out why RDF supports this use case so well; with RDF the label can contain as much supporting evidence as it wishes, which helps to build reputation

15:39:40 <RalphS_> ... also, this shows why rules should be first-class RDF so they can be searched and queried with the same sets of discovery tools

15:39:57 <RalphS_> ... to permit rules to add to reputation

15:42:03 * chaalsBOS waves to Reto

15:42:27 <Mkflow> hi to all

15:44:26 <reto> raplhs: do you have an example for rules/reputation?

15:47:00 <sh1mmer> reto in BOS?

15:47:56 <reto> sadly not

15:48:06 <reto> btw, hi

15:48:11 <sh1mmer> hi

15:48:16 <sh1mmer> yeah its been a while

15:53:00 <chaalsBOS> reto, ralphs is out

15:53:08 <chaalsBOS> oh, no he's not.

15:55:31 <reto> hi chaals! coffee i know

16:00:22 <dajobe> erika!

16:01:47 <RalphS_> rules/reputation example -- I don't have one to copy & paste but Kato-san suggested in his slides that one could find people with similar tastes by comparing their filtering rules

16:04:29 <danbri> daniela:

16:04:37 <danbri> i work for vodafone... a big phone company

16:04:48 <danbri> ...not here to talk about detailed rdf tech issues

16:04:55 <danbri> ...but presenting use cases around sw tech

16:05:03 <danbri> ...and huge value we see that relates sw tech to our business

16:05:06 <danbri> ...some around content metadata

16:05:12 <danbri> ...and some around adult content

16:05:21 <danbri> vodafone: 150 million subscribes

16:05:25 <danbri> ...headquarters in uk

16:05:29 <danbri> vodafone.com

16:05:34 * danbri won't clone his slides

16:06:09 <danbri> ...helping to develop Mobile Web Initiative

16:06:31 <danbri> ...prev director of content management at thestreet.com

16:06:45 <danbri> ...where i got the Metadata religion

16:06:51 <danbri> ...news, articles etc about companies

16:07:04 <danbri> ...financial news has a built-in ontology... stockticker symbols

16:07:19 <danbri> ...but if you have a reference db that connects those symbols to industries etc

16:07:28 <danbri> ...we switched to an xml-based content management system in 98

16:07:47 <danbri> ...coded up ticker symbols in markup

16:08:00 <danbri> ...helpd us slice/dice content, sell biotech feeds to biotech industry

16:08:13 <danbri> ...better target/package content without great manual editorial effort

16:08:18 <danbri> ...so that has driven my perspective

16:08:34 <danbri> ...bnefore that worked w/ Nature and Science and other sci titles

16:08:39 <danbri> Vodafone Live

16:08:44 <danbri> er

16:08:51 <danbri> "Vodafone Live!", rather

16:08:54 <danbri> ...essential a portal

16:09:06 <danbri> ...ui over some content, retrieved from other srcs via a webservice interface

16:09:10 <danbri> ...device neutral format of content

16:09:17 <danbri> ...particulars of device, characteristics

16:09:24 <danbri> ...because we're brokering content

16:09:31 <danbri> ...we have a built-in payment mechanism

16:09:36 <danbri> ....buy a ringtone etc

16:09:38 <danbri> ...music download

16:09:43 <danbri> ...itunes-like thing

16:09:48 <danbri> ....all seamless to the customer

16:09:53 <danbri> ...have the billing relationship

16:09:59 <danbri> ...so we play a brokering role

16:10:13 <danbri> ...background material on mobile content

16:10:26 <danbri> ...challenges for mobile web, not just (relatively closed-world) mobile portals

16:10:35 <danbri> ...right content at right time, less willingness to browse

16:10:42 <danbri> ...idea of multiple use content starts to breakdown

16:10:47 <danbri> ...prepare content for the mobile channel

16:10:51 <danbri> ...small versions of headlines

16:10:59 <danbri> ...customers are not as forgiving as they are w/ the web

16:11:32 <zool> because of the cost recovery model?

16:11:32 <tlr> How many percent of searchers look beyond hit 10 at Google, again?

16:11:45 <danbri> ...per usage fees for tracks, not data rate

16:12:03 <danbri> ...still expectation that customers are paying, so need to be v clear re relevance

16:12:09 <danbri> ...search results must be relevant

16:12:11 <danbri> ...personalisation

16:12:24 <danbri> ...common example we hear re marketing is amazon suggestions

16:12:56 <danbri> ...in amazon contxt on desktop, people can ignore irrelevant suggestions

16:13:08 <danbri> ...on mobile, it takes up all screen, costs customer trust

16:13:18 <danbri> esp true on mobile phones, as a strong branding relationship customer/carrier

16:13:23 <danbri> ...latter put logo on phone, etc

16:13:27 <crschmidt> (bigger screen!)

16:13:35 <danbri> ...'vodafone is a quality mark', that's how carriers differentiate

16:13:41 <danbri> ...on quality and perception of quality

16:13:52 <danbri> ...also regulatory issues in europe, incl. uk

16:14:06 <danbri> ...we are being held to a different responsibility levels compared to generic desktop/isp situation

16:14:20 <danbri> ...an arugement for why this is necessary: these mobiles are often not under parental control

16:14:28 <danbri> ...u can monitor kids surfing on a shared pc, vs a web-phone

16:15:19 <danbri> ...example of wanting content relating to Madonna

16:15:41 <danbri> ...danger of trying to sell people Madonna ringtones while they read about the Weeping Madonna"

16:15:49 <danbri> ...what are we doing now?

16:16:00 <danbri> ...search on mobile

16:16:36 <danbri> ...re search, see a bunch of categories

16:16:47 <danbri> ...tryting to provide a search experience, choose amongst existing categories of info

16:16:57 <danbri> ...and match to various content providers

16:17:00 <danbri> ...using rdf metadata

16:17:05 <danbri> ...simple file, content partners put on sites

16:17:13 <danbri> ...about a year ago

16:17:25 <danbri> ...when we impliemented that, we saw revenue for those who used the rdf jump

16:17:50 <danbri> ...somewhat amazing to us

16:18:02 <danbri> ...we had to sell rdf to our content providres

16:18:42 <danbri> ...vodafone live is our "closed world"

16:19:02 <danbri> ...also 'friendly content' where we have some rel'n w/ content provider

16:19:55 <danbri> ...we also integrating use of what Phil described into content from partners

16:19:59 <danbri> ...filtering on a hyperlink

16:20:08 <danbri> [takes to whiteboard]

16:20:33 <danbri> ...eg. we're providing search result list to people

16:20:40 <danbri> ...they searched on 'all videos by ...'

16:20:43 <danbri> ...and see results

16:20:58 <danbri> ....search results A and B are videos w/ no adult content

16:21:10 <danbri> ...but C comes from an adult provider, w/ violent content, nudity etc.

16:21:26 <danbri> ...we can check a users personal profile, and choose to do a number of things

16:21:48 <danbri> ...grey out the link, or not show it, ...

16:21:53 <danbri> ...we need fine-grained level

16:22:00 <danbri> ...as definitions vary betwee ncountries

16:22:08 <danbri> ...eg netherlands vs egypt

16:22:40 <danbri> ...rdf info in the links [missed detail]

16:23:31 <[GNU]> moin

16:23:57 <danbri> Mobile Web Initiative

16:24:09 <danbri> ...drafting a 'best practices working group' within w3c

16:24:30 <danbri> ...MWBP WG... will make a Mobile OK trustmark

16:24:35 <danbri> ...idea of A, AA, AAA conformance

16:24:39 <danbri> ...but machine-readable

16:24:43 <danbri> ...analagous to WAI

16:24:49 * danbri wonders relationship to EARL

16:25:03 <danbri> ...issue: who certifies that content is mobile friendly?

16:25:05 * erim thinking similar things with EARL, Annotea

16:26:18 <danbri> ...we want things like "last time we checked, this content was mobile-firendly or tuned v closely to the mobile experince

16:26:26 <danbri> Questions

16:27:10 <danbri> (Bijan will now talk after lunch...)

16:29:38 <danbri> phil: danny ash wouldn't put metadata on her site, thru fear of blocking

16:29:56 <danbri> ...urban myth that yahoo downrate things w/ pics labels

16:30:06 * danbri stops scribing

16:35:36 <danbri> ...[...] "who has vouched for ability of some site to self-label correctly

16:35:44 <libby> I gues the regulatory environment has tied them in knots...

16:36:07 <danbri> ...right now we are doing adult content filtering for mobile web, but all based on url blacklists/whitelists

16:36:14 <danbri> ...couple of things in it i don't like, eg. format of urls

16:36:48 <danbri> phil: porn sites label acdurately

16:36:59 <danbri> ... in that they generally don't lie about their porn-ness

16:37:14 <danbri> ...we are thinking about doing something in space daniel suggests

16:37:41 <libby> [beerintheevening.com is blocked as adult in uk vodafone system]

16:37:53 <danbri> ['eve' maybe?]

16:38:26 <danbri> ...

16:38:45 <libby> ['beer'?]

16:38:54 <libby> heh

16:39:08 <KjetilK> I think it could help people would rate self-labels, a quick click in the browser UI to mark the label as inaccurate or fraudulent

16:39:41 <KjetilK> if someone that you trust marks a self-label as untrustworthy you would not bother about it

16:40:06 <KjetilK> this would encourage people to provide accurate descriptions about themselves

16:44:04 <danbri> kaz: (raised q of cultural diffs between japan and US re attittudes to filtering, censorship...)

16:44:13 <PatrickS> ['in thee'] ;-)

16:44:32 <danbri> danny: transparency is key

16:45:08 <danbri> ...ppl want to be able to choose

16:45:19 <danbri> ...real user choice

16:47:57 <danbri> frank: some prev schemes have involved manadatory categories

16:48:01 <danbri> ...blocking if that missing

16:48:59 <danbri> phil: pics was found to be 1st-ammendment friendly in us courts

16:49:07 <danbri> ...transparency is important

16:49:13 <danbri> ...but parents don't care about that

16:50:17 <danbri> chaals: Australians filter violence but not care about sex

16:51:46 <libby> phil: aol deuchland does not make its erotic content available until 10pm at night!

16:53:25 <danbri> (swig f2f head count: 34)

16:58:20 <tlr> lunch is ready

16:58:41 <libby> :)

16:59:03 <crschmidt> lunch? I'll be right there! :)

17:00:37 <danbri> danielapplequist: i was involved in prism

17:00:42 <danbri> ...some discussion re mobile aspects

17:00:54 <danbri> ...some interest, but not that much motivation

17:01:01 <danbri> ...also poss around idealliance

17:01:27 <chaalsBOS> .t gmt

17:01:29 <phenny> Tue, 01 Mar 2005 17:01:27 GMT

17:02:20 <libby> ---->lunch

17:47:26 <dviner> dajobe gone for the night?

19:05:36 <bandri2> bijan "but wait, there's more" parsia showing rdf/sw goodness from a WS-Policy mapping

19:08:23 * bandri2 is impressed

19:08:37 <dajobe> what?

19:08:51 * chaalsBOS is impressed (and interested, cause EARL is looking to do this kind of stuff)

19:10:13 <dajobe> admin notice: dc_swig and chump web site will be offline for a while - server maint

19:10:39 <dajobe> although I notice somebody killed dc_Swig again earlier

19:11:43 <edd> villains!

19:11:53 * crschmidt coughs at erika

19:15:56 <dviner> hi dave

19:18:56 <dajobe> hi dave

19:19:48 <dviner> wonder if we could chat offline for a sec?

19:20:06 <dajobe> I've jsust this sec started dinner

19:24:09 <chaalsBOS> i'm sorry dave...

19:25:43 <RRS>http://sw.deri.org/~aharth/2005/02/swigf2f/slide1.html

19:25:43 <dc_swig> E: http://sw.deri.org/~aharth/2005/02/swigf2f/slide1.html from RRS

19:25:58 <RRS> E:|Semantically Interlinking Online Community Sites

19:25:58 <dc_swig> Titled item E.

19:26:21 <RRS> E:Andreas Harth, SWIG F2F, 1 March

19:26:22 <dc_swig> Added comment E1.

19:28:00 <RRS> I hear Andreas pronounce SIOC like "shock" <en-us>

19:28:29 <sandro> Me too. He said it was the Irish word for "frost"./

19:28:42 <Cloud_> Indeed it is

19:28:48 <sh1mmer> hey John

19:28:53 <Cloud_> Hi Tom

19:29:14 <sh1mmer> I don't see you onthe attendance list

19:29:15 <sh1mmer> :)(

19:29:24 <Cloud_> I wish :)

19:30:02 <sh1mmer> just think about not being cold and in 3' of snow

19:30:19 <sandro> Shockingly low show of hands on "who out there has a FOAF file".... Like 6 hands of maybe 35 people...?

19:30:31 * crschmidt has one

19:30:32 <sh1mmer> yeah

19:30:48 <sh1mmer> I wonder how many people are scard about Eric's info suddenly appearing

19:30:50 <sh1mmer> heh

19:30:55 <sh1mmer> especially eric

19:31:12 * erim is scared

19:31:34 * crschmidt has several million foaf files

19:31:41 <sandro> Are you at the IG meeting crschmidt?

19:31:46 * erim both excited and scared

19:31:47 <crschmidt> Nah

19:31:52 <crschmidt> no need for me to be there :)

19:31:56 <sandro> If you want to could LJ, then I have on too. But I didnt cound it.

19:32:00 <sh1mmer> i would h ave bought you beer

19:32:01 * RRS surprised that erim is surprised

19:32:08 <sandro> If you want to count LJ, then I have on too. But I didnt count it.

19:32:18 * crschmidt counts LJ!

19:32:27 * erim not surprised .... just excited / scared :)

19:32:30 <crschmidt> Those files may not be very full, but they're still FOAF files.

19:32:43 * erim suggests folks try http://eliyon.com some time

19:32:52 <crschmidt> sh1mmer: that would be illegal

19:32:59 <crschmidt> sh1mmer: i wouldn't want you to get arrested

19:33:11 <sh1mmer> crschmidt, oh right you are just a babe arfen'nt ya?

19:33:21 <crschmidt> 2.5 weeks short of 21

19:33:24 <crschmidt> ^dob crschmidt

19:33:25 <sh1mmer> heh

19:33:25 <julie> 1984-03-19

19:33:32 <sandro> No, you're right, crschmidt. I guess mostly I hesitate because (1) I dont have much control over it, and (2) I'm in professional mode today. :-)

19:33:36 <sh1mmer> i am impressed that i haven't been carded once yet

19:33:56 <sh1mmer> maybe it's the beard

19:33:57 <erim> i was particulary interested in seeing how integrated the sw.deri.org results were... lots of stuff i worked on, very little of which I thought was descxribed in FOAF files

19:34:16 <sandro> I noticed that too, eric.

19:34:43 <sh1mmer> erim, maybe youshould go around and hit all the people that are describing your life a little too much :P

19:34:51 * erim lol

19:35:42 <Cloud_> Bring back danbri's transcriptions!

19:35:58 <bandri2> thanks,but sorry can't trascribe everything!

19:36:18 <Cloud_> Ah there you are, well thanks for the ones so far

19:36:32 <sh1mmer> yeah sorry, i specialise in useless social banter more than transcription

19:37:11 <bandri2> marja on annotea

19:37:19 <bandri2> ...on annotea shared bookmarks... 2nd phase of annotea

19:37:31 <bandri2> ...i'm a usability person

19:37:43 <bandri2> ...inspiration was real user problems

19:37:51 <bandri2> ...reviews of specs at w3c was one motivating use case

19:37:55 * Cloud_ transfers soul to Cloud

19:38:07 <crschmidt> sandro: yeah, the "I can't control it very much" is what hits most people would have first in mind when not mentiong those files

19:38:11 <bandri2> (slides are on her laptop, hopefully will be on Annotea site soon)

19:38:39 <bandri2> ...labelling alike to physical bookmarks, tabs in book / scribbles etc

19:38:44 <bandri2> ...shows oriiginal annotea ui

19:38:47 <bandri2> ...pencil mark

19:38:52 <bandri2> ...separate annotaiton window

19:39:05 <bandri2> ...other clients were made too

19:39:28 <bandri2> ...was a technical architecture, used ALgae queries, diff RDF stores (locally and in the server)

19:39:32 <sandro> My college had an app that would give you the top-view diagram of a room with everyones names at their computers. I want that for this room!

19:39:37 <bandri2> ...could combine the page at the client

19:40:11 <bandri2> sandro, we have that for this _chat_room... http://crschmidt.net/semweb/rdfbot/online_users.cgi?channel=swig&image=1

19:40:22 * chaalsBOS notes that not everyone knows how to use bookmarks...

19:40:23 <bandri2> 2nd phase: shared bookmarks and topics

19:40:31 <bandri2> showing running live in firefox

19:40:33 <bandri2> toolbar

19:40:55 <bandri2> [no pagemarks | quick bookmark | configure | search ] toolbar

19:41:50 <bandri2> shows an annotea-extended bookmark view pane in firefox

19:42:07 <bandri2> title, recalls, topics (w/ multiple topics), description, ...related links

19:42:36 <bandri2> (Josek, Aliman... did you guys continue talking re Annotea topics <-> SKOS, after swad-e meeting?)

19:42:51 <bandri2> ...can also navigate from a page that has been collaboratively bookmarked

19:43:14 <bandri2> ..can give a bit more info also when bookmarking, eg. using subtopics

19:43:27 <bandri2> ...show bookmarks in a hierarchy

19:43:47 <bandri2> (some diffs between firefox and mozilla re window layout)

19:43:57 <bandri2> ...click topic links, goes to page

19:44:41 <aliman> yeah jose & I had a chat this morning :)

19:45:03 <bandri2> great

19:45:10 <bandri2> ...can select stores to use

19:45:14 <bandri2> ... or store to web

19:45:15 <aliman> After we met in France I spent a good while thinking about how to make SKOS Core work with bookmarks ...

19:45:17 <bandri2> ...more bookmarks

19:45:35 <bandri2> ...can search the bookmarks

19:45:40 <aliman> in the end I thought that SKOS Core wasn't right for internal representation of bookmarks within application ...

19:45:55 <aliman> however should be easy to do a SKOS bookmark export.

19:46:02 <bandri2> ......searches descriptions too

19:46:26 <bandri2> ...if you guys could write up a brief compare/conttrast would be good

19:46:29 <bandri2> just a mail msg etc

19:46:41 <bandri2> marja: tried to make it v simple so would be used

19:46:47 <bandri2> ...we give uris for bookmarks and topics

19:46:59 <bandri2> ...find different information

19:47:16 <bandri2> ...metadata can be repurposed; linked to user profiles

19:47:32 <bandri2> ...eg. pass that to amazon or other such online businesses

19:47:38 <bandri2> ...for more focussed search

19:47:48 <aliman> bandri2 that 'compare/contrast' comment to me and josek?

19:48:22 <bandri2> yes, sorry

19:48:29 <bandri2> (and marja...:)

19:48:40 <bandri2> ..also experiments w/ Spam classification

19:48:46 <bandri2> ...shows example at w3c

19:49:02 <aliman> josek is the rdf vocab you use for bookmarks published on the web?

19:49:40 <josek> Aliman, I believe so... digging the URL. I think I forgot to update it with the latest version (related to seeAlso)

19:50:07 <bandri2> marja: also link to filtering

19:50:12 <bandri2> ...this is somewhat more gentle filtering

19:50:20 <bandri2> ...people have more choice about filtering out, in, ...

19:50:28 <bandri2> ...can change the groups according to what you're doing

19:50:29 <josek>http://www.w3.org/2002/01/bookmark

19:50:30 <dc_swig> F: http://www.w3.org/2002/01/bookmark from josek

19:50:31 <bandri2> ...hobby vs work

19:51:04 <bandri2> Qs:

19:51:09 <josek> F:|Annotea bookmark vocabulary. Missing an update

19:51:09 <dc_swig> Titled item F.

19:51:21 <bandri2> ...ericm: Annotea related to Simile's notion of a piggybank

19:51:58 <bandri2> q (stefano): how do you submit data back to the repository?

19:52:06 <bandri2> ...do you have some sort of persistent repository?

19:52:15 <bandri2> marja: annotea annotations use annotea servers...

19:52:31 <bandri2> ...we wanted to try out use of rdf files too

19:52:39 <bandri2> (...locally?)

19:52:47 <bandri2> ...experimenting w/ bookmark server, getting it working

19:52:54 <bandri2> ...getting files working w/ the annotations

19:53:00 <bandri2> ...experimentign w/ ericp

19:53:09 <aliman> josek what's updated in the bookmarks vocab?

19:53:25 <aliman> anything added/removed?

19:54:41 <bandri2> danbri: how have you found Mozilla/firefox to code with?

19:54:53 <bandri2> marja: i use their rdf stuff, is good once you learn where docs are, and who to ask...

19:55:23 <bandri2> phil: annotea servers... who runs these? could i make money hosting such?

19:55:28 <bandri2> chaals/eric/marja: yes, could do

19:56:20 * bandri2 not scribing discussion

19:56:28 <josek> Aliman, some few things added for aliasing of names, shortcuts. I need to look at it for some minutes to remember what changed

19:56:38 <josek> no major changes, though

19:57:16 <crschmidt> is there any plans for non-daytime #swig f2f for beers or something?

20:02:15 <bandri2> tom croucher

20:02:24 <bandri2> uni sunderland, The North, England, Europe.

20:02:27 <crschmidt> (aka sh1m)

20:02:49 <bandri2> ...combining properties together to form an Inverse FUnctional Property

20:03:00 <bandri2> ...ie uniquely identify things that don't have URIs

20:03:20 <bandri2> (also called 'unambiguous property constellations' by sandro)

20:03:23 <crschmidt> ( a la geo:Point )

20:03:30 <bandri2> ...exploring with this, use of features of objects to describe them

20:03:40 <bandri2> (yes, chris)

20:03:54 <bandri2> ...ie. composite keys in database [via Frank]

20:04:14 <bandri2> ...in terms of identitfication in real world, we use characteristics of things

20:04:24 <bandri2> ...so we do this in RDFland via a bunch of properties

20:04:29 <sandro> MY calling it that predates OWL, when the term was "unambiguous property" instead of the current "inverse functional property"

20:04:36 <bandri2> ...want to specify a way to represent these clusters of properties

20:04:48 <bandri2> ...lots of ways to do this

20:04:56 <bandri2> ...some dispute (w/ dan, others) re use of ordered lists

20:05:13 <bandri2> phil t: fundemantal maths is set theory, so order should be irrelevant

20:05:30 <bandri2> enrico franconi(sp.?): has a description logic based on ?? ..has some exploratory work

20:05:37 <bandri2> ...we wrote a paper a while back

20:05:38 <bandri2> @@url

20:05:43 <bandri2> ...some bits of owl in there

20:06:06 <bandri2> issues to do w/ identity... you can identify things pretty easily... and describe how to identy things w/ properties

20:06:08 <sandro> I assume: http://www.inf.unibz.it/~franconi/

20:06:10 <bandri2> ...but need to be careful

20:06:26 <eaon> tom and joes paper is at http://osiris.sunderland.ac.uk/~cs0tco/eswc2005.pdf

20:06:42 <libby>http://osiris.sunderland.ac.uk/~cs0tco/eswc2005.pdf

20:06:42 <dc_swig> G: http://osiris.sunderland.ac.uk/~cs0tco/eswc2005.pdf from libby

20:06:44 <bandri2> ...some clear issues of how to describe the way you must use properties together to create something unambiguous

20:06:53 <bandri2> ...granulaity: how specific is the data we're working with

20:06:55 <bandri2> ...the context

20:07:10 <bandri2> ...a restaurant at "21 John Street" when contextualised to Sunderland UK

20:07:18 <bandri2> ...if you want to go into the technical detail

20:07:32 <bandri2> ...looked at a subsumption hieracthy to show which properties relate at which levels

20:07:34 <libby> G:|Situation and Identity A Generalisation of Inverse Functional Properties by Tom Croucher and Joe Geldart (pdf)

20:07:34 <dc_swig> Titled item G.

20:07:41 <bandri2> ...final one is temporality

20:07:45 <bandri2> ...properties change over time

20:08:01 <bandri2> ...need to make sure that you've decided and clearly set down if the properties change

20:08:03 <bandri2> ...vs diff object

20:08:12 <bandri2> ...so need to say which properties have a temporal nature

20:08:19 <bandri2> ...can change one's name, but not one's date of birth.

20:09:01 <bandri2> +1 on that (foaf uses 'static inverse ...'

20:09:11 <bandri2> scribing... can someone pls try capture for logs

20:19:35 <dajobe> ooh IETF announced "The IESG has approved the following document: - 'The 'tag' URI scheme ' as an Informational RFC"

20:19:40 <dajobe> congrats

20:20:00 <sandro> Thanks. I heard this last week. Still don't believe it.

20:20:29 <sandro> Of course it should have been granted informational status without discussion 4 years ago.

20:20:50 <dajobe> yeah

20:20:58 <sandro> After all the discussion it had, it should have been standards track, IMHO, but I'm sure not going to fight over it.

20:23:56 <dajobe> somehow the shortest specs take the longest time

20:24:40 <sandro> They're short enough that peoople actually read them? :-)

20:24:50 <dajobe> nah

20:24:57 <dajobe> that's why you put the good stuff at the start!

20:25:23 <bandri2> hmm agenda screwup, i forget sandro was supposed to be moved earlier :(

20:25:24 <bandri2> sorry sandro

20:25:27 <sandro> But with the big ones, people dont feel entitled to complain, because they didnt study the whole thing? :-)

20:25:37 <bandri2> let's have u first after coffee if that's ok?

20:25:41 <sandro> After break is fine, bandri2 .

20:25:46 <bandri2> phew ;)

20:25:53 <sandro> I miss "danbri".

20:25:59 <bandri2> bandri2 is now known as danbri

20:26:34 <danbri> amit presents 'Meteor-s'

20:26:40 <danbri> .g meteor-s

20:26:48 <phenny> meteor-s: http://seds.lpl.arizona.edu/nineplanets/nineplanets/meteorites.html

20:33:20 * danbri updates http://esw.w3.org/topic/SwigAtTp2005

20:38:05 <aliman> I just posted a guess at mapping between SKOS and BM to mailing list ..

20:38:07 <aliman>http://lists.w3.org/Archives/Public/www-rdf-interest/2005Mar/0000.html

20:38:07 <dc_swig> H: http://lists.w3.org/Archives/Public/www-rdf-interest/2005Mar/0000.html from aliman

20:38:27 <aliman> H:|Tentative SKOS <=> BM Mapping

20:38:27 <dc_swig> Titled item H.

20:57:07 <sandro>http://www.w3.org/2004/12/rules-ws/cfp

20:57:07 <dc_swig> I: http://www.w3.org/2004/12/rules-ws/cfp from sandro

20:57:33 <sandro> I:|W3C Rules Workshop

20:57:33 <dc_swig> Titled item I.

21:20:22 <bandri2> PatrickS shows cool work behind www.forum.nokia.com

21:20:24 <bandri2> slides forthcoming

21:20:37 <crschmidt> ooh

21:20:49 * crschmidt is interested in that now that he's become a ForumNokia addict

21:21:09 * erim thinks the semweb is working.... :)

21:21:40 <bandri2> +1

21:30:03 <erim> patrick: we heavily use owl cardinality and sameas

21:30:15 <erim> ... but very little else

21:31:37 <erim> patrick: re selling this internally... it was more of an evolutionaly buy in

21:32:16 * crschmidt uses the RDF/XML profiles of the phones for stuff sometimes

21:32:22 <crschmidt> ^javaPlatform Nokia6600

21:32:45 <julie> Configuration/CLDC-1.0, Profile/MIDP-2.0, rdf:Bag

21:32:49 * crschmidt watches disks spin frantically to fill mysql caches

21:33:27 <erim> crschmidt, any of this stuff you working on written up?

21:33:39 <erim> re to RDF/XML profiles

21:33:48 <crschmidt> erim: I just use it via julie for asking simple questions

21:33:54 <crschmidt> a la what you just saw

21:33:55 <erim> ah

21:34:18 <crschmidt> I had someone request a dump of all the profiles at one point, but he dissapeared and I totally forgot about it until just now

21:34:40 <crschmidt> I did create some scutterplans for the files so I could just sic my little spider on them at one point

21:37:15 <bandri2>http://beghin.inria.fr/

21:37:15 <dc_swig> J: http://beghin.inria.fr/ from bandri2

21:37:33 * crschmidt waves to patricks

21:37:46 <bandri2> J:|KMP

21:37:46 <dc_swig> Titled item J.

21:44:08 <Mkflow> hi

21:45:04 <RalphS_>http://www-sop.inria.fr/acacia/soft/kmp.html

21:45:04 <dc_swig> K: http://www-sop.inria.fr/acacia/soft/kmp.html from RalphS_

21:45:30 <RalphS_> K:| [English] description of INRIA Knowledge Management Platform (KmP)

21:45:30 <dc_swig> Titled item K.

22:05:51 <libby> ++ on using doap...I want to do this for collectiosn of apps and demos in swbp

22:06:19 <libby> in fact I do but I want someone else to create the entries :)

22:06:38 <libby> we're goign to talk about this on thirsday in swbp group

22:06:51 <libby> thursday

22:07:34 <libby> danbri had a cool suggestion for events like conferences to be sent to the list as rdf and then collected - I think this might work for apps and demos too

22:08:07 <libby> talk to me if youre interestedin this

22:08:50 <libby> we're struggling with how to manage the data effectiovely

22:09:54 * PatrickS slides http://lists.w3.org/Archives/Public/www-archive/2005Mar/0002.html

22:09:54 <PatrickS> \a

22:11:05 <libby> ok to chump them patrick?

22:12:02 * crschmidt wonders what his computer will do with a .ppt

22:12:17 <PatrickS> Yes

22:12:37 <libby>http://lists.w3.org/Archives/Public/www-archive/2005Mar/0002.html

22:12:37 <dc_swig> L: http://lists.w3.org/Archives/Public/www-archive/2005Mar/0002.html from libby

22:13:31 <crschmidt> heh. it just whines.

22:14:04 <libby> could someone title that for me? my ppt is playing up

22:14:30 <crschmidt> L:| Slides, From PatrickS

22:14:30 <dc_swig> Titled item L.

22:14:47 * crschmidt is opening PPT now, will retitle once he knows what the title is ;)

22:15:02 <sh1mmer> L:Nokia are really doing it, and then open sourcing it. Go Nokia!

22:15:03 <dc_swig> Added comment L1.

22:15:22 <crschmidt> the powerpoint logo on the mac is kind of cool: orange and glowy

22:15:40 <dc_swig> Titled item L.

22:15:57 <josek> what libby just said makes me think of FOAF for projects... relationships between projects

22:16:48 <libby> thanks patrick

22:17:36 <crschmidt> oh, cool: the phone summaries are actually generated from the RDF

22:17:44 * crschmidt supposes he doesn't have to look at both anymore, then :)

22:23:56 <PatrickS>http://swdev.nokia.com/fnsearch

22:23:56 <PatrickS>http://swdev.nokia.com/fnosig

22:24:52 <PatrickS>http://swdev.nokia.com/voc

22:25:09 <RalphS_> K:| English-language description of INRIA Knowledge Management Platform (KmP)

22:25:09 <dc_swig> Titled item K.

22:27:28 <libby> L:testign for patrick

22:27:28 <dc_swig> Added comment L2.

22:27:46 <libby> L2:""

22:27:47 <dc_swig> Deleted comment L2.

22:28:47 <PatrickS>http://swdev.nokia.com/fnsearch

22:28:47 <PatrickS>http://swdev.nokia.com/fnosig

22:28:47 <dc_swig> Added comment L2.

22:28:54 <crschmidt> PatrickS: No phones with "unlimited heap size" ?

22:29:23 <PatrickS>http://swdev.nokia.com/voc

22:29:23 <dc_swig> Added comment L3.

22:30:05 <libby> L2:""

22:30:05 <dc_swig> Deleted comment L2.

22:30:08 <libby> L3:""

22:30:08 <dc_swig> Comment L3 not found.

22:30:18 <libby> sorry, L3 worked out ok

22:30:27 <libby> ok, now L2 is L3

22:30:39 <libby> L2:""

22:30:39 <dc_swig> Deleted comment L2.

22:30:54 <PatrickS> L:http://swdev.nokia.com/fnsearch

22:30:54 <dc_swig> Added comment L2.

22:31:32 <PatrickS> L:http://swdev.nokia.com/fnosig

22:31:32 <dc_swig> Added comment L3.

22:32:08 <dc_swig> Added comment L4.

22:32:25 <PatrickS> L: http://swdev.nokia.com/voc

22:32:27 <dc_swig> Added comment L5.

22:33:11 <sh1mmer> I should have pimpedmy sooon to be semweb'd postcardworld map project

22:33:14 <sh1mmer> ohwell

22:33:33 <sh1mmer> right

22:33:35 <sh1mmer> beers

22:33:40 <sh1mmer> ciao internet land

22:33:43 <PatrickS> L: http://swdev.nokia.com/facets/RDF_Facets.pdf

22:33:43 <dc_swig> Added comment L6.

23:27:58 <tav|offline> tav|offline is now known as tav