Friend of a Friend (FOAF) IRC Chat Logs for 2004-08-20

01:44:15 <crschmid1> crschmid1 is now known as crschmidt

07:22:43 <bblfish> hehe: foaf camp pictures with "depicts beer" :-D

07:28:24 <mortenf> surprise, heh :)

07:29:07 <Treenaks> where?

07:31:53 <mortenf> see http://rdfweb.org/topic/FoafCampPhotos

07:32:46 <libby> heh: http://swordfish.rdfweb.org/photos/2004/08/19/2004-08-19-Pages/Image29.html

07:39:19 * danbri looks for scribe help here

07:39:27 <danbri> mark on FOAFScape

07:39:40 <danbri> - standalone Java foaf browser

07:39:57 <danbri> - two pane + tabs on screen...

07:40:02 <danbri> - local repository

07:40:09 <danbri> browser-like toolbar

07:40:20 <danbri> - shows graphical representation of foaf document

07:40:21 <libby>http://www.iis.uni-stuttgart.de/foafscape/

07:40:27 <danbri> - classes rendered as different icons

07:40:41 <danbri> right pane = graph thingie; left pane = detailed overview

07:41:09 <danbri> - includes basic html browser via Java component

07:41:24 <danbri> - boxed icons are linked

07:41:48 <danbri> - layout algo needs improving (he says)

07:42:01 <danbri> - can include images about the persons, specify label of the node

07:42:14 <libby> - ohh, pictures of people

07:42:20 <ephidrina> layout of box on left needs work ;)

07:42:22 * danbri wonders if jordan here yet

07:42:26 <dajobe-lap> (it's all orange t-shirts here)

07:42:30 <danbri> - database in background

07:42:39 <danbri> - gets startign dataset from a google query

07:42:46 <dajobe-lap> yes, jordan's to your right

07:42:48 <danbri> - tried "allinurl:foaf filetype:rdf daterange: etc

07:42:50 <dajobe-lap> dajobe-lap is now known as dajobe

07:42:53 <libby> yarden?

07:43:23 <danbri> (someone ask him if he's ok being up next, short cwmclone talk?)

07:43:42 <danbri> - shows HSQL database manager

07:43:54 <danbri> - over jdbc

07:44:19 <danbri> - search function currently quite basic; searching for a pattern inside a property

07:44:25 <danbri> - planning to integrate rdql

07:44:56 <danbri> - (?x foaf:name "Mark") finds files which contain triples where the name contains "mark"

07:45:53 <libby> - filters

07:46:02 <danbri> - future things; impliment layout improvements, also discover/represent incoming links in database

07:46:17 <danbri> - share local repository in a p2p-ish way, so information crawled only once

07:46:35 <danbri> - there's an export, exports URLs of the files you have

07:47:27 <danbri> Qs: is it based on Jena?

07:47:28 <danbri> - yup

07:47:43 <danbri> Jeen: How is HSQL? we did some tests but were dissapointed

07:47:57 <danbri> - up to 100mb is ok; tried 2-300mb you get memory problems, it gets slow.

07:48:16 <danbri> - but there are options re server; can be in-process, or on different machines

07:48:27 <lsimons> dan: is it open source

07:48:39 <lsimons> not yet, but plan is gpl

07:48:45 <lsimons> first clean up :-D

07:48:51 <libby> late sept/early oct

07:49:11 <lsimons> dan: if you encounter special kinds of relationships or nodes, can you customize those?

07:49:17 <lsimons> like "collegue"

07:49:19 <danbri> - checking mortenf's example for xfn

07:51:03 <lsimons> you can see them on the left hand side

07:51:09 <lsimons> what about the graph?

07:51:09 <danbri> - things show up w/ 'null' for namespace prefix

07:51:52 <danbri> - in graph, 'drankBeerWith' shows up fine :)

07:52:20 <lsimons> and additionally plans to use OWL...

07:52:34 <danbri> - uses last portion of url for label, rather than getting schema

07:53:23 <danbri> damian: if you're getting the schema you can use schema info, i18ns better

07:54:17 <danbri> - it works at the instance level rather than schema

07:54:27 <mattmcc> Although rdfs:labels in schemas don't always make friendly labels for applications..

07:54:30 <LeoS> I think this may well integrate with gnowsis

07:54:53 * ephidrina thinks fantasy foafware ...

07:55:10 <libby> ----Jeen, sesame

07:55:15 <danbri> (someone scribe? ah thanks)

07:55:18 <libby>http://openrdf.org

07:55:30 <petermika> open-source rdf database

07:55:35 <libby> sesame is open source RDF db with rdfs support

07:55:55 <libby> originally EU project; then open sourced at Aduna

07:56:52 <libby> [seems to be a network problem wrt the demo for some reason]

07:57:45 <libby> - architecture: JDBC/RDBMS - postgres, mysql; inmemory, and native storage, binary format

07:57:54 <libby> are the backends

07:58:16 <libby> - SAIL storageand inference layer - abstraction, handles rdf resaoning bits

07:58:54 <libby> - on top of that upload, query (SeRQL) RQL, export, admin

07:59:13 <libby> - above taht access api, graph and reppository apis

07:59:57 <libby> - can deploy as a server, commuinicate via api local or via http; or use it as a library

08:02:22 <libby> - description of API

08:02:42 <libby> - SERQL: see ch 5 in the user manual

08:03:49 <libby> - slightly different syntax from RDQL; <uri> {node} ;= splitpoint

08:04:40 <libby> - e.g. find people who know people (have a knows relation)

08:05:07 <libby> danbri: what does 'distinct' mean in this case?

08:05:37 <libby> jeen: jeen knows several poeple - with distnct just get unqiue results on the resultset

08:05:53 <libby> danbri: rdf doesn't guarantee that these are different people

08:06:53 <libby> - comparaison operators. also has contrust query - result is an rdf grah or a transformed RDF graph

08:07:34 <libby> danbri: can you look inside xml literals?

08:07:57 <libby> jeen: no, and people haven;t asked: - let us know!

08:08:34 <libby> - separate download of the parser (RIO)

08:08:41 <libby> [I use this :) ]

08:09:09 <libby> - small lightweight, does Ntriples, turtle...[missed some]

08:10:07 <danbri> (next talk? OWLchestra?)

08:10:12 <libby> - remote http api, same functionality as local library - fairly trivial to replace one with the other

08:10:41 <libby> - repository API doesn;t do v fine-grained access; graph api allows fine grained, but you do that on a local graph.

08:11:11 <libby> libby: implementing DAWG stuff?

08:11:40 <libby> jeen: thinking about BRQL - already have the functionality in serql

08:11:54 <danbri> (the same functionality; not BRQL itself)

08:16:20 <libby> --owlschestra - benjamin Nowack

08:16:35 <libby> - a vocabulary editor built into a web publishing system

08:17:09 <dajobe> sesame native store doc http://www.openrdf.org/forum/mvnforum/getattachment?attach=3

08:17:15 <dajobe> off http://www.openrdf.org/forum/mvnforum/viewthread?thread=179#752

08:18:04 <libby>http://www.appmosphere.com/pages/en-owlchestra

08:18:35 <libby> - nice aspects of web editor e.g. multi-user support for free, no sync problems

08:18:50 <libby> - permissions, groups etc

08:20:08 <libby> - could do validation, forms generation, generate rdf/xml, auto-add typing triples, e.g. autoadding superclasses you've used subclasses of so other scutters don;t have to do inferencing

08:20:27 <libby> - 'site kit' - php-based web publisher

08:21:22 <libby> - used it for the w3photo vocab

08:23:23 <libby> - http://bnowack.de/w3photo/

08:24:58 <libby> - forms-based, drop down lists for classes (max 50 or so)

08:25:51 * dirkx Does anyone have a nice picture on line of a tshirt ?

08:26:33 <danbri> there's a closeup in http://moblog.nicecupoftea.org/ but a bit dark

08:26:58 * mortenf looks for one

08:27:01 <libby> - documenttaion section, can include text and bits of example code etc

08:28:02 * danbri wonders abotu a public browser for foaf+dc+rss+cc+musicbrainz+speaksreadswrites+zodiac+xfn+geo etc

08:28:11 <libby> - can add detailed information about serializing, spacing, xml/rdf etc, rdf:ID

08:29:34 <libby> - generate custom html

08:29:50 <libby> - linked documentation generated

08:30:17 <libby> - integrate example, klinks in the docs

08:32:38 <libby> - html part has components, e.g. images

08:33:18 <libby> - and these page items have rdf/xml representations

08:33:52 <libby> (yay!)

08:34:56 <libby> danbri: if you'rea passive user of ontologies and you want to navigate around lots at once? do you have an example to play with?

08:36:16 <libby> - no but simple to program extensions

08:36:43 <libby> danbri: people often ask what viocabs are available; be appreciated thinks.

08:37:05 <libby> ...likes the mixing

08:37:37 <libby> ----alberto - rdfstore

08:38:58 <libby> - lightweieght database system for rdf

08:39:05 <libby> -first perl, now C

08:39:15 <libby> - rdql support, plus additiions

08:39:39 <libby> - orginally stanford api, now a bit more OO

08:39:46 <libby> - built in uniciode support

08:40:03 <libby> - compiled on linux-like machines

08:40:17 <libby> - hashed stoage, key values, berkeleey db

08:40:27 * danbri wonders about windows

08:41:02 <libby> - tcip storage daemon

08:41:08 <libby> [missed some sorry]

08:41:17 <libby> - detail of tables: 10

08:41:24 <danbri> we should get ppl's presentations linked in from agenda, for record

08:41:48 <libby> - datatypes, lang, contexts

08:41:50 <libby> yeah

08:42:13 <binzac> and context is used in all ,my hacks :-)

08:42:16 <libby> - indexing: abstract it into a huge bitmap

08:42:43 <libby> - similar for properties and strings

08:43:34 <libby> - blindly atm: not internationalization or stop words yet. seems to work with chinese though

08:43:36 <balbinus_> balbinus_ is now known as balbinus

08:45:26 <libby> - next dawg, brql

08:45:29 <libby> - index geo data

08:45:47 <libby> - ultimate dream: joijning queries over several stores

08:46:05 <binzac> yes, simple indexing but if works well for simple "shapes", ie checks some lat/long boundaries

08:46:50 <binzac> it is not a dream.. we need it for any application. data is around the world... now hacking with multiple queries but it makes all this more copmplex

08:47:10 <danbri> quads...

08:47:17 <zool> binzac that sounds nice.... i have a postGIS backed store that runs simple spatial queries against nodes with geo properties

08:47:38 <binzac> should talk later...

08:48:02 * zool nods def geowanking little bof or sthing :)

08:48:03 <libby> [I missed that q, anyone want to add it?]

08:48:12 <binzac> we try to make this native in rdql, ie avoid run part of query in rdql and part out in postGIS

08:49:42 <zool> oh thats interesting. yeah i just have the spatial queries behind methods in a net api - don't have a query language implementation in this store - i'd like to find one we could just drop into it ...

08:50:59 <libby> - alberto illustrtaes installation

08:51:30 <libby> - alberto races against the clock

08:52:24 * libby very impressed

08:52:35 <danbri> (who up next? coffee?)

08:52:42 <mortenf> (sounds good)

08:52:53 <libby> yeah cofee

08:52:59 <petermika> I'm up, but would rather take coffee myself :)

08:53:23 * JibberJim votes for coffeee

08:53:25 <libby> yay! - alberto's query works. installed and running in 5 mins!

08:53:33 <JibberJim> :-)

08:53:37 <libby> it's scheduled for nowish; lots of time later

08:54:02 <libby> - rdf templates: www.asemnatics.com/rdflets

08:55:09 <libby> [alberto was working on this last night till about 12. we all v impressed]

08:56:00 <libby> - shows directory with template file, xslt, and rdql query

08:56:01 * lsimons votes 4 coffee 2

08:56:08 <libby> - want nested queries really

08:57:18 <libby> - an example is http://foafcamp.asemantics.org/attendee.html

08:58:42 <libby> - rdql generates xml, and then xslt to html

08:59:37 <libby> q: how do you manage these sparse matrices?

08:59:50 <libby> alberto: use compression libraries; otherwise own code

09:00:21 <libby> dirk: compression is to keep the io low, not to make it smaller in size

09:00:46 <jeen>http://www.asemantics.net/presos/SWAD-E/SWADe-rdfstore.html

09:01:01 <libby> dram, you beat me to it!

09:01:06 <libby> s/dram/damn/

09:01:07 <jeen> :)

09:04:13 <libby> this is agreat photo: http://www.wasab.dk/morten/2004/08/photos/foafcamp/2/image-29.html

09:04:28 <Treenaks> mad hacking :)

09:04:36 <mortenf> yeah, i like it

09:06:25 <libby> scaling: search is linear and simple search is linear

09:06:56 <libby> danbri: in rdql is clause order preserved for execution?

09:07:00 <libby> - not right now

09:07:39 <dajobe> I don't preserve the order either in rasqal's execution engine

09:09:14 <bblfish> out of context: but I see on foaf log that there are a lot of osx users at foafcamp. Does anyone know how to make vi in Terminal the default app for rdf files?

09:09:34 <[GNU1> lots of apple laptops out there in the foaf world

09:09:45 <[GNU1> ahm :)

09:10:21 <libby> - bsd-style license

09:11:08 <libby> dan: have you tried with a foafcrawler?

09:11:30 <libby> - YARC (yet another rdf crawler) - is a plan; use queries to filter

09:11:34 <libby> [nice idea]

09:11:44 <libby> - in perl

09:12:00 <libby> - foaf is messy with the bnodes

09:13:02 <mortenf> coffee now...

09:14:35 <bblfish> I have been looking for the equivalent of xterm -e 'vi myfile.txt', but it is annoyingly difficult to do... ah well.

09:17:07 <[GNU1> bblfish: ctrl-click on the file, then open with... and choose your own little script to open with, mark the "as default" checker

09:17:21 <[GNU1> i am unsure about the labels... german macosx

09:18:19 <[GNU1> bblfish: and if unsure if vi is your favorite app vor editing .rdf... have a look at subethaedit :)

09:19:22 <bblfish> [GNU1: yep but if that won't help you sepcify that you want vi in Terminal to edit your rdf. That will just open Terminal. (the contrl thing)

09:19:24 <bblfish> I'll have a look at subethaedit

09:21:12 <[GNU1> oh... Terminal.app that is... not an xterm

09:21:32 <[GNU1> no, sorry... dont know how to control Terminal :/

09:23:24 <swh> bblfish, gvim?

09:24:03 <bblfish> mmh. yes, but gvim seems to require x. There was a gvim I just downloaded for OSX, but the fonts are so ugly...

09:24:46 <eaon> subethaedit rocks

09:24:56 <eaon> even though i miss vi(m) commands

09:26:07 <bblfish> On OSX one really gets spoiled by beauty. It is difficult to digest something ugly after a while. Just got subetha downloaded...

09:26:58 <eaon> mostly it's not about uglyness really - it maybe seems like it's the uglyness that prevents you from using it - but it's really the non-os-xish (or unusable) feeling

09:27:39 <bblfish> I like good fonts. Subetha does not highlight n3 I notice...

09:28:07 <[GNU1> actually the things that really hit you hard are the X11 apps :/

09:28:33 <[GNU1> like gaim... (the only app on macosx supporting multi user chats on jabber)

09:29:50 <eaon> [GNU1: there's psi, but it's a qt app but it looks like it's faking the os x style ;)

09:30:15 <[GNU1> but also in the category of "wonnabe"

09:30:39 <bblfish> Nothing is perfect.

09:31:10 <eaon> i hope adium is going to have better jabber support soon

09:31:11 <[GNU1> hmm, isnt?

09:31:20 <[GNU1> adium is near perfect :)

09:31:30 <eaon> they're using libgaim anyway

09:31:35 <eaon> yes

09:32:04 <eaon> anyway, breakfast - bye ;)

09:32:07 <[GNU1> :

09:32:08 <[GNU1> )

09:32:22 <bblfish> ok. have a good time.

09:37:23 <danbri>http://moblog.nicecupoftea.org/archives/images/default/2004-08-20_103210_21285_0.jpg

09:38:25 <LeoS> flink

09:38:32 <LeoS> friends and links

09:38:55 <danbri> ||||||||

09:39:05 <danbri> ?

09:39:05 <libby> like the name flink

09:39:10 <libby> ---peter mika

09:39:25 <LeoS>http://prauws.cs.vu.nl:8080/flink

09:39:36 <bblfish> hehe. I like Doug Adams take on subethaedit. [[

09:39:38 <bblfish> Over, believe it or not, something called the SubEthaNet. [...] I really didn't foresee the Internet. But then, neither did the computer industry. Not that that tells us very much of course - the computer industry didn't even foresee that the century was going to end.

09:39:39 <bblfish> ]]

09:39:41 <libby> data mining of socila networks + a web interface

09:39:56 <jeen>http://prauw.cs.vu.nl:8080/flink

09:40:07 <jeen> (no s in prauw ;))

09:40:19 <LeoS> thx

09:40:50 <libby>http://www.cs.vu.nl/~pmika/research/foaf-ws/mining.html

09:41:11 <libby> - scutter...

09:41:54 <libby> - babster, javamail

09:42:08 <danbri> (bibster, I think)

09:42:09 <libby> bibster, even

09:42:10 <danbri> .g bibster

09:42:12 <phenny> bibster: http://bibster.semanticweb.org/

09:42:26 <libby> - rdf format for bibtec

09:42:48 <libby> - emails, foaf profiles and bib refs, sespate stores

09:42:58 <libby> - also webmiining from google

09:43:07 <libby> - geo locations lookup via ws

09:43:24 <libby> - some reasoning - smushing ('dentity reasoning'

09:43:25 <libby> )

09:43:31 <dajobe> smushing++

09:43:34 <libby> - soem fuzzy string matching

09:43:43 <libby> (for names)

09:43:51 <libby> - sesame - owl samesas

09:44:20 <dajobe> (we've used the sesame rules stuff for the skos demo, to write the skos relationship rules)

09:44:23 <libby> - JUNG - a java api for network analysis

09:44:44 <danbri> (interesting re skos/rules; didn't realise)

09:45:14 <libby> - to visaulization briwsing, export to SNA packages

09:45:20 <jeen> (sesame custom inferencing: http://www.openrdf.org/doc/users/ch03.html#d0e611)

09:46:06 <libby> - shows a nice graph

09:46:06 <libby> - and a geo visulaization

09:46:33 <libby> (of europe)

09:47:11 <libby> - points out mysterious NW <->SE line of the geo semweb

09:48:50 <libby> group photo: http://swordfish.rdfweb.org/photos/2004/08/20/2004-08-20-Pages/Image8.html

09:49:10 <libby> weird yellow/oragnge effect that graham predicted

09:49:31 <jeen> - weighted 'knows' relation, based on cooccurrence of names on the web

09:50:03 <libby> - not open, not all foaf data

09:50:31 <libby> [just for the daml type people?] - 167

09:50:34 <libby> 167 peple

09:50:59 <danbri> (semweb conference PC, isn't it?)

09:51:04 <jeen> 167 is a set of 'well known' Semantic Web ppl

09:51:08 <libby> [knows automatically derived?]

09:51:13 <jeen> danbri, sorry, you're right

09:51:14 <libby> ah right

09:51:32 <danbri> shows that conference PCs are clique-based, perhaps?

09:51:36 <libby> - mails they have sent to one another, publications in common

09:51:51 <libby> not foaf-galway I don;t think

09:51:54 <libby> re cliques

09:52:00 <libby> there's at least 2 cliques anyway

09:52:02 <libby> ;)

09:52:21 <libby> - importance of each person to another, degree of closeness

09:52:22 <danbri> its people the organisers know or know by reputation plus a few, generally, I think

09:53:19 * libby spills tea on laptop

09:53:35 <JibberJim> sugary?

09:54:04 <libby> damian: acadmics make their info open, via pubs - would it work for 'normal' people?

09:54:16 <libby> peter: can easily change scutter to add more people

09:54:36 <JibberJim> publications, not pubs, academics don't spend all their time in pubs giving out info, that's a viscious slur libby

09:54:38 <libby> danbri: bloggers might ahev similar characteristics to academics in terms of status, publishing a cross-refs

09:54:42 <libby> heh

09:54:51 <libby> nah, no sugar (phew)

09:55:37 <libby> - preliminary prrogramme swad-e foaf galway http://www.w3.org/2001/sw/Europe/events/foaf-galway/programme.html

09:57:35 <petermika> <comment to pres> you need to set your browser to use JDK 1.4.* to see the applet

09:57:56 <libby> ----jo walsh

09:59:57 <mortenf> (technical problems, please hold)

10:00:10 * mortenf tries to scribe

10:00:35 <mortenf> - geopolitical foaf, interesting use cases for foaf, from the us

10:01:19 <mortenf>http://space.frot.org/geopolfoaf.html

10:01:36 <mortenf> ... civic space

10:01:38 <mortenf> ... indyvoter

10:01:53 <mortenf> ... backed by triple store, with Class::RDF

10:02:11 <mortenf> ... can be exported to foaf

10:02:31 <mortenf> ... lots of advocacy, socal networks

10:02:51 <mortenf> ... fundraising re howard dean campain

10:03:14 <mortenf> ... money buys physical space, canvassing etc.

10:03:28 <mortenf> ... advokit, helps manage groups, campains

10:03:44 <mortenf> (campaigns)

10:04:10 <mortenf> ... copyright req - public data

10:04:15 <mortenf> ... they rule

10:04:39 <mortenf> ... "fund race 2004"

10:04:47 <shellac> (adovkit: http://www.voter2voter.org/?)

10:04:48 <mortenf> ... who in local area have donated to whom

10:05:12 <mortenf> ... data in public domain, geocoder.us etc.

10:05:28 <mortenf> ... in europe: restricted data, copyright laws

10:05:45 <mortenf> ... "they work for you" - MP's by post code

10:06:06 <mortenf> ... proceedings etc. (custom xml format)

10:06:09 <nostromo> in Europe I can go to the bar and ask ,at least in Las Matas :-)

10:06:21 <mortenf> ... some geo stuff integrated

10:06:46 <mortenf> ... europarliament with voting records etc.

10:06:59 <mortenf> ... open, BI

10:07:09 <mortenf> ... distributed queries

10:07:25 <mortenf> ... monitoring of voting machines

10:07:42 <mortenf> q: statistical analysis?

10:07:49 <mortenf> azz: don't really know

10:07:55 <mortenf> (hmm)

10:08:28 <mortenf> ... watching out for legislation, micro voting, online voting

10:08:49 <ephidrina> any news on the rumour that they're calling in UN observers in nov?

10:08:50 <mortenf> ... advocacy dev meeting

10:08:58 <mortenf> ... hoping to do something along these lines for europe

10:09:07 <mortenf> ... public databases etc.

10:09:54 <mortenf> Q: reason for using triple store?

10:10:15 <mortenf> azz; started using Class:DBI

10:10:21 <mortenf> (bah)

10:10:34 <mortenf> ... need to dynamically alter tables etc.

10:10:42 <mortenf> ... some performance problems

10:10:56 <mortenf> ... (after moving to triple store)

10:11:16 <mortenf> Q: do you keep track of relationship origins?

10:11:30 <mortenf> joeldg: no, just the relation

10:11:59 <mortenf> Q: UK data republished?

10:12:08 <mortenf> ... yep, in XML/RSS

10:12:16 <mortenf> ... perhaps some topic maps

10:12:39 <mortenf> danbri: would like something re political (views)

10:13:24 <mortenf> a: in SF lists of people voting for candidates

10:13:30 <gk> Did anyone catch that URL of Jo's links?

10:13:45 <mortenf> gk, it's in the logs, see topic

10:14:02 <gk> mortenf, thanks.

10:14:14 <mortenf> end of jo

10:14:15 <mortenf> -------------

10:14:21 * bengee didn't bring his wiki ID to foafcamp, can't add http://www.appmosphere.com/pages/fc-photos (currently only 3 pics) to http://rdfweb.org/topic/FoafCampPhotos ...

10:14:30 * mortenf adds it

10:14:44 <bengee> thanx :)

10:16:14 * mortenf looks for new scribe...

10:17:15 <mortenf> yarden katz: pychinko: Rete-based CWM clone

10:17:19 <libby> --- yarden katz

10:17:29 <mortenf> (thx :)

10:18:01 <jeen>http://www.mindswap.org/~katz/pychinko-foaf-camp.html

10:18:05 <libby> - motivation - faster reimpl of cwm

10:18:57 <libby> - implement forgy 1982 ('Rete')

10:19:07 <libby> - vs cwm's topolgical sort

10:20:24 <libby> this is great bengee: http://www.appmosphere.com/prod/media/1_short_demos_resized.jpg

10:20:26 <libby> heh

10:20:43 <bblfish> Just to complete the question earlier: you can run the following apple script and it will open the vi in Terminal.

10:20:45 <bblfish> on open inputfile

10:20:46 <bblfish> tell application "Terminal"

10:20:48 <bblfish> activate

10:20:49 <bblfish> do script "vi " & quoted form of POSIX path of inputfile

10:20:51 <bblfish> end tell

10:20:52 <bblfish> end open

10:20:54 <bblfish> Ok. back to RDF... phew.

10:20:58 <gk> Rete complexity is linear on size of input (is that input data or ruleset?)

10:21:05 * bengee has a nice pic stitcher :)

10:21:10 <libby> - rete builds a network of rules at startup and a new fact is filtered through rules network making the least amount of redundant checks

10:21:41 <bblfish> bblfish is now known as bblfish_lunch

10:21:56 <libby> - works esp well when you have: lots rules, lots of recursive application, and a persistent Rete

10:23:17 * zool tries to remember where mortenf's daily gps tracks live, again

10:23:36 * mortenf finds them

10:24:20 * danbri ponders hookability of zool's Perl RDF stuff w/ Albertos... (especially crawler-meets-storage...), wonders if there's a Rete impl in Perl too

10:25:09 * mortenf points to http://www.wasab.dk/morten/2004/08/18-gps.rdf - seealso 19-gps.rdf from yesterday

10:25:21 <libby> - woah re benchmarks!

10:25:41 <zool> there's no proper rules/inference stuff in our store, apart from IFP smushing... would be nice to borrow fucntinality

10:25:44 <libby> - 2000 facts: 21 secs vs cwms 11 mins

10:25:45 <zool> thx mortenf

10:25:56 <zool> spiffy

10:26:59 <danbri> q+ to ask re test suite ideas for rule engine interop?

10:27:10 * danbri misses w3c Zakim bot (manages queues...)

10:28:30 <libby> - why bother? because community of python users

10:28:35 * alberto danbri: yes good idea - our stuff should be pretty straightforward to use from a perl script - I can help out zool if needed - I will update some PODs mans in next days

10:28:53 <danbri> that'd be cool, alberto

10:29:26 <danbri> see http://search.cpan.org/~zooleika/Class-RDF-0.12/RDF.pm and nearby

10:30:04 <gk> danbri, re q+, I'd have thought a start point to get existing developers' test suites, and filter them to common cases.

10:30:27 <danbri> q+ to ask how much code, whether a cwmcloneclone in perl, .js (for mozilla) - how big a task?

10:30:38 <libby> - rdql query impl

10:30:53 <libby>http://svn.mindswap.org/pychinko

10:31:40 <alberto> danbri: we have got RDFStore::Object which is similar

10:32:05 <danbri> there's a crawler in there someplace. will look in break. or zool might have pointers...

10:32:08 <alberto> RDFStore :: Object which is similar to Class :: RDF

10:32:18 <libby> [btw, ILRT mail is borked, likely to be for the weekend: use libby.miller at gmail.com for me instead - ta!]

10:32:30 <libby> yarden: "it's AI!"

10:32:38 <danbri> all: ssssh! ;)

10:32:52 * jeen agrees with yarden :)

10:32:53 <libby> don;t ask, don;t tell ;)

10:33:59 <libby> - work is by yarden, bijan and kendall

10:35:25 <libby> danbri: this is v cool and shows you what you get for going to the library and reading ;)

10:36:08 <libby> danbri: other langauges/

10:36:30 <libby> - most languages have a Rete impl already, e.g. eclipsse for C

10:36:35 <libby> - lisp

10:36:51 * zool checking out RDFStore::Object - sounds spiffy

10:36:54 <libby> - normally don;t have to implement the algorithm - alrreday done

10:36:57 <jeen> (clips, not eclipse)

10:37:02 <libby> d'oh!

10:37:03 <libby> thanks

10:37:35 <alberto> zool: let me know if you need any help - or just add to it -

10:38:15 <libby> danbri: thoughts on interop testing for rdf rules engines/

10:38:27 <libby> - yep we have a small test suite - unit testing

10:39:12 <danbri> [I think I got to move my talk/session after lunch folks; needs more work]

10:39:29 <libby> ok, well we 're a bit short on time anyway

10:39:36 * danbri nods

10:39:55 * libby not v prepared either :(

10:40:34 <jeen> it's stand up pres... make it up as you go!

10:40:59 <libby> heh, yeah

10:41:06 * libby tries to remember the point

10:41:45 <libby> ..of the work

10:42:01 <libby> jeen: does it scale?

10:42:38 <libby> - sorta

10:42:49 <libby> [can anyone be more specific? ;)]

10:44:07 <gk> Thinking more about the "why bother?" point, as opposed to just using CLIPS or JESS or ... I think having an engine that can take account of the idiosychrasies of CWM built-ins and the like is very important.

10:44:37 <libby> damian: rules are easy to understand....[...]

10:45:00 <gk> rules are functional ;-)

10:45:33 <danbri> see http://rdfweb.org/topic/FoafAxioms for foaf rule attempts

10:46:53 <libby> - you can impl owl dl with a set of rules

10:50:03 <danbri> (I think he said OWL Full too, or most of it...)

10:50:08 <jeen> (not owl dl, but a subset)

10:52:37 <libby> sorry

10:52:42 <libby> scribe tiring....

10:52:46 <libby> peckish

10:52:51 <zool> :)

10:53:16 <danbri> mortenf re photos...

10:53:34 <danbri> http://rdfweb.org/topic/FoafCampPhotos ->

10:53:56 <libby> - point out yourself using this tool and send morten the triples

10:54:05 <danbri> -> http://www.wasab.dk/morten/2004/08/photos/foafcamp/2/edit.html

10:55:59 <libby> - tick the ones taht are you; choose yourslef in teh dropdowns at the bottom of the page; or if you're not on thelist go to the next page and you canb add your name and IFP

11:23:37 <sh1m> just ignore me. im not really here

11:28:16 <dajobe> hello

11:29:57 <dajobe> test festival

11:30:16 <nostromo> hi

11:30:32 <dajobe> turn up the volume

11:30:59 <nostromo> I think this is a male voice

11:31:29 * nostromo is testing the festival plugin for gaim

11:36:13 * mortenf waves goodbye - heading home...

11:39:04 <danbri> when we restarting? 2pm?

11:50:15 <[GNU1> i know it's to general question, but.. is raptor the prefered tool to use?

11:52:25 * danbri scribbles more slides

11:52:53 * nostromo can shut it off

11:54:24 <nostromo> is it off?

11:54:28 <nostromo> it is

11:55:15 <crschmidt> heh

11:55:23 <crschmidt> festival is neat...

11:55:25 <crschmidt> although a bit creepy

11:55:42 <xavier> s/creepy/crappy/

11:55:46 <zool> theres no decent free femail voices for festival that i know of

11:56:22 <crschmidt> jess doesn't let me play with it when she's around, cause the voices all creep her out.

11:56:26 <xavier> and it can't do any audio stream processing... as in, play while thinking of what to say next

11:57:05 <zool> i made a festival voice out of my voice once but it came out all crackly and fuckedup

11:57:06 <xavier> i like rab... mmm, british accents

11:57:54 <zool> i like when you get festival to read a very long random stream of words with no punctuation

11:58:38 <xavier> heh

11:58:58 <xavier> i've been off-and-on working on a lojban voice for festival.

11:59:21 <xavier> it's sorta nifty, but it presently lacks any ability to take breaths

12:02:17 <zool> that sounds great

12:03:41 <danbri> can someone take notes?

12:03:49 <bblfish_lunch> bblfish_lunch is now known as bblfish

12:03:55 <danbri> LeoS announching FOAFNaut + Gnowsis intergration

12:03:58 <danbri> demos on screen

12:06:08 <milesdavis> what's the link?

12:08:24 <libby> ---graham on DL and haskell

12:08:25 <libby>http://www.ninebynine.org/Software/HaskellDL/DLExploration.lhs

12:10:09 <zool> (literate haskell)++(++)

12:10:12 <zool> that rocks so hard

12:13:53 <libby> aw 'HUGS'

12:14:05 <libby>http://haskell.org/hugs

12:16:16 <LeoS> gnowsis + foafnaut: http://130.89.140.35:8668/foafnaut/index.html

12:16:34 <LeoS> seealso: http://rdfweb.org/topic/FoafCampGnowsisAndFoafNaut

12:16:40 <eaon> LeoS: hacked it together?

12:16:48 <LeoS> Jup.

12:16:53 <LeoS> hi eaon btw

12:17:02 <eaon> cool very nice :)

12:17:05 <eaon> hey LeoS

12:17:13 <LeoS> michi: is the passed uri open to you?

12:17:33 <eaon> the wiki works the foafnaut url still loads

12:17:44 <eaon> seems like it's going to time out in a bit :)

12:18:00 <LeoS> thats ok, its a javasxcript rdf parser, it takes time

12:18:24 <eaon> hehe

12:18:50 <LeoS> eaon: actually you are one of the milestones in my foaf file :-)

12:19:02 <LeoS> I will add that you know grenz,..,, mom

12:19:15 <eaon> oh and please never give libby the phone again when i'm totally confused and tired ;)

12:19:42 <eaon> so no the first link doesn't work for me :/

12:20:26 <eaon> LeoS: add http://zine.niij.org/data/about-me ? :)

12:20:40 <libby> heh, poor eaon

12:22:07 <LeoS> done

12:22:08 <LeoS>http://130.89.140.35:8668/foafnaut/foafnaut?uri=gnowsis://leo.gnowsis.com/outlook/contact/00000000ECD4B99358B9814B9DAFE2255CD8AE9AE4C32000

12:22:33 <LeoS> but you won't ssee it in the foafnaut rendering yet, we don't follow seeAlso yet

12:23:06 <LeoS> but i polan to hack this. some like "RSS feed of foaf" -> when your foaf file changes, I wanna know and perhaps add the new ones to my knows

12:23:29 <LeoS> so i will start polling all your foafs from now on :-)

12:23:36 <LeoS> all your files are belong to us

12:23:42 <eaon> hehe

12:23:49 <eaon> anyway, link doesn't work here :/

12:26:31 <DanC> loggerf, pointer?

12:26:31 <DanC> See http://www.ilrt.bris.ac.uk/discovery/chatlogs/foaf/2004-08-20#T12-26-31

12:27:34 <libby> heya danc

12:27:41 * DanC wonders how much of the swap test suite pychinko does

12:28:13 <libby> I'll ask him in a bit

12:28:33 * DanC sees borken image http://www.mindswap.org/semweb/foafcamp/rete-network.jpg

12:29:41 <DanC> grumble... in-your-face URL that's not even a link (http://svn.mindswap.org/pychinko/)

12:29:48 <DanC> but cool that the code is released!

12:30:55 <DanC> hmm... Makefile refers to run-sc.py ; I don't see it

12:32:38 <DanC> main seems to be http://svn.mindswap.org/pychinko/pychinkafon.py

12:32:43 <DanC> back later, perhaps...

12:36:21 <LeoS> just entered this to gnowsis, eaon: [zelt] [knows] [mõks], [grenz]. [mõks] [knows] [zelt],[grenz].

12:37:51 <milesdavis> DanC: link fixed

12:38:26 <milesdavis> DanC: the main code is "main.py" :>

12:39:23 <LeoS> eaon: check out the monochrom gathering on http://130.89.140.35:8668/foafnaut/index.html

12:39:54 <alberto> Libby & Daniel presented http://www.w3.org/2001/sw/Europe/events/foaf-galway/papers/pp/validating_rdf/

12:41:32 <JibberJim> s/daniel/damey/

12:42:02 <libby> s/daniel/damian/

12:42:20 <libby> ----zac: from FOAF to XOAX

12:42:40 <libby> [zac suggests signing papers with foaf profiles]

12:44:01 * libby 's talk: http://www.w3.org/2001/sw/Europe/events/foaf-galway/papers/pp/validating_rdf/

12:44:21 <eaon> LeoS: that link still doesn't work for me :/ screenshot?

12:44:22 <libby> zac talks about rfid on beer bottles

12:44:35 <libby> - annotating beer

12:44:52 <libby> - the beer-bottle community

12:45:06 <libby> - connectying communities

12:46:14 <libby> - more serious - sharing stuff between countries, data which is itself confidential/public/classified

12:46:27 <libby> - usa 16 agencies, 64 users

12:46:33 <libby> EU: multiply by 25 states

12:46:56 <libby> - e.g. schengen information system, poor data protection

12:47:23 <LeoS> eaon: http://www.dfki.uni-kl.de/~sauermann/2004_08_21_foafcamp/monochrom.png

12:48:56 <libby> [can;t find schengen cost, thouigh sis 2 ~ 33m euro]

12:50:16 <eaon> nice

12:51:14 * danbri catches up, grins as DanC descends into self-parody

12:51:33 <danbri> DanC, they shipped some really cool stuff! maybe supress the url-in-your-face complaint until tommorrow?

12:51:50 <libby> - simpler is better

12:52:09 <libby> - zac finds 'owl and other fantasies' book

12:52:31 <libby> zac: questions:

12:52:56 <libby> - who and what needs a global model? (smaller "global' models like foaf?)

12:53:09 <libby> - is the glovbal model big briewther III ?

12:53:12 <libby> broieth

12:53:17 <libby> brother!

12:53:34 <libby> - performance, optimisation modelling - sclable?

12:53:44 <libby> -too simple?

12:53:49 <libby> - below dinosaur line?

12:54:08 * alberto libby: where can I put the link of short preso this morning?

12:54:31 <libby>http://rdfweb.org/topic/FoafCampProgramme

12:54:38 <alberto> cheers

12:54:53 <libby> - who wants to play?

12:55:24 <libby> 1. proposal for rdfs and foaf for organising and sharing intelligence and security information (research)

12:55:41 <libby> 2. MEATBUS - global tracking of foods etc

12:56:04 <libby> 'BEEFSAFE' ;)

12:57:25 <libby> - next foafcamp - Tuscany?

12:58:12 <libby> - a workshop in Rome in the spring? geospacial data perhaps?

12:58:32 <libby> ~ maybe also combining vocabs

12:58:46 <libby> - 'practical' 'deploying' semweb?

12:59:41 <libby> ----coffee

13:00:17 <crschmidt> next foafcamp - US somewhere!

13:00:20 <crschmidt> :)

13:00:45 <libby> heh

13:12:13 <[GNU1> is there a way using redland to only find_statement() the thing in a specified language?!

13:21:48 <danbri> uncool/transient url for (too many?) slides, http://homer.w3.org/~danbri/foafcamp/foaftalk-export.pdf

13:34:50 <libby> ---final foafcamp session - danbri

13:35:03 <libby> - foaf todo list

13:35:14 <libby> <danbri>uncool/transient url for (too many?) slides, http://homer.w3.org/~danbri/foafcamp/foaftalk-export.pdf

13:35:48 <libby> [hopefully he will link a better url from FoafCampProgramme later]

13:36:24 <libby> dan's dayjob is standardisation: and it's slow, expensive and boring...

13:36:47 <libby> - foaf is experiment in grassrootsd, faster deveelopemnt

13:37:18 <libby> - foaf characterisation is hompages information right now

13:37:41 <libby> - people is too large a topic; people are interesting in relation to other stuff - events, other people, docs etc

13:38:31 <libby> - rdf is an approach for partitioning vocab development to different communities

13:38:56 <libby> - so the images people aren't doing stuff about geo while geo peopel are doing images stuff etc

13:39:12 <libby> - can be used/generated by social networkign sites

13:39:37 <libby> - foaf just uses 'knows' - more detail is not machine-checkable

13:39:58 <libby> - 'friend' is social embarrassing

13:40:12 <libby> [perhaps we easily embarrassed ;)]

13:40:52 <libby> - designed to be extended, foaf files can be linked; some modeling idioms e.g. workplaceHomepage

13:41:15 <libby> dan struggles with laptop

13:42:18 <libby> - a dictionary of terms not a file fromat (this has been difficult to explain)

13:42:32 <libby> - freedom also a big problem

13:42:59 <libby> - nice for the people writing the rdf; crappy for the consumers of rdf

13:43:04 <crschmidt> (FOAFnet aims to solve that problem by setting standards for a specific use case: that of sites wanting to use interop for social networking)

13:43:15 <gk> (Can mix commonly used vocabs with private/grassroots special terms.)

13:43:29 <libby> could use validation techniques :)

13:43:35 <libby> - freedom is important though

13:43:48 <libby> - rdf is foaf's guarantee for decentralized control

13:44:22 <libby> - foaf isn;t a standard, process isn;t a democracy, broad scope means that chartering would be difficult

13:44:47 <libby> - don;t want to spend a lot of time standardizing, not 5 years for 15 properties

13:44:56 <libby> - but we do want something better than we have

13:45:18 <alberto> well said danbri :) bottom-up vs. top-down approach to write vocabularies

13:45:25 <libby> - lots of parallel complimentary vocabs, e.g. musicbrainz, DC, RDF iCal GEO, etc etc

13:46:03 <libby> - how do we get confidence in the stability of these vocabs without spending our lives standardizing?

13:46:37 <libby> - lots of foaf stuff is based on experiences of danbri in DC community since 1997. very cool people...

13:47:09 <libby> - dc predated use RDF; foaf uses it as it's basis

13:47:22 <libby> - "DC community is wonderful" :)

13:48:17 <libby> - DC elements are ok and useful, but scoping is a very hard problem; v open process measnt consensuis was slow and difficult and tricky to extend becaues of the scoping issues. endless metadata harmonization meetings :/

13:49:14 <libby> - lessons from DC: changign teh ns URI is very very very expensive. DC moves to 1.1, different namespace. they got burned, used that, loved to a 'living ns' approach

13:49:20 <libby> s/loved/moved

13:50:47 <libby> - foaf uses living ns approach - the spec will change as we learn more.

13:50:57 <libby> - l;earn from how peopel actually use the properties

13:51:50 <libby> - DC usage chaired by the very excellent Tom Baker; term-level decision record and workflow: not at the document level

13:52:19 <libby> - tom is in best practices group - working on a vocabulary managmenet taskforce to look at this stuff (dan and I are on it too)

13:52:45 <libby> - also SKOS (theasurus)_, foaf, dc, etc etc.

13:53:06 <libby> - stable etc is not really defined; most foaf properties are unstable

13:53:50 <libby> - can we trust foaf? how do people know they can? socila side of vocabulary management. what's 'just enough' standardization?

13:54:17 <libby> - practical issues: is the ISP paid up? do we have the domain name for 10 years?

13:55:22 <libby> - the process for tweaking foaf spec is not completely clear. explaining why, what tests etc.

13:56:09 <libby> - "which bits might change? how did the bits get like that? how do we know the meaning is clearly deinfed? what problems do they solve? are they adequately documented and tested?

13:56:30 <libby> - e.g. DC:creator is pretty tricky

13:56:33 <LeoS> libby: i do a video of it. altready divxing parts of it

13:56:47 <libby> a video of what leoS?

13:56:53 <libby> danbri?

13:56:55 <LeoS> oh dan

13:56:59 <libby> heh, cool

13:57:04 <libby> he'll like that ;)

13:57:34 <libby> - documentation: are there translations?

13:58:14 <libby> - dan can change the meaning of millions of document around the world before we can stop him!

13:58:48 <libby> - but stability is important. and we need to work out, or the big guys will do it all (including the lovely fluffy ones like w3c :)

13:58:54 <crschmidt> heh, i can kind of do that to people's friends pages on LiveJournal - I scrape a bunch of comic feeds that are friended by over 20,000 users

13:59:04 <libby> - not just foaf, all these vocabs

13:59:47 <LeoS> this talk is already on divx at http://www.dfki.uni-kl.de/~sauermann/2004_08_21_foafcamp/

14:00:00 <libby> - things to do:

14:00:29 <libby> - work with swbp to establish conventions for trust in OS-like grassrotts vocab developemnnt

14:00:59 <JibberJim> Is that a beer can danbri is opening?

14:01:10 <libby> - stnadrds track? W3C has lots of process, but not vocabularies really mostly

14:01:11 <JibberJim> surely having a beer in the middle presenting is pushing it a bit!

14:01:22 <libby> oh well we're quite informal here ;)

14:01:34 <lsimons> JibberJim: its actually a standard practice here. The professors do it all the time!

14:01:35 <libby> - DC has got term-centric usage board

14:01:39 <libby> heh

14:01:52 * lsimons thinks its nice to see people sniff up our culture

14:02:09 <JibberJim> hehe, I think I should stick around here tonight, so I need to find a place to stay, any recommendations?

14:02:10 <libby> - so maybe W3C? not sure. more than this though: find opuit how to do it rather than rubberstamping

14:02:46 <lsimons> JibberJim: drienerburght inbetween the foo people

14:02:49 <libby> - need to have a think about the process a bit - write it down. foafcamp and foaf-galway both parts of this process

14:03:07 <lsimons> or try and stay here maybe

14:03:09 <libby> - set asside the perrenial "falling in scope" question

14:03:25 * bengee adds some photos to to http://www.appmosphere.com/pages/fc-photos

14:04:06 * mattb watches a bit of video, wonders about coming over to catch the end

14:04:08 <libby> - instead qs like: has it been rtranslated?are there tools that produce and consume it? does it use sensible modleling conventions? plausible usecase documentation? what does it overlap with?

14:04:15 <libby> heya mattb :)

14:04:21 <mattb> hey hey

14:04:27 <libby> we have about 90 mins of discussion after dan finishes

14:05:01 <libby> - everythign in foaf can be done better elsewhere e.g. people's names)

14:05:15 <libby> - but it is used and real and the Semweb

14:05:20 <JibberJim> drienerburght is full isn't it on friday, that's what the foafcamp site says?

14:05:35 <lsimons> across the road is expensive though

14:05:40 <lsimons> what's it called again

14:06:13 <libby> - W3C is interested in this soprt of area, and PIMs, calendar etc; perrhaps not the best place? but a good place to discuss good ways of doing this stsuff; more open to non-traditional stakeholders

14:06:25 <mattb> jibberjim: yeah, i couldn't get a twin room when i tried to book two weeks ago

14:06:28 <mattb> had to get two singles

14:06:28 <libby> the best western?

14:06:43 <lsimons> yeah

14:06:57 <libby> - lightweight hosting on w3c?

14:06:57 <lsimons> there's a list at http://wiki.oreillynet.com/eurofoo/wikis.conf?EuroFooHotel

14:07:07 <libby> best western is v close

14:07:18 <mattb> yeah, we walked past it on the way here from town

14:08:00 <nostromo> I have a room in Best Western (the expensive one) Broeierd, sharing is not that bad: 75€

14:08:14 <libby> todos for discussion: Semweb best practices, process for community-informaed vocab maintenance; and fixes to spec and website (beta.foafproject.org)

14:08:17 <nostromo> (let's see if I'm doing UTF-8 or ISO-8859-15 with € sign

14:08:24 <lsimons> they still have rooms tho

14:08:24 <libby> yeah, looks good

14:09:15 <libby> - interested in funding - EU? to do some of this as dayjob?

14:09:21 <libby> ---questions:

14:09:44 <libby> yarden: would encourage a foafcore kind of effort - limited subset agreed on; also cater for owl dl people?

14:10:06 <libby> danbri: yees, but not sure how to do it, and don;t want to change the ns. different ways of doing this

14:10:14 <libby> - examples etc

14:10:30 <libby> - Dl thing could be one for discussion

14:10:33 <libby> (loater)

14:11:14 <libby> graham: sceptical about testcases for e.g. rdf ical vocab? opinions?

14:11:39 <libby> d: RDF Core was easy-ish; not claer in these cases what the black and white questions are

14:12:17 <libby> g: might be useful to think about testcases to develop this a bit more

14:12:28 <libby> d: list of sample files is v useful

14:13:12 <LeoS> this talk as DIVX http://www.dfki.uni-kl.de/~sauermann/2004_08_21_foafcamp/

14:13:33 <eaon> LeoS: filming with your webcam? :)

14:13:42 <LeoS> with the canon

14:13:53 <eaon> ah

14:14:11 <libby> stefano: RDf stuff for last 4 months. schemaweb.info was the most useful place he found. shopping basket approach. q: lightweight process for schemas in geneneral, not standardisation for foaf instead?

14:14:20 <libby> danbri: yep, this what trying to say :)

14:14:30 <LeoS> eaon: shall i bring you a foafcamp shirt to vienna?

14:14:53 <LeoS> costs €8.5

14:15:01 <eaon> if libby allows it :)

14:15:07 * eaon does puppy eyes

14:15:12 <libby> heh

14:15:13 <eaon> thats perfect :)

14:15:52 <LeoS> i think the organisers need the €8.5 instead of the shirts :-)

14:16:24 * eaon gives 10 because he couldn't come to spend more money

14:16:28 * alberto yeah leoS, right....

14:16:41 <libby> libby: mixing nsamespaces together is very very hard to document sensibly

14:16:53 <libby> yes please re 8.5 :)

14:16:58 <libby> give us the money!

14:17:12 <libby> stefano: we use owl properties to map:

14:17:28 <bblfish> LeoS: does the divx contain video? Just getting sound here.

14:17:36 <LeoS> ok eaon, deal done. i will advance the cash for you, you sure with 10€?

14:17:42 <libby> danbri: this often works only for a limited context

14:17:59 <eaon> LeoS: yip

14:18:02 <LeoS> bblfish: on my desk it does. the sources where mov/jpg but i virtualdubbed them

14:19:10 <LeoS> eaon: size as small as possible, as far as i know your physical dimensions.... ?

14:19:30 <eaon> LeoS: same as you i'd say

14:19:32 <libby> graham: rfc coming out shortly re schema registry

14:19:39 <golbeck_> golbeck_ is now known as golbeck

14:19:48 <libby> minumum is L leo/eaon

14:20:02 <eaon> thats fine :)

14:20:14 <LeoS> we just checked, i have L

14:20:56 <LeoS> hm, that would be good for a scheme. t-shirt sizes of all attendees of a conference, to help the organisers buy the shirts. We can put this in our foaf files

14:21:11 <gk>http://www.ietf.org/internet-drafts/draft-klyne-msghdr-registry-07.txt

14:21:14 <libby> dirk: RFC/ internet drafts - 2/3 not product of ietf process, just individuals or groups

14:21:17 <milesdavis> where do i get those sexy green swad t-shirts?

14:21:28 <libby> - might be a way forward for stability

14:21:35 <libby> milesdavis: you ask me nicely

14:21:43 <libby> (but I only have a few atm)

14:21:48 <gk> Above URI is forthcoming message header field registry proposal (http://www.ietf.org/internet-drafts/draft-klyne-msghdr-registry-07.txt)

14:21:48 <milesdavis> libby: please? :)

14:21:54 * JibberJim thinks you should sell them libby

14:21:55 <libby> heheh, maaaaybe

14:22:01 <libby> we're not allowed to...

14:22:11 <libby> come and see me afterwards miles, see what I can do

14:22:14 * JibberJim thinks you should sell them without the SWAD logo on the back :-)

14:22:15 <milesdavis> oh really - why not?

14:22:16 <milesdavis> ok

14:22:21 <milesdavis> will do

14:22:24 <libby> dirk: suggests an informational RFC

14:22:29 <libby> EU project has paid for them

14:22:36 <gk> (It's not a "schema" registry ... see libby comment several lines above)

14:22:51 <libby> sorry gk, gettign a bit tired

14:23:12 <libby> rogier: put the best pratice as part of the specification?

14:23:19 <gk> (libby, sure, just making sure the record isn't too wrong.)

14:23:29 <libby> d: think we can ssplit that into SWBP?

14:23:32 <libby> gk: ta

14:23:38 <eaon> LeoS: kingmob == reo ;)

14:23:47 <LeoS> idea: not publish emails of others in oyur foaf file!

14:24:05 <libby> well that's sort of polite, but not a rule....

14:24:16 <eaon> LeoS: you're publishing mine ;)

14:24:28 <milesdavis> libby, fwiw, I'm a french citizen so EU lovesssss me

14:24:36 <libby> heh

14:24:39 <azz> the rule I've used is to use whatever form they've used in their own FOAF file...

14:25:06 <libby> who are you milesdavis? do you really really love the semantic web?

14:25:20 <LeoS> yeah, and I just said here that this has to stop

14:25:20 <kingmob> LeoS: hey :)

14:25:23 * milesdavis = yarden

14:25:28 <libby> d'oh!

14:25:38 <dajobe> lol

14:25:41 <libby> heh, of course yarden, not sure what sizes I have

14:25:50 <milesdavis> :)

14:25:57 * milesdavis pleased by obviousness

14:26:02 <libby> :)

14:26:17 <kingmob> LeoS: eaon sent me - he said you have some foaf camp shirts for sale, maybe?

14:26:27 <kingmob> wow. that sounded like a phrase from a bad espionage movie.

14:26:42 <LeoS> hi kingmob. 10€ per shirt, what size?

14:26:58 <JibberJim> you getting a mark-up there LeoS?

14:27:15 <LeoS> I am the courier to vienna actually

14:27:16 <libby> anyone, feel free to add notes here....

14:27:18 <bblfish> the avis are nice. thanks

14:27:22 <kingmob> LeoS: large, preferably.

14:27:29 <libby> d:foaf shoul cover these examples of usage, and doesn;t adequately yet

14:27:51 <LeoS> knigmob: didn't you used to collect this ring around your belly?

14:27:55 <libby> lliz: maybe use techniques for spam-proof mails

14:28:13 <libby> ...i.e. the good one will be private

14:28:40 <libby> ...the others will be heavily filtered

14:28:45 <kingmob> LeoS: yeah, i became a little fat in spring, but it was just a phase.

14:29:04 <kingmob> i was going for the "bald fat unshaved geek with stains all over his shirt" look

14:29:08 <kingmob> but it just wasn't for me.

14:29:26 <libby> then and now pictures?

14:29:35 <libby> 'before' and 'after'

14:29:35 <LeoS> kingmob: but you were already quite near to the goal

14:29:50 <eaon> haha

14:30:01 <eaon> come down guys ;)

14:30:19 <LeoS> whoa, we just had a power out but luckily these accus work

14:30:27 <kingmob> LeoS: a man whose clothing represents a major violation of the geneva convention and the charta of human rights shouldn't talk about looks :)

14:30:33 <LeoS> otherwise you would have seen a much of "has left"

14:30:57 <LeoS> kingmob: i wanna meet this man!

14:30:57 <libby> dan: favours of owl: owl:uptight and owl:scuffy

14:31:15 <kingmob> LeoS: look at some pictures of the speis, then.

14:31:28 <LeoS> libby: send him the link to his videos on http://www.dfki.uni-kl.de/~sauermann/2004_08_21_foafcamp/

14:31:29 <kingmob> i still have nightmares about your faux jaguar fur coat.

14:31:32 <libby> owl:dl is problematic with foaf

14:32:01 <libby> - esp using mbox_sha1sum

14:32:30 <libby> yarden: DL version would mean not importing DC

14:32:37 <libby> d: don't do that

14:32:59 <libby> y:IFP is problematic with a lot of DL reasoners

14:33:08 <libby> d:we don;t use datatypes

14:33:58 <milesdavis> http://rdfweb.org/pipermail/rdfweb-dev/2004-August/013611.html ron alford

14:34:02 <libby> peter: has done this recently (also ron-something from mindswap's recent mail

14:34:04 <libby> ya

14:34:05 <libby> ta

14:34:25 <libby> dan: foaf is used a lot with non-DL compatible vocabs

14:34:43 <libby> y: perhaps not a big deal; depends what the foaf community decides in terms of reasoning

14:35:05 <libby> d: maybe throw away stuff that isn't DL for them - a version

14:35:30 <libby> d: make Dl reasoners more tolerant?

14:36:13 <libby> d:IFPs are massively useful

14:36:42 <libby> (used all the time in real life e.g. stock-tickers)

14:37:13 <libby> d: tell everyone to use pallette (sp?)

14:37:18 <milesdavis> pellet

14:37:21 <libby> sorry

14:37:37 <libby> thanks

14:37:44 <milesdavis>http://www.mindswap.org/2003/pellet/index.shtml

14:37:59 <libby> d:what do people want to say and can;t?

14:38:04 <pmika> OWL DL version of FOAF: http://www.cs.vu.nl/~pmika/foaf/foaf.owl

14:38:11 <libby> - which version of a manifesto to sign up to

14:38:30 <pmika> see comments for what needed to be changed compared to the OWL-Full version

14:38:44 <libby> thanks peter

14:39:02 <milesdavis> Another OWL DL version of FOAF: http://www.mindswap.org/2003/owl/foaf

14:40:29 <libby> libby: how we going to fix these todos, esp process

14:40:31 <libby> ?

14:40:50 <libby> d: as we develop the website some of these will come out in the wash

14:41:14 <libby> d: irc meetings? - a regular heartbeat for the project

14:41:27 <libby> d: we don;t really have any real tie assigned to it, until v recently

14:41:41 <libby> - every 2-4 weeks perhaps?

14:42:11 <libby> dirk: can;t really hash it out in sufffienct depth on irc

14:42:14 <crschmidt> (more common meetings are often better - gives more people a chance to attend, etc., more drive/purpose on a regular basis)

14:42:15 * bengee would like to volunteer to build a dictionary-like front-end/browser for the foaf terms and their documentation...

14:42:19 <libby> d: but could be a heartbeat for that

14:42:40 <libby> bengee, I'd like to see that v much

14:43:08 <libby> chris,. more often as in every 2 weeks

14:43:09 <libby> ?

14:43:20 <libby> leo: irc has timezone issues

14:43:31 <libby> d: move it around?

14:43:40 <crschmidt> libby: yeah

14:44:11 <crschmidt> Deanspace got good meetings doing it at 7pm eastern time, but they're much more US based, so can't really be used as an example, I suppose

14:44:26 <libby> libby: uncertainly at the moment: not really much time to put into FOAF at the moment :(

14:44:36 <libby> uncertainty

14:44:57 <libby> d: wants next project to be foaf-realted to help with this issue

14:45:11 <crschmidt> developers/etc. work better with goals - what needs to be done to encourage FOAF? can a team get together with knowledge in a certain area in order to do it? etc.

14:45:42 <crschmidt> Meetings can be a good way to meet, say "here is what needs to be done, if you want to help and think you can, split off and do it, we'll talk again in a week, two weeks, etc."

14:45:44 <libby> gk: more formal processes cost a lot of money; attempts to make the process more formal without more cash isn;t perhaps going to work

14:45:49 * milesdavis likes danbri's idea of formalizing the technical, low-level issues of hosting a vocab - e.g. domains/backups/server location/responsibility of maintainer..

14:46:04 <crschmidt> (more observation than specific to foaf)

14:47:03 <libby> [least beer-centric foafmeet ever, believe it or not ;)]

14:47:44 <libby> (a previous meet: http://swordfish.rdfweb.org/photos/2003/01/04/)

14:48:06 <libby> stefano: move open source approach into processes that define something; thionks this is possible.

14:48:20 <milesdavis> dude, it's like the same people !

14:48:31 <pmika> writeup from two months ago on making FOAF DL compatible at http://www.cs.vu.nl/~pmika/foaf/validating.rtf

14:48:59 <libby> d:sounds plausible.....

14:49:07 <libby> scribe missing stuff, sorry

14:49:44 <libby> leos: needs vcard, calendar schemas etc; want to use these and foaf in commercial stuff

14:50:11 <libby> d:round trip through vcard files for foaf?

14:50:50 <libby> zac: apache started with a problem and tried to fix it....start by making something...get some funding for this :)

14:50:56 <libby> ...get something for people to start using

14:51:37 <libby> d:yeah. learn from DC: victim of it's own success in a way: tech stuff didn't keep up

14:51:52 <libby> ...miutated the community - didn;t finish it's initial task

14:52:06 <libby> d: in summary: yes

14:52:31 <libby> d: foaf: "a substandards organisation" ;)

14:53:28 <libby> [another foafmeet: http://swordfish.rdfweb.org/photos/2003/05/30/index.html]

14:53:42 <libby> d: loose bunch of peopel who can work together = foaf

14:54:16 <LeoS> gk: jabber has also a community process

14:54:25 <libby> gk: could look at the jabber community process perhaps? became unweildy, wen tto IETF in the end after commercial interest made it difficult

14:54:44 <LeoS> leo: jabber was not üpeople in charge. here you have people open source & comercial interest

14:55:54 <libby> d: worries he might be a bottleneck; control is important, but RDf is a safety valve for extensibility

14:56:16 * libby wishes DanC was about, wonders what he recons about this sort of thing

14:56:42 * libby personally agrees with dirk that an ietf rubbrstamp might well be the way forward

14:57:09 * DanC waves from a telcon

14:57:14 <libby> the SWBPG work on ideas for versioning etc is going to take a while: it's basically research

14:57:21 <libby> heya danc

14:57:51 <libby> s/rubberstamp/RFC or something/

14:57:55 <libby> ahem

14:58:24 <libby> I guess a w3cn note, if possible, might do the same job...

14:59:41 <libby> zac: (sorry missed that)

14:59:52 <libby> dirk: has been asking poeple; what is foaf to you?

15:00:12 <libby> - vocab/my data/your data/aggregation/transient docs

15:00:17 <dngor> "I have no friends. FOAF is a barren wasteland of angst."

15:00:34 <binzac> just about early stage... we now as in some 92.. much early for reals tandardization.... this is more a "zero phase" :-)

15:00:40 <JibberJim> FOAF is an opportunity dngor

15:01:01 <libby> [missed some stuff sorry]

15:01:25 <libby> dirk: is standardisation going to riun the community, favour some opinions over others?

15:01:30 <libby> ruin

15:02:00 <libby> i.e. I guess it fixes foaf's meaning, thereby alienating people

15:02:07 <libby> potentially

15:02:42 <libby> dirk: if the group shows consensus, stanrd is quick and simple; otherwise could be bad

15:04:23 <libby> d: 'foaf' was in use before foaf....we got traction from a cute phrase ....but perhaps never very clever what it was. lots of people thought file format. added some stuff now: persoinal profile document

15:04:54 <libby> ...lots of realted tech work: scuttering, etc

15:05:22 <libby> ...builds ona nice tight rdf spec, which might help with some issues

15:06:03 <libby> [scribe drifted off sorry :/]

15:06:12 * DanC finishes telcon, tries to figure out what "this sort of thing" referred to...

15:06:29 <gk> (Just to correct the record above... I didn't claim that commercial interest made JAbber community process unweildy, just that moving to the IETF seemed to approximately coincide with some level of commercial interest.)

15:06:30 <libby> ...urn/url @semantics work maybe?

15:06:48 <libby> of dear, sosry gk, I just dont seem to be able to scribe you accuately!

15:07:20 <libby> sory 'this sort of stuff' = informal vocabulary developement

15:07:21 <gk> (Libby, it's OK - just clarifying the record :-)

15:07:25 <libby> thanks :0

15:07:26 <libby> :)

15:07:32 <DanC> not foaf in particular, libby?

15:07:46 * DanC notes the irony in referring to RDF schemas as "informal"

15:08:04 <libby> well, foaf and then others...

15:08:23 * gk DanC, :-)

15:08:31 <DanC> I have much more to say about the specific case of foaf than the general case.

15:08:40 <libby> oh aye? :)

15:08:46 <libby> what do you think?

15:08:51 * libby interested

15:09:05 <DanC> um... well, I guess I have pointed questions more than advice.

15:09:30 <libby> dirk's point: using foaf:// or similar

15:09:31 <DanC> well, my first advice is: develop test cases. that will help with all sorts of things

15:10:10 <libby> we looked at that a little above...danbri's take was that foaf testcases are much more difficult to specify as 'yes/no' testcases

15:10:17 <gk> DanC, that was discussed earlier, not rejected, but there was some concern that some aspects of FOAF aren't really testable.

15:10:26 <libby> yeah

15:10:34 <libby> I dunno though. could bear more discussion

15:10:46 <DanC> yes, it's difficult, but I still think it will help. I think the foaf community has a notion of "smushing" that can now be grounded in RDF+OWL specs, and is quite testable.

15:10:54 <gk> I agree. I think the effort would be illuminating.

15:11:13 <libby> what else danc?

15:11:32 <libby> liz: UI issues are v interesting, difficult

15:11:40 <DanC> pointed question: is foaf more valuable as a stable artifact or an ongoing service?

15:12:15 <ronwalf> Mostly wrt owl-dl and foaf, I've been suggesting that foaf be made /friendlier/ to owl-dl reasoners, and they can ignore what they can't reason over

15:12:18 <libby> depends on your point of view I guess

15:12:26 <libby> +1 ronwalf

15:12:33 <gk> DanC, yes - identity reasoning is clearly testable. But some information less so (how is use of foaf:knows usage actually testable?)

15:12:47 <DanC> developing test cases can illuminate things like "is there sufficient interest to do the sort of engineering work necessary to 'finish' foaf"?

15:13:21 <DanC> foaf:knows seems to be observable from a number of tools; I can imagine useful tests about the combination of rdfs:seeAlso and foaf:knows.

15:13:28 * mattb wonders if it's too late to come over to logica and say hi

15:13:47 <libby> we're out of the room in 15 mins mattb. but we'll be around here

15:14:07 <mattb> ok, i'll come see

15:14:11 <mattb> heh, dajobe advises otherwise elsewhere...

15:14:20 * eikeon wonders if jordan is about?

15:14:25 <dajobe> well...

15:15:05 <libby> danc: there are some foaf users who want a standard asap; I kinda like it as an rdf experiment, and a fun thing. but I'm probably in a smaller community

15:15:06 <DanC> re artifact/service: it's easy to trust artifacts that don't change. you just cache them locally or whatever. it's more subtle to trust that somebody will maintain the foaf vocabulary in a way that benefits the community. but potentially much more valuable.

15:15:29 <libby> okey, yeah

15:15:49 <libby> very interessting questions but will take time to think about....

15:16:01 <DanC> hmm... even artifacts that don't change are subject to various risks: patents, copyright, trademark.

15:16:05 <milesdavis> eikeon: i'm here

15:16:09 <gk> DanC, I won't argue your point; I don't have strong views. I was trying to give a sense of some of the discussion here.

15:16:27 <libby> mattb: apparantly other buikding at 6

15:16:38 <gk> (FoafCamp meeting is breaking up now)

15:16:39 <DanC> so really, if you want lots of people to invest in foaf, you'll eventually need the sort of legal infrastructure IETF/apache/W3C provides.

15:16:50 <libby> we have to leave danc....

15:16:51 <mattb> libby: which other building?

15:17:02 <DanC> perhaps another time, libby

15:17:03 <libby> probly the one you are in diesenburg or something

15:17:25 <mattb> libby: cool, we're all by the bar

15:17:26 <eikeon> milesdavis, didn't know all your aliases I see ;) -- Hi there.

15:17:47 <libby> I think very interesting potentially. did you see the new SWBP taskforce that tom baker is leading in this area?

15:17:56 <libby> cool mattb, see you ina bit

15:18:27 <DanC> which area? talking to me, libby?

15:18:28 <ronwalf> http://www.mindswap.org/2004/owl/mindswappers#Jordan.Katz foaf:alias "milesdavis"

15:18:34 <libby> yep, I was danc

15:18:43 <libby> I'm not sure of the TF's bname...

15:18:47 <DanC> tom baker... Dr. Who? ;-)

15:18:48 <libby> s/bname/name/

15:18:57 <libby> yeah.

15:18:59 <libby> no

15:19:00 <libby> :)

15:19:07 <libby> dublin core guy

15:19:22 <DanC> ah. no, hadn't seen that he's coordinating a new BPDWG TF

15:20:06 <libby>http://www.w3.org/2001/sw/BestPractices/VM/

15:20:11 <libby> not much happened yet

15:20:33 <libby> - it being holidays

15:21:53 <libby> looks like fairly big overlap with what we want

15:21:59 * libby in the way, better go

15:22:03 <libby> l8rs!

15:23:47 <DanC> ciao

15:26:05 * gk waves goodbye...

16:11:35 <bblfish> seems to be an error on the foaf log

16:12:00 <bblfish> Internal server error it says: http://www.ilrt.bris.ac.uk/discovery/chatlogs/foaf/latest

16:13:09 <crschmidt> loggerf, pointer

16:13:09 <crschmidt> See http://www.ilrt.bris.ac.uk/discovery/chatlogs/foaf/2004-08-20#T16-13-09

16:23:23 <lsimons> ANNOUNCEMENT for all FoafCampers: change of plans wrt dinner. Tim Oreilly has invited us all to join them for dinner, so instead of going to downtown Enschede dinner will be in a place across the road from the Drienerburght, and it'll start at like 19:00.

16:23:48 <lsimons> (if for no particular reason you can't make it in time, ask directions at the Drienerburght reception...)

16:37:53 * eikeon wishes he could ssh over for dinner ;)

18:23:07 <joeldg> some major updates http://historyagent.com/

19:29:52 <Jipp> More foafcamp pictures: http://www.kwark.org/Gfx/2004/2004Week34/

20:57:28 <golbeck_> golbeck_ is now known as golbeck

22:57:32 <eaon> gnight #foaf

23:07:00 * Emmy good night all :)