Friend of a Friend (FOAF) IRC Chat Logs for 2004-09-02

03:44:47 <xavier> foaf:dateOfBirth - sup wit' dat?

08:01:07 <teefal> we are on tv

08:01:16 <Cloud> Cloud TV

08:04:42 * teefal is easily impressed with simple technical advances

08:05:42 <marccanter> hey look - it's me!

08:05:43 <shellac_> shellac_ is now known as shellac

08:05:46 <jbond> it's all too loud!!!!

08:05:56 <marccanter> he's very American in that sense

08:06:10 <jbond> But in another sense...

08:06:45 <jbond> foaf. Taking our data back. One friend at a time.

08:07:06 <marccanter> real issues - hhmmmm

08:07:15 <libby>http://www.w3.org/2001/sw/Europe/events/foaf-galway/papers/fp/technical_and_privacy_challenges/

08:07:27 <libby> Joseph Smarr, Plaxo

08:07:27 <marccanter> don't ask what happened to Sean Parker......

08:07:53 <teefal> boycott until mac & firefox version?

08:07:56 <jbond> A firefox version?

08:08:02 <jbond> Using web standards?

08:08:04 <marccanter> your address book is our fair game

08:08:17 <marccanter> there CERTAINLY are NOT 3M people USING Plaxo

08:08:32 <marccanter> there are 3M names that have been rippe doff... and stuffed into.....

08:08:39 <libby> how about logging here, backchat on the unlogged #foaf-galway?

08:09:44 <danbri> makes sense to me

08:10:29 <marccanter> ALL data shoul;d be in FOAF

08:10:32 <danbri> joseph "Plaxo naturally has more stuff than FOAF on its own currently expresses

08:10:34 <marccanter> Danbri and Libby decide

08:10:46 <marccanter> - all in favor -hit the # key 3 time

08:10:46 <danbri> ...we'd like to use FOAF, but how to extend, augment etc., which other schemas to use or invent, ....

08:10:48 <marccanter> ###

08:11:13 <danbri> ...it'd help if FOAF community took a leadership/moderating role, pointers to other schemas, deployed practice, ...

08:11:33 <danbri> ...we'd know what services were trying to represent, and either pull new terms into FOAF or find suitable other schemas

08:11:40 <danbri> ...other challenge: handling permissions

08:11:45 <danbri> ...foaf files currently v public

08:11:59 <danbri> ...diff to Plaxo etc and other services that offer more private data-sharing environments

08:12:03 <libby> [I really think this is an excellent summary of the issues]

08:12:05 <danbri> ...orkut friendster etc too

08:12:06 <teefal> best practice guidelines ... how to extend, what extensions exist, how to interoperate

08:12:12 <danbri> +1 good summary :)

08:12:19 <marccanter> here here

08:12:37 <danbri> ....plaxo allows you to choose what to share, and we'll autogen a vcard at a url for you

08:12:48 <danbri> ...you're creating multiple diff urls for diff foaf files, etc

08:12:56 * danbri duplicating slides, backs off from scribing too much

08:13:06 <marccanter> defined as = hack

08:13:11 <teefal> secret urls have no real protection ... leave trail in digital logs

08:13:59 <teefal> encrypting with all your contact's public pgp key is a management nightmare

08:15:04 <libby> ...distributed authentication would be ideal - but how?

08:15:33 <marccanter> perfect setup for my talk

08:16:58 <libby> ..."clinks" - contact links - personal uris which can be used via pgp to collect information enrypted to you

08:17:02 <teefal> clink "magic url" approach is just shifting the problem around

08:17:22 <danbri> imho FOAF's core businesss is data format, not APIs (of which there'll be many), protocols etc.

08:17:24 <libby> ...managing the privacy issues

08:17:31 <danbri> (data formats are easier :)

08:17:48 * mortenf fails to locate "clink" on the web

08:18:12 <marccanter> here here

08:18:13 <libby> ...sharing foaf from plaxo etc likely to be opt-in

08:18:37 <libby> ...educate the users about the this feature, why useful to them

08:18:56 <jbond> Faced this on Ecademy. You can have it opt out, but you need to severely limit the data exported. Full data export requires specific authentication.

08:18:59 <libby> ...chicken and egg problem - not that many foaf apps

08:19:00 <marccanter> mmmmmm - I love Chicken with my eggs

08:19:08 <marccanter> or was that Eggs with my chicken?

08:19:23 <danbri> Q: Is a FOAF/RDF file with my personal data intrinsically differenet from an XHTML/XML file with my personal data in? Or just easier to scrape/process?

08:19:29 <teefal> opt-in is bad for business ... educating users, supporting it ... ROI for company on optional features is non-trivial

08:19:33 <libby> ...potentially compromising privacy of contacts

08:19:41 <libby> ...not just self

08:19:56 <marccanter> BTW - do NOT assk him what Plaxo's business model is.....

08:20:07 <libby> ...US law: owner of the data decides what to do with it, e.g. if I give you my business card you can do what you want with it

08:20:08 <jbond> teefal: No. Which is why we're all careful about automatic display of personal data in html. Same applies with FOAF

08:20:36 <libby> ...currently plaxo no contact sharing - potential future problem

08:20:48 <libby> ...but the foaf knows is the interesting bit

08:21:21 <libby> ...ecademy: for contacts just name and hashed email, no more

08:21:29 <danbri> ..."just say eough to wire up the network"

08:21:46 <marccanter> we need some centrazlied, 3rd party service -to worry about all this shit - so we can :get on with it"

08:22:17 <libby> ...much plaxo data is not about plaxo members: definitely want people to allow to export that data (which is theirs)

08:22:26 <jbond> Not a problem for foaf aggregators because you get the key and no heresay information

08:22:45 <libby> ...mboxsha1sum is useful but eventually you need to share your real mbox (to get in touch)

08:23:00 <libby> ...Foaf community could describe best practices for this

08:23:02 <jbond> The key is *my* email address

08:23:11 <teefal> "sha1sum not useful" .. but could a forwarding service exist (or does one)?

08:23:30 <Cloud> a central service?

08:23:34 <teefal> even plaxo

08:23:47 <marccanter> that knows nothing about teh data - but an infrastructure facility

08:24:04 <teefal> simply hash between sha1sum and real email, just to protect it

08:24:04 <Cloud> we could sha1sum all e-mails and do some matching ;-)

08:24:17 <jbond> cloud: that's what we all do

08:24:31 <libby> ..sure advocacy for foaf isn;t going to persuade people; some (solvable) issues need fixing

08:24:33 <shellac> Dirk (@semantics) has a DNS system for finding people

08:24:47 <danbri> there is a risk of brute forcing common names + punctuation + @ + list of domain names

08:24:47 <marccanter> coolio

08:24:47 <jbond> shellac: url?

08:24:56 <shellac> erm...

08:24:59 <danbri> .g ddds dirk asemantics foaf

08:25:02 <phenny> ddds dirk asemantics foaf: http://rdfweb.org/pipermail/rdfweb-dev/2004-March/012809.html

08:25:20 <danbri> google++

08:25:43 <shellac> it's like the URN resolution scheme in life sciences

08:25:54 <libby> q: what level of permissions does plaxo use?

08:26:33 <sh2m> sh2m is now known as sh1m

08:26:39 <libby> a:business and personal; business anyone can have it who has mail address; personal have to give it to them personally. Potentially could do much more complex permissions

08:26:41 <teefal> rules to define subsets for distribution and permissions

08:26:42 <danbri> aside: I've just rewritten http://rdfweb.org/topic/FOAFCommunityProcess based on discussions here and at FOAFCamp; will talk about it later in breakouts.

08:27:58 * mortenf reads...

08:29:17 <marccanter> extensibility

08:29:20 <marccanter> permissions

08:29:23 <marccanter> authentication

08:29:26 <marccanter> privacy

08:33:11 <dick_> now we get to hear what Marc *really* thinks

08:33:52 <danbri> marc canter on foafnet foafnet foafnet foafnet foafnet foafnet foafnet foafnet foafnet foafnet

08:34:38 * mortenf finally finds clinksystems.com

08:34:51 <danbri> ...foafnet ...industry consortium... motivated by making products

08:34:59 <danbri> ...not research... take foaf into marketplace

08:35:03 <danbri> ...companies working together

08:35:10 <danbri> ...but companies are about competing for $$$

08:35:21 <teefal> only a certain pool of money ... teefal believes there's enough to go around :)

08:35:29 <libby>http://www.w3.org/2001/sw/Europe/events/foaf-galway/papers/pp/foafnet/

08:35:32 <dajobe> & €€€s

08:35:32 <danbri> ...opensource world... hold hands sing kumbaya... may hold out against the Microsoft and Apple's of this world

08:35:45 <danbri> ...watched sixapart ship a product using foaf

08:35:52 <mortenf> & DKR's

08:36:33 <danbri> danbri interruption: sixapart stuff was really good, and they shipped code

08:36:40 <danbri> marc: marketing... baselevel foaf-based spec

08:36:52 <danbri> ...u can't ask comercial companies to deploy something loose and openended

08:36:59 <danbri> ...strategy is toehold, foot in the door, ...

08:37:04 <danbri> ...hardest part is 1st step

08:37:19 <danbri> ...baselevel spec, + basic UI guideliens, are driven by those issues listed by joseph

08:37:23 <danbri> ...endusers control their own data

08:37:25 <teefal> cannot get companies to follow spec based on looseness... sounds like RDF is in trouble :)

08:37:32 <danbri> ...theme of these talks, "I want my data back"

08:37:41 <teefal> loose is good

08:38:14 <danbri> btw http://www.w3.org/People/Connolly/drafts/web-research for "I want my data back"

08:38:29 <danbri> ...you're talkign about research + individual basis, but we're talking about huge companies

08:38:37 <danbri> ...all the issues joseph brought up

08:38:47 <danbri> ...we have ~30 companies interested but mostly lurking

08:39:07 <danbri> ...vehicle for keeping momentum going fwd

08:39:11 <danbri> ...goals: get foaf adopted

08:39:22 <danbri> ...for me, is re social networks, attributes being moved around

08:39:27 <danbri> ..." what software isn't about people?"

08:39:33 <danbri> ...so crucial to future of software

08:39:38 <danbri> ...baselevel foafnet is

08:39:46 <danbri> name / face / sha1sum / weblink / list of friends

08:39:59 <dick_> <laughter/>

08:40:05 <danbri> ...of all the wonderful things we can do w/ foaf, lets focus on these base things

08:40:11 <danbri> ...notice email is hashed

08:40:15 <danbri> ...how to get things achieved

08:40:26 <danbri> OK I'm not transcribing this slide; will get it off marc.

08:41:03 <danbri> ...u need to have list of sites, know they're compatible, certified etc

08:41:29 <danbri> [I disagree regarding 'certification' registry --- but yes, service discovery is important]

08:41:43 <danbri> ...hot linking --- ongoing synch

08:41:50 <danbri> ..."Where is the 'master' foaf file?"

08:42:02 <teefal> sounds like santa's list

08:42:22 <danbri> ...techs: WS-federated, Liberty alliance, ...

08:42:24 * sh1m wants a pony

08:42:32 <danbri> ...the enterprise world

08:42:34 <dick_> I have a pony

08:42:37 <danbri> ...we downloaded sourceID...

08:42:50 <teefal> <foaf:santaAttitude>naughty</foaf:santaAttitude>

08:42:51 <dick_> did it complile? :)

08:43:02 <danbri> [danbri aside -- nearby: WS-Addressing submitted to W3C, see http://www.w3.org/Submission/2004/05/ ]

08:43:10 <danbri> ...different personas

08:43:18 <danbri> ...I want only the religious right to see my nudie photos

08:43:37 <danbri> ...within a walled garden, can make this work; trick is to make it work between systems.

08:43:37 * mortenf smiles

08:43:37 <shellac> marc wants religious right to see his nudie photosw

08:43:37 <danbri> +1

08:43:41 <shellac> damn!

08:43:43 <danbri> ...lot of discussion about relationships

08:44:01 <danbri> ...whole world... foafnet, peopleaggregator, ... some use existing relaationship namespace

08:44:02 <sh1m> shellac, I am glad I am a commie ;)

08:44:05 <danbri> ...how to put meaning into relationships

08:44:16 <danbri> ...we put a bare min of rdf stuff into the foafnet

08:44:24 <danbri> ...why? I hear a lot of anti-rdf flack

08:44:31 <danbri> ...I'm a big guy and can take it!

08:44:42 <dick_> feed them cake

08:44:43 <danbri> ...but that won't make those catholics eat pork, or something

08:44:44 <shellac> metaphors going south

08:44:51 <danbri> ...so we de-emphasised the rdf fancy stuff

08:44:57 <danbri> ...cos ppl see world thru xml glasses

08:45:03 <Cloud> marc = henry kissinger

08:45:07 <danbri> ...i'm playing henry kissenger role!

08:45:18 <danbri> ...we've had some sociologists working with us, one brought up a good point

08:45:25 <danbri> re import/export, we'll have a checkbox

08:45:28 <teefal> opt-in again

08:45:36 <danbri> ...unless clicked off, your name doesns't go out in someone's foaf

08:45:37 <jbond> foafnet: http://www.socialtext.net/foafnet/

08:45:41 <danbri> (in foafnet export I guess)

08:46:08 <danbri> ...so do you need opt-in per person? eg. ashamed of one particular relationship?

08:46:43 <danbri> ...only viable rep system today is ebay, cos big closed system

08:46:47 * jbond thinks multiple personas is 1995 thinking. One the 2004 net they leak into each other

08:46:52 <danbri> ...big world, that we haven't touched on yet, whic h is foaf groups

08:47:06 <danbri> [I think secrecy is under-rated, re personas]

08:47:13 * mortenf just don't find personas interesting

08:47:14 <dick_> multiple personas don't leak if you wear a tin foil hat

08:47:25 <danbri> ...foaf:Group ... more small networks is realistics

08:47:29 <danbri> ...talk about process

08:47:35 <sh1m> dick_, or take your pills.

08:47:39 <danbri> ...bit more vocab

08:47:47 <danbri> ...business models, plaxo et al aggregating wewb services

08:48:21 <danbri> what's in common re our needs?

08:48:26 <teefal> aggregating web services? not sure about the verb... repurposing them? repackaging?

08:48:47 <Cloud> marfia canter

08:49:04 <danbri> ...I dream of a world where we have our own 'mafia' where you can be in diff groups, flow data, etc...

08:49:18 <danbri> ...equal ms and apple, be rteally big, 'cos we're Us, in control of own destiny

08:49:29 <danbri> ...thinking of typekey

08:49:36 <shellac> teefal: I think like amazon & google services are repackaged together

08:49:38 <danbri> ...bloggers were suffering from blog spam

08:49:49 <danbri> ...early example of auth / security system

08:50:11 <danbri> [note to self, writeup pgp/imap single sign on proposal]

08:50:46 <danbri> ...I spent a lot of time grokking this stuff, unless there's a company around it, someone motivated, things don't happen

08:51:01 <teefal> unless company wakes up in morning and does single-vision stuff.. things won't happen (like the web for instance :)

08:51:03 <danbri> ...need ppl doing these things, and skip seems to have it

08:51:13 <jbond> s/skip/sxip/

08:51:38 <danbri> dick hardt

08:51:42 <danbri> ...activestate, w/ oreilly

08:51:59 <danbri> can someone scribe so i can listen?

08:52:06 <danbri> brain and hands don't work at same time

08:52:13 <danbri> ...i sold activestate last fall

08:52:20 <danbri> ...since then focussed on making skip happen

08:52:20 <mortenf> goal of sxip:

08:52:24 <danbri> ta!

08:52:24 <jbond> too may passwords on net

08:52:28 <mortenf> privacy issues

08:52:52 <mortenf> (tech pause)

08:53:14 <jbond> Apple laptop problems!

08:54:07 <mortenf> we've got go!

08:54:22 <mortenf> ... repetetive registration

08:54:39 <mortenf> "on the internet, nobody knows you're a dog"

08:54:48 <mortenf> ... member sites, home sites, root site

08:55:13 <mortenf> ... home site: trusted by user, used by used, e.g. plaxo, ecademy

08:55:25 <mortenf> ... different personas

08:55:44 <mortenf> ... member sites: "sxip in", info from home site(s)

08:56:02 <mortenf> ... users select information to release

08:56:15 <mortenf> ... tech: uses shared cookie

08:56:26 <mortenf> ... and dns

08:56:42 <mortenf> ... GUPI, globally unique persona identifier

08:56:52 <teefal> very foaf sounding... "a guppy"

08:56:58 <jbond> Auto-discovery of web service entry point via DNS TXT record. Very cool

08:57:26 <dajobe> similar to the DDDS that Dirk @semantics showed a couple of times

08:57:27 <mortenf> ... trust not necessary between home and member sites

08:57:28 <jbond> GUPI makes me think of New Orleans Voodoo Loahs.

08:58:00 <mortenf> ... one home site is authoritative for one GUPI

08:58:16 <mortenf> ... possible to use diff homes

08:58:25 <mortenf> ... user-centric system

08:58:37 <mortenf> ... includes privacy management

08:58:53 <mortenf> ... uses p3p

08:59:02 <teefal> p3p-like

08:59:07 * jbond wonders. homesite.com.sxip.net is used so that there's a common domain for cookie passing via the browser. Could the FOAF Soup javascript Hack from yesterday be used?

08:59:47 <dajobe> the diff here is that the user doesn't need to be a DNS expert to set up a DDDS thing

09:00:10 <mortenf> ... any site can be a home or member site

09:00:29 <mortenf> ... simple, extensible, low barrier of entry

09:01:57 <mortenf> ... business model: user: free, dev's: os license, member: domain name, home: certificate

09:01:58 <jbond> AFAIK, integration with SXIP only requires ability to set redirect headers. So any scripting language.

09:02:40 <mortenf> ... timeline: demo is online, dev kits in the pipe, java last in september

09:02:55 <mortenf> q (marc): old site today?

09:03:28 <dajobe> https://sxip.org/

09:03:29 <mortenf> a: still old site, wanted to try it out

09:04:07 <danbri> PANEL 10-10.45: I want my data back" - panel: Julian Bond, Marc Canter, Joseph Smarr, Andreas Harth, Timothy Falconer, Dan Brickley (moderator)

09:04:22 <mortenf> q (tom): what about foafnet with foaf not over http?

09:04:55 <mortenf> a: only http, personal home site possible (localhost)

09:05:16 <mortenf> q (henry): can you tell it to trust other sites?

09:05:45 <mortenf> a: different sites say/know diff things

09:06:10 <mortenf> ... can trust slashdot on slashdot membership info

09:07:05 <mortenf> q (?): how crucial is dns in sxip?

09:07:20 <mortenf> a: it works, scales

09:07:45 <mortenf> ... uses dns to find end point

09:07:49 * danbri suggests Panelists make their way down to front during Qs

09:08:03 <danbri> 10:04] <danbri>PANEL 10-10.45: I want my data back" - panel: Julian Bond, Marc Canter, Joseph Smarr, Andreas Harth, Timothy Falconer, Dan Brickley (moderator)

09:08:40 <mortenf> q (joseph): how are foafnet getting people/companies on board?

09:09:08 <mortenf> a: "don't lock people in"

09:09:22 <danbri> aside possible goal: an explict foaf+rdf/outlook mapping

09:09:38 <mortenf> ... "the rise of the About Me page"

09:09:38 <sh1m> orple

09:09:45 <mortenf> heh

09:10:19 <mortenf> ... we want to have something to show, then pressure...

09:12:02 <Cloud> panel session

09:12:04 * mortenf looks for new scribe

09:12:18 <sh1m> i just wanna be able to irc on marc lovely pate

09:14:12 <danbri> can someone scribe?

09:14:32 <sh1m> trying to lock someone into a closed env is a waste of time

09:14:50 <sh1m> the easier we make to let people move the data around the faster the net will grow

09:15:13 <JonR> Yesterdays photos are up on http://storymill.net

09:15:21 * Arnia prods sh1m to scribe

09:15:22 <sh1m> how do you create an env where lots of companies (who aren't ms) make it their business to create a network

09:16:01 <sh1m> ...if are all work together and are open, then we get a much better result

09:16:38 <sh1m> danbri: do you think its possible to get end users to push demand with features

09:16:49 <sh1m> dick: yeah users will push and it will snow ball

09:16:56 <sh1m> danbri: even without ms and google?

09:17:22 <sh1m> dick: most people on the net don't know what blogging is

09:17:24 <sh1m> danbri: but they will

09:18:35 <sh1m> dick: right but it starts up on the edges

09:18:36 <sh1m> joseph: the more you can get data, the better

09:18:36 <sh1m> joseph: the public perception is important, privacy, not locking customers in is really important

09:18:53 <sh1m> joseph: maybe foaf is a vehicle for selling that as the way to avoid lock in

09:19:07 <sh1m> ... it might be hard to convince vc's that there is a business model without the lock in might be a problem

09:19:20 <sh1m> ... but there is still space for a business model with feature growth

09:19:55 <sh1m> timothy: we are seeing a really aprehention about seeing products that a closed

09:20:23 <sh1m> ... these things might be around in 50 or 100 years and if some company is closed and goes out of business

09:20:36 <sh1m> ... loosing that amount of anotation on photos is a big issue

09:20:36 <libby> ..."it's like losing your photos in a fire"

09:20:40 <libby> tim++

09:20:50 <sh1m> ... the only safty is to spread it around

09:21:36 <sh1m> andreas: it important that all the data locked in weblogs or usenet servers but being able to wrest control would be ideal

09:21:52 <sh1m> ... a more broad data model using foaf where you can store and access and control the data

09:22:24 <sh1m> ... i want to be able to authenicate my friend for providing this access. i don't know if i would able to pay each person for that

09:22:46 <sh1m> marc; the point of foaf net is not to perminently limit the data but to make incremental steps

09:23:18 <sh1m> ... dick and i have only know each other a month but dick is not here to extract money from the little guy its about enabling participating

09:24:14 <sh1m> ... if you support and sponsor the page we get a logo at the bottom of the page

09:24:53 <sh1m> julian: is it a stick or a carrot? how do we get foaf adopted by organisations, corps and jsut people

09:25:05 <sh1m> ... helps to know how rss and blogs started

09:25:25 <sh1m> ... there a huge number of people who visit blogs but woudln't know how to make one

09:25:40 <sh1m> ... rss is on the bbc google and newsreaders are pretty mainstream

09:26:13 <sh1m> ... right now we are concentrating on generating foaf and encouraging use of foaf, but there aren't so many applications which use it

09:26:23 <sh1m> ... without the killer only app we are making write only data

09:26:43 <sh1m> danbri: i wanted to touch about consumer apps, we have touched on vcards and ical

09:27:12 <sh1m> ... its really grown into something substantial and we haven't really kept up, we want to make it into a good OSS

09:27:27 <sh1m> s/OSS/open source project/

09:27:47 <sh1m> ... my boss is asking should we make this a standard

09:28:13 <sh1m> ... i am here to ask should w3c get involved and what can we do

09:28:32 <sh1m> ... should we use vcard or replicate and map?

09:29:19 <sh1m> joseph: i find astonishing we have been developing ways to describe addys for 4000 years

09:29:27 <sh1m> ... every country has its own informal standard

09:29:34 <Cloud> s/joseph/julian

09:29:42 <sh1m> ... ive come to the conclusion of something which is a subset of vcardf

09:30:00 <sh1m> ... happy to use the vcard as an extensino

09:31:07 <sh1m> dick: i come from a pragmatic background and the question is who cares, and that is where people who have an itch will scratch it

09:31:35 <sh1m> ... for foaf to be successful you should to find the people with the use cases and get them to fix it

09:31:42 <sh1m> joseph (really this time)

09:32:03 <sh1m> the vcard is a good place to start, and outlook actually keep a lot more info that vcard does

09:32:28 <sh1m> apple is also based on vcard, and they have extended vcard for im and deferentiation

09:32:45 <sh1m> vcard does have some issue, and we could do a better job of articulating the format

09:33:05 <sh1m> plaxo has some NL expertise and that could be used to breaking up and looking after addresses

09:33:42 <sh1m> maybe deserves its own namespace, but i would be happy with it in foaf

09:34:04 <sh1m> timothy: i am happy mixing things in

09:34:26 <sh1m> ... but developers are having issue with the rdf paradigm but they see "this isn't foaf"

09:34:45 <sh1m> ... while they can gradually make the shift, but if it were all there it would help speed things up

09:35:09 <sh1m> danbri: with rss 1 for example, we stripped everythign into a bunch of namespaces, it was political

09:35:32 <sh1m> danbri: its nicely decentralised but at the same time we took it a bit too far

09:35:55 <sh1m> joseph: the harder you make it to make a tool to write this stuff the less people are going to get involved

09:36:23 <sh1m> joseph: making it like sxip where there is just a little bit of codce you can copy and paste and woosh off you go

09:36:37 <sh1m> danbri: people should look at dajobe's position paper

09:36:50 <sh1m> danbri: i dont want it in the spec without tool support

09:37:05 <sh1m> andreas: from a technical perspective i dont care as long as there are mappings

09:37:19 * danbri thinks more Qs from the audience overdue, looks for hands

09:37:26 <sh1m> andreas: there is also a lot of data available in legacy formats

09:37:45 <sh1m> andreas; from a modeling point it would be nicer

09:37:58 <sh1m> marc: when i think of foaf it should a wrapper

09:38:25 <sh1m> marc: built in would be good

09:38:37 <Cloud> gc (Galway Core)

09:39:35 <Cloud> sh1m: too many namespaces in foaf

09:40:11 <Cloud> (referenced)

09:40:49 <sh1m> julian: maybe the main part of foaf is providing encouragement for consicenty

09:41:07 <sh1m> ... we are not going to stop people extending it but we can encourage best practice

09:41:24 <sh1m> ... foaf has issues with being the back bone but it doesn't need to reinvent an address namespace

09:41:55 <libby> [what's missing from these breakout sessions? http://rdfweb.org/topic/FoafGalwayBreakoutSessions ]

09:42:10 <sh1m> danbri: w3c is good at doing things like big solid standard, foaf is good at getting things out there

09:42:44 <sh1m> ... the foaf part of my job is having fun with you guys, and w3 part is gettting those standard there

09:42:48 <dan-uk_> libby: RDF on the Desktop?

09:43:00 <libby> ah yeah

09:43:06 <libby> were there any notes form that?

09:43:07 <sh1m> doing this in the foaf namespace you might as well be using xml

09:43:07 * libby looks

09:43:12 <sh1m> who said that?

09:43:38 <sh1m> you do get the value from using the ontology

09:43:50 <Cloud> (bblfish)

09:43:59 <bblfish> yes

09:44:12 <sh1m> danbri: you can mix data freely, do you change the specs a little to fix stuff

09:44:32 * edd scrolls back, notes killer app is overrated and anyway not what a decentralized tech needs

09:44:52 <sh1m> henry: if a person has an mbox, which maps to this space in the vcard or create a special ontology

09:45:24 <sh1m> bblfish: if you provide subproperties you can make that work

09:45:41 <dick_> please shout your name

09:46:11 <sh1m> julian: plee to the tool developers, i dont want to have to write that stuff, when we are selling foaf we are selling rdf. we need some help.

09:46:54 <sh1m> dick: as a previous tool developer, tim brey phoned us up and asked what do we need to support "xml in perl" if you want to get foaf out there talk to the tool developers so that they can work with foaf

09:47:25 <dick_> s/tim brey/Tim Bray/

09:47:31 <sh1m> joseph: build cool applications, that runs off foafnet. but i dont want to reinvent and use

09:47:41 <sh1m> ... use it so that people can plug into

09:47:58 <dajobe> ObPlug redland www.redland.opensource.ac.uk - LGPL/Apache2 license

09:47:58 <sh1m> danbri: its not just if the def. of the spec is ok, its whether its used an deployed

09:48:11 <pmika> panelists: rdf is still too hard to sell it... (People think it's XML...) dont' even mention OWL.

09:48:24 <pmika> message to tool developers: let me focus on making interfaces, not ontologies

09:48:30 <sh1m> are you guys looking at ldap, active directory integration

09:48:36 <sh1m> joseph: yes.

09:48:40 <pmika> build cool apps to support foaf!

09:49:11 <mortenf> coffee break...

09:49:11 <libby> feel free to add notes from breakout sessions to http://rdfweb.org/topic/FoafGalwayBreakoutSessions

09:49:22 * danja arrives just in time for coffee

09:50:14 <jeen> sounds like you're having a good time up there in ireland

09:50:42 <libby> yes indeed

09:50:51 <libby> last night was fun

09:51:27 <jeen> not another foam party I hope?

09:52:38 <libby> no, a medieval banquet!

09:52:45 <libby> with poems!

09:53:21 <libby> "ah the songs of the old country! they'll melt your face"

09:53:48 <danja> lol

09:54:01 <jeen> wow, nice :)

09:54:26 <libby> I'm sure they'll be lots of photos. there's a few on my site

09:54:47 <libby>http://swordfish.rdfweb.org/photos/2004/09/01/

09:57:00 <danja> heh, I like danbri's handy-size business card

09:57:37 <danja> good turn out - was it 100 then?

10:06:46 <bengee> heh, nice photos

10:07:50 * bengee hopes they didn't 4get to print a biz card for libby, too..

10:09:07 <jeen> 'excuse me, good lady, who lives in that castle?'

10:09:30 <iand> more photos collected here: http://del.icio.us/tag/foaf-galway

10:21:37 <libby> 102 or 105, something like that

10:22:42 <libby>http://www.w3.org/2001/sw/Europe/events/foaf-galway/papers/fp/keyword_extraction_from_the_web/

10:22:53 <libby> --Keyword Extraction from the Web for FOAF Metadata

10:22:58 <libby> golbeck! how are you?

10:23:15 <libby> Junichiro Mori speaking

10:23:27 <golbeck> hi libby

10:23:31 <golbeck> everyone

10:23:38 <golbeck> i'm a bit better today, i think.

10:23:40 <golbeck> thanks :)

10:23:44 <golbeck> how was the castle dinner last night?

10:23:48 * mortenf welcomes golbeck back to the living :)

10:23:54 <libby> fun, yeah

10:23:59 <libby> glad you're feeling better

10:24:02 <golbeck> thanks, mortenf

10:24:08 * libby hoping to finally meet golbeck f2f

10:24:14 <golbeck> me too :)

10:24:16 <sh1m> yay! golbeck :)

10:24:19 <sh1m> golbeck++

10:24:34 <golbeck> i still have foaf apparel for a couple people who must claim it

10:24:38 <sh1m> are we (someone) minuting presentation?

10:24:43 <libby> yeah

10:24:45 <sh1m> ok

10:25:03 <sh1m> demo of extraction of social network from the web

10:25:07 <libby> ...shows a cool demo of various links between people and things

10:25:40 <sh1m> ...background of SW, social Networking, and FOAF

10:25:48 <libby> ...creation of metadata is a problem

10:25:53 <libby> ...need a way to create it

10:26:04 <libby> ...facilitate and accelerate

10:26:22 <libby> ...much foaf data is made by hand or using foaf-a-matci

10:26:41 <libby> ...but there's a lot of personal information on the web alreday

10:26:57 <sh1m> this presentation: Keyword Extraction from the Web for FOAF Metadata - Junichiro Mori, Yutaka Matsuo, Mitsuru Ishizuka, Boi Faltings

10:26:57 <libby> ...proopose keyword extraction method to create foaf data form existing data

10:27:15 <sh1m> presented by Junichiro Mori

10:28:01 <libby> ...keywords looking for are important entities/attributes to describe people, e.g. foaf, card

10:29:38 <libby> [missed how they did it :/]

10:29:49 <libby> ...search for a name on google e/

10:29:57 <libby> e.g. "dan Brickley"

10:30:08 <libby> ......use top 10 documents

10:31:08 <libby> ...html tag deletion, parts opf speech processing, ....

10:31:44 <libby> ...problems: is there an optimal number of documments to use to extract keywords?

10:32:02 <libby> ...depends on the algorithm used by the search engine

10:32:12 <libby> ..."Tom Baker" problem ;)

10:32:18 <libby> ...people with the same names

10:32:47 <libby> ...partial solution is adding affiliation to the query: Tom Baker GMD

10:32:55 <danbri> .g doctor who tom baker

10:32:56 <sh1m> oooh oooh see the inherent "unambigious property constelation" in the system

10:33:21 <libby> ...score each term using co-occurrence

10:33:26 <phenny> doctor who tom baker: http://www.imdb.com/name/nm0048982/

10:33:46 <nmg> sh1m: unambiguous property constellation?

10:33:47 <libby> .g tom baker gmd

10:33:50 <phenny> tom baker gmd: http://www.desire.org/html/research/deliverables/D3.5/d35peer2.html

10:34:11 <libby> ...term appearance in the same web page

10:34:24 <libby> ...calculated using jaccard coefficient

10:35:35 <libby> ..e.g. Dan Brickley and SWAD ~ 2000 webpages; one or other much more

10:35:55 <libby> ..the calculation of the intersection is expressive

10:36:02 <danbri> danbri is now known as swad

10:36:29 <libby> ...problem using co-occurence information: different contexts for the same person

10:36:44 <libby> ...context is a viewpoint, background, situtaion, status

10:36:54 <swad> swad is now known as danbri

10:38:01 <libby> ...e.g. Dan Brickley; Foaf; swad separtely; dan brickley and swad; dan brickley and foaf file

10:38:16 <libby> ...people can be contexts of another person

10:38:53 <libby> see http://www.w3.org/2001/sw/Europe/events/foaf-galway/papers/fp/keyword_extraction_from_the_web/ section 3.1

10:40:35 <danbri> [one thing of interest; you can use those links to find people whose foaf files and other rdf docs likely provide more info about me, so a good way to leverage google...]

10:41:19 <libby> [how are these terms classified I wonder? technical term, name, org name...]

10:41:35 <mortenf> [yeah, if you're the only dan brickley...]

10:41:55 <libby> :) thanks tony

10:42:08 <libby> - manual extraction of the types of keyword

10:42:51 <libby> ...then categorise the webpages found as different things: homepage, event page, weblog etc

10:43:00 <libby> ...weblog looks useful as a source of information

10:43:23 <libby> ...stats of keyword type per type of page

10:43:36 <libby> ...weblog also useful

10:44:22 <libby> ...weblog can have personal information which isn not included in people's personal web page

10:45:18 <libby> ...so how do we create foaf files to from this information?

10:45:28 <pmika> weblog can be useful only if you analyse people who have blogs...

10:45:53 <libby> ..useful for finding interests of people

10:46:22 <libby> ..problems: is the keyword really right? could get someone to check - intereactive creation of foaf files

10:47:26 <libby> ...hoe is the property of the keyword decided? NLP? machine learning?, entity extraction?

10:47:42 <libby> ...how do we express the context in a foaf vocabulary?

10:48:13 <libby> ....they extracted a social netowrk form the web; they combined it with the keywords

10:48:32 <libby> ...keywords -> semantics to the social network: services like finding a community

10:49:09 <libby> ...shows a diagram using Dan brcikley and Libby Miller - the keywords found include Jan Grant, ILRT, RDF, Europe project

10:49:17 <libby> ...can get some information abotu what this network means

10:49:25 <libby> ...have to be careful about privacy

10:49:40 <danbri> [except my privacy; I'm obviously asking for it! :-]

10:50:18 <libby> ...future work incliudes: evaluation of extracted keywords, precision and recall; ...integration with e.g. foaf-a-matic and other foaf creation systems

10:50:54 <libby> [I'm very interested in tools to help creation; could be analysing and feeding back existing foaf information as well as data mining form the web]

10:51:27 <libby> ...their system only works for people very active on the web, e.g. researchers

10:51:50 <libby> --questions

10:52:01 <danbri> [can someone closer type the question?]

10:52:17 <libby> q:applying this to a different docain, real life example: how would you do this?

10:52:29 <pmika> is this specific to this community?

10:52:33 <libby> ...also how long does the processing of the data take?

10:53:09 <libby> a:researchers - information is available. time taken, esp keyword extraction - 3 days

10:54:15 <libby> q: maybe use the same method to try and work out is SWAD is e.g. a project or an organisation?

10:55:04 <ldodds> I would be interested to know what the general feeling is about scraping data about users. If one takes just whats publically available, but republish it as FOAF, are there privacy (or other?) issues?

10:55:18 <libby> ----next speaker

10:55:19 <libby>http://www.w3.org/2001/sw/Europe/events/foaf-galway/papers/fp/trust_aware_decentralized_recommender_system/

10:55:19 <libby> Moleskiing: a Trust-aware Decentralized Recommender System

10:55:29 <libby> speaker: Paolo Massa ,

10:55:49 <libby> ...aim is to recommend skiroutes to other users

10:56:47 <libby> ...goal of the site is to make ski mountaineering safer

10:57:24 <libby> ...people chooses route using weather conditions, users the route and then reports back: decentralized assessments of snow conditions

10:57:34 <libby> ...want to automate this process

10:58:04 <libby> www.moleskiing.it - site in italian about routes: describe your trip

10:58:40 <dajobe> bblish suggests using babelfish (which he created)

10:58:48 <libby> ...goasl is to recommend to every single user routes which are secure and enjoyable for them

10:59:10 <libby> ..collaborative filtering has various problems: sparse data nd malicious attacks

10:59:52 <libby> ...trust-aware recommender system is the solution

11:00:19 <libby> ...ask users to express their trust about other users; all trust statemenets are public

11:00:36 <libby> ...bit like blogrolls, ebay, epinions, p2p, even google

11:01:08 <libby> ...trust statements (links) are weighted e.g. X trusts Y .7/1

11:01:58 <libby> ...agreate them together and then use trust metrics: use existing edges to predict non-existing edges

11:02:18 <libby> s/agreate/aggregate/

11:02:42 <libby> ...moleskiing predicts for ever user, trust scores of other users, based on the social network

11:03:34 <libby> ...nothing to do with liking people

11:03:35 <bblfish>http://babelfish.altavista.com/babelfish/trurl_pagecontent?url=http%3A%2F%2Fwww.moleskiing.it%2Fmski%2Fhome.do&lp=it_en

11:03:43 <libby> ...trust metrics wiki [missed url]

11:04:05 <libby> ...data is exported in foaf using foaf extension from Jen Golbeck

11:04:15 <golbeck>http://trust.mindswap.org

11:04:15 <libby> trust.mindswap.org

11:05:20 <libby> person1 - trust:trust9 -> person2 - dc:subject -> some particular route (by url)

11:05:48 <libby> ...also can exopress what they think about the route: secure, interesting, etc

11:06:19 <libby> ...hompage sows ski routes: commnets in last 15 days, rated secure by majority of users, ranked in order of trust

11:06:37 <libby> ...didn;t want to create yet another site to login to

11:07:25 <libby> ...you put some stuff in your *own* foaf file, with soem trust and skiing information; ping moleskiing.it and we aggregate it

11:07:54 <libby> ...using bookmarklets http://hublug.hubmed.org/archives/000175.html [?]

11:08:25 <sh1m> nmg, finally found the reference http://rdfig.xmlhack.com/2004/03/22/2004-03-22.html#1079970341.819230 that's the original conversation.

11:08:41 <libby> ...login with you foaf - how? - various possibilities, maybe semanticv cookie, create login....not sure yet

11:08:43 <sh1m> nmg, "unambigious property constellations" is Sandro's term.

11:08:52 <nmg> sh1m: thanks

11:09:08 <sh1m> nmg, basicly collection of properties which make an IFP (I call them composite IFPs or cIFPs

11:09:25 <libby> ...phiolosphy: don;t wnat data to be centralized and locked in; interop with other small skiing communities - share users

11:09:33 <nmg> I thought that was probably what you meant

11:09:34 <libby> ...they will send data but don;t want to make it public

11:10:53 <sh1m> nmg, probably should have given you the one sentance explanation earlier but i click through about 6 mo of the scratch pad looking for hte link

11:11:02 <libby> ...real community site; will report back on login with foaf, and sharing data betweehn communities; maybe make distrust dataments private so people can be more honest

11:11:13 <sh1m> nmg, looking through all the '03 pages by mistake didn't speed it up

11:11:17 <libby> ...personally he would like to be able to login using his foaffile

11:11:26 <nmg> sh1m: there was some discussion on the webont list about this last year

11:11:56 <sh1m> nmg, right. although I am approaching it from a slightly different angle

11:12:06 <libby> q: is it computationally expensive?

11:12:14 <sh1m> nmg, from an ontological point of view its trivial

11:12:14 <libby> a: no using trust metrics ...

11:12:37 <libby> q: have you run into any inconsistencies? can get these problems

11:12:47 <sh1m> nmg, but in terms of taking a dataset and identifying and optimising for the use of cIFPs has issues.

11:13:38 <libby> a: goal of trust metrics is to be able to incorprate these differences of opnion: the differences in trust may be because A likes it and B hates it

11:13:42 <libby> ---next speaker

11:13:54 <libby>http://www.w3.org/2001/sw/Europe/events/foaf-galway/papers/fp/challenges_of_foaf_characterization/

11:14:02 <libby> The Challenges of FOAF Characterization.

11:14:02 <libby> John C. Paolillo and Elijah Wright

11:14:09 <libby> John speaking

11:14:46 <golbeck> some trust metrics of interest: Eigentrust (a page rank modification), Richardson et al's metric (ISWC 03), Zeigler, <shamelessPlug>mine at trust.mindswap.org</shamelessPlug>, and paolo maintains a list at http://moloko.itc.it/trustmetricswiki/moin.cgi/FrontPage

11:15:16 <libby> ...why characterise foaf?

11:15:27 <libby> ...does the design of foaf fit its intended purpose?

11:15:38 <libby> ...how are the resources of foaf actually used?

11:15:50 <libby> ...impacts of technological barriers, social affiliations etc

11:16:00 <libby> ...inetrested in learning about the spread of foaf itself

11:16:15 <libby> ...challenges for foaf characterisation

11:16:24 <libby> ...more so than web characterisation

11:16:46 <libby> ...problems of gettign a represeantative sample; distorting effects of large sites

11:16:52 <libby> ...data quantity and quality

11:16:56 <libby> ...ethical considerations

11:17:14 <libby> ...informed consent, security, authentication need more thought

11:18:23 <libby> ...sampling issues; random isn;t possible; snowballing is (suttering); the former is representative but noot possible because we don;t know where all the foaf is; the latter is not representative, and has a bias towards well-connected files

11:18:49 <libby> ...will never find unconnected bits

11:19:23 * danbri hopes people try to minimise data islands (another motive for foaf:knows being broad definition, so people aren't afraid to make RDF islands into RDF hypertext)

11:19:24 <libby> ...try to get around this with a large set of points; would like everyone to try characterising foaf

11:19:29 <libby> - more data

11:19:38 <libby> [yeah, like wikis]

11:19:53 <libby> ...used jim ley's crawl from March 2004

11:20:07 <libby> ...1G mysql dump+indices

11:20:41 <libby> ...lots of data from live journal; makes it difficult to characterise foaf

11:21:05 <libby> ...by 4th day of the scutter run, no longer getting foaf from anywhere except LJ

11:21:07 <crschmidt> (lots of foaf data *is* from livejournal. that is a characteristic ;))

11:21:27 <libby> (I think they're looking for a different sort of characterisation)

11:21:37 <aharth> why to replicate all that livejournal data?

11:21:38 <crschmidt> (i know, i know, i'm just playin.)

11:21:40 <libby> :)

11:21:49 <libby> ...scutter gets stuck; and all machine generated

11:22:29 <crschmidt> LJ FOAF is good for generating graphs, but not a lot else

11:23:25 <nmg> crschmidt: why so?

11:23:36 <danbri> hi chris :)

11:23:38 <libby> ...shows a graph shoing last modifcation dates; they figure foaf hasn't been around long enough to make lots of modifications

11:23:38 <crschmidt> nmg: no data. no names, no images, nothing other than contact information

11:23:48 <libby> ...LJ data is excluded

11:23:52 <crschmidt> all you get is :knows and :interests

11:24:12 <libby> ...shows 2 peaks; adrian's photos

11:24:14 <nmg> ...which don't link up with anyone else's foaf data

11:24:22 <libby> (ajft.org)

11:24:36 <libby> ...so this didn;t really work as a way of showing how foaf was spreading

11:25:11 <libby> ...another problem was the size of the available adta: modest crawl; 6.73million triples, 196 namesapces

11:25:21 <libby> [wow! 196! cool!]

11:25:53 <libby> ...just from foaf ns, 3.81m; 839,934 are instances of foaf:knows relation

11:26:24 <libby> ...interest is very high too; foaf files might be made more efficient by looking at the way interest is encoded

11:27:12 <libby> ...274,00 distinct individuals (65 billion possible pairs); not feasible or tractible; been trying to reduce the graph into meanful cluisters

11:27:27 * danja blinks at 196

11:27:28 <libby> ..17,000 distinct subjects

11:27:36 <libby> 5.3 million length 2 paths

11:27:43 * ldodds wonders if there's a list of the namespaces

11:27:44 <libby> ...too big!

11:28:01 <libby> [could grab it off jim's dump leight?]

11:28:05 <libby> leigh, even

11:28:14 <ldodds> yes, I was being lazy :)

11:28:18 <libby> ...average linklage between people is 3 or 4

11:28:19 <danja> leight of and

11:28:28 <libby> heh

11:29:13 <libby> [if you do a logs search of foaf you'll find it, from when these guys were asking about it ;)]

11:29:17 <danbri> re reciprocality, Chris mentioned that in LJ, there is a strong culture of reciprocality.

11:29:37 <libby> ...48 individuals linking between the LJ and the rest of teh foaf community

11:29:42 <crschmidt> (strong enough to the point that i got a cranky email for un-:knows-ing someone in my file yesterday *sigh*)

11:29:50 <libby> hehhe

11:29:54 <ldodds> LJ and places should probably dump a complete export. would make it easier for scutters I suppose.

11:29:59 <libby> can people be un-known?

11:30:01 <crschmidt> ldodds: LJ can't, really

11:30:12 <libby> ...some ethical observations

11:30:17 <crschmidt> libby: this was my LJ file, and :knows is really just the equivilant of reads_weblog

11:30:45 <libby> ...youthful users: exceptionally sociable, consumers, legally vulnerable (RIAA, porn etc)

11:30:51 <Cloud> same with plink, you may want to un-know someone who 'knows' you

11:30:59 <Cloud> and appears on your plink profile

11:31:04 <libby> ...typical YASN is only viable bexcause of these youngsters

11:31:18 <libby> [mikssed a bit]

11:31:29 <libby> ...informed consent is very important

11:31:30 <crschmidt> ldodds: LJ can't really dump the data in any way. ignoring the fact that 4 million files is just a lot of data, getting it all out in any way isn't a simple task

11:31:30 <dan-uk_> crschmidt: cf. the LJ culture of "friends-only"

11:31:33 <Fides> Oddly enough almost all the people I know on LJ are 25+

11:31:43 <crschmidt> (not that it can't be done, just that it most likely won't be :))

11:32:14 <libby> q: will you keep the data fresh and up to date?

11:32:15 <crschmidt> Fides: you don't know me? *sniff* In reality, that's a huge exception: the average age is 17-18, http://www.livejournal.com/stats/

11:32:21 <ldodds> I guessed as much, but it'd probably be less load on the site to have a single export (which could be mirrored)

11:32:23 <libby> a: not really still grappling with the issues

11:32:43 <libby> q: whats the technical bottleneck for you guys?

11:33:00 <libby> a: need to d princviple components anaalyssis on huge dataset; scalability in general

11:33:03 <crschmidt> ldodds: yeah, but since LJ deals with 700 dynamic pageloads/second, i doubt foaf is even a percentage which matters to them at this point

11:33:42 <crschmidt> (please note that I am completley and totally unrelated to LJ in nay official way, just some guy who knows out it works)

11:33:42 <libby> q: any advice for scutter writers?

11:34:19 <crschmidt> "Polite scutters cache, impolite scutters get banned." -LJ

11:34:21 <libby> a: run lots of scutters and popool the data probably the best thing

11:34:38 <danbri> +1 re more scutters and expose data

11:34:50 <libby> q: voice suport by mail to tom croucher for a big dataset of foaf at sunderland university

11:34:56 <libby> ---next spaeker

11:34:57 <Fides> Probably. I think the is a small subculture that came to LJ from other areas (mailing lists etc) who are on the whole female, 18-60

11:34:57 <ldodds> seems sensible. can also have scutters targetted to particular sources/communities

11:35:04 <ldodds> mattb had some thoughts on that I think

11:35:08 <sh1m> The People's Portal: Ontology Management on Community Portals - Anna V. Zhdanova

11:35:15 <Fides> Since that is the group of people I hang out with... ;-)

11:35:15 <libby>http://www.w3.org/2001/sw/Europe/events/foaf-galway/papers/fp/peoples_portal/

11:35:22 <libby> link is boked in programme, sorry

11:35:30 <sh1m> present the people's portal, a community portal with ontology managemenr

11:35:34 <libby> no probs, thanks chris!

11:36:05 <sh1m> vision for better interoperability using SW

11:36:32 <sh1m> ...involves new standards and software

11:36:47 <sh1m> ... what is the state of the art? we have many tools to make metadata

11:36:48 * bengee thinks (re scuttering/entry points) Mark Giereth's approach to use google to get an initial set of (not connected) foaf files makes sense, too

11:37:01 <sh1m> ...little motivation, need to make people use the tools to make the SW reality

11:37:29 <sh1m> ... if people didnt exchange non semantic information but use semantic data instead

11:37:36 <sh1m> ... looking at making this vision a reality

11:37:43 <sh1m> ... look at orkut screen shit

11:37:50 <sh1m> s/shit/shot/

11:37:53 <sh1m> (opps)

11:37:56 <golbeck> ha

11:38:15 <sh1m> ... cannot control data, and so it devolves to email

11:38:17 <mortenf> hehe

11:38:26 <sh1m> ... for every issue you have to write emails

11:38:44 <sh1m> ... allow users to build ontologies on a community portal

11:39:05 <sh1m> ... stephan goes to a portal and makes a new ontology and starts using it, then i can use it too

11:39:22 <sh1m> ... what is needed to get to the vision of interoperability without the overhead?

11:39:48 <sh1m> ... need to marry community web portals with collaberative ontology construction tools

11:40:19 <sh1m> ... env. comrpises of ontology management env. and a community semantic web portal

11:40:29 <sh1m> ... lots of research topics around this

11:41:11 <sh1m> ... what about implementation efforts?

11:41:32 <sh1m> ... our goal is to develop a semantic community portal that allows users to control the information structure of the portal

11:42:14 <sh1m> ... possible to distringuise different levels of ontologies

11:42:39 <sh1m> ... user data like telephone numbers, also describe communities, personalisation, and the portal themselves

11:42:51 <sh1m> ... all different levels of ontologies which power the portal

11:43:12 <sh1m> ... users can log in and view their data, modify it, and extend an ontology with another attribute or class

11:43:34 <sh1m> ... user can see the status of the ontology items, so if it is a stable or not ontology

11:43:45 <sh1m> ... speicificly adapted to the deri website

11:44:12 <sh1m> ... basic ontologies to describe people, projects, using extensions of foaf, and we use the status for concepts

11:44:48 <sh1m> ... if it used for publishing then it is "stable" if it is user proposed it is "testing" if a user proposal is popular it might become "stable"

11:45:04 <sh1m> ... currently 91 people on the DERI sysmte

11:45:21 <sh1m> ... generates standard pages at hte front end

11:45:33 <sh1m> ... allows user to display custom data without programming skills

11:45:49 <sh1m> ... converters between uis and ontolgies and instances

11:47:39 <sh1m> ... uses jena, sesame, but being moved to rdf/xml + php

11:47:39 <sh1m> ... using RDFS and OWL

11:47:39 <sh1m> ... future work, have done basic functionality, need to look closer at publishing cycles and user uses

11:47:57 <sh1m> ... conclusion: speificiation for a framework for collaborative ontology development in a community site

11:48:06 <sh1m> ... make the DERI site totally semantically enabled.

11:48:24 <sh1m> ... interesting to launch another use case, 'semantic classmates' or other social networks

11:48:50 <sh1m> ... look at the changes of how a social networking site would be with instant publishing

11:49:25 <sh1m> ... requirements for the DERI use case, version alignment, different ontologies

11:49:34 <sh1m> ... need to describe communities not just users

11:49:55 <sh1m> ... we haven't found as many schemas we could reuse. wish to encourage development of ontologies.

11:50:47 <sh1m> morten: how do you handle input and output of new terms that are added to the ontology

11:50:55 <sh1m> morten: how do you present to the user

11:51:11 <sh1m> anna: sometime we use navigation on different pages, now mostly on one page

11:52:02 <sh1m> q: dropping sesame and move do you know what the performance is like on the new system?

11:52:17 <sh1m> anna: we are looking for better data integrity issues

11:52:31 <sh1m> big performance differnce on PC vs Mac

11:52:43 <sh1m> new system = jena

11:53:01 <sh1m> q: is the semantic web site still live?

11:53:21 <sh1m> its not but it should be

11:53:29 <sh1m> its might be a bit out of date

11:53:46 <sh1m> damian: do you have a base ontology that you can derive from?

11:54:16 <sh1m> damian: if i wanted to add a new kind of localised name, i would need three properties could I do that on the interface?

11:54:37 <sh1m> damian: or geograhpic information, a bnode with lat and long attached

11:54:50 <sh1m> anna: quite simple ontologies but we are looking to improve them

11:56:09 <sh1m> tom: how do you deal with users adding lots of ontologies which are essentially the same?

11:56:20 <mortenf> lunch!

11:56:29 <sh1m> anna: all the users on the same site use the same ontology

12:59:53 <iand> see http://rdfweb.org/topic/FOAFCommunityProcess

12:59:56 <libby> danbri's session: looking at http://rdfweb.org/topic/FOAFCommunityProcess

13:00:00 <libby> :)

13:01:39 <danbri>http://rdfweb.org/topic/FOAFCommunityProcess

13:01:40 * bengee finds http://rdfweb.org/topic/FOAFCommunityProcess really great. not easy to implement, though..

13:01:56 * ldodds just reading

13:03:03 <libby> ....when foaf was young, wvenryone knew each other so codification of the process wasn;t so important

13:03:23 <libby> ...gradually it became clear that we needed to articulate some of these assumptions

13:03:30 <libby> ...hence this document

13:03:41 * dan-uk_ thinks there would be lots of ontology mapping if you let people create their own predicates, to cope with duplicates.

13:04:19 <libby> ...foaf has learned from DC, which wasn;t particlarly smooth, as they were creating a process as they went along

13:04:59 <libby> ...DC had 1.0 and 1.1 versioning in their namespace; the rdf spec used to say this, and it turend out to be wrong...

13:05:12 <libby> [this was a nightmare]

13:05:29 <libby> ...lots of docs everywhere with the old namespace in so lots of messy code

13:06:00 <libby> ...foasf ns is managed a sa living document. it will change. 0.1 in the ns is not going to change, for the DC-type reasons

13:06:34 <libby> ...the document in the wiki above is about building a sane 'process' thingy around a living document

13:06:53 <libby> ...Tom Baker (not Dr Who) is spent a lot of time doing this for DC

13:07:12 <libby> ...can;t quite borrow the DC process - because foaf structures are more complex: bt we can learn from it

13:07:44 <libby> ...stable, unstable and testing propertioes and classes in foaf. trtying to work out what this means, learning from business and academia

13:08:37 <libby> ...some of the idems in the new doc havenl;t been lives up to, e.g. nice clear list of changes, e.g. an rss feed. the mailing list is too busy to track

13:09:37 <libby> ...dan changed the domain of homepage yesterday with a cvs commit; changes thousands of foaf documents. he's not going to do it again :) need to talk to people first, and also make sure the documentation tracks the changes accurately

13:10:18 <libby> ..there are different types of changes to make; some are 'worse' thnan others in that they have more effects and may lead to falseness

13:10:57 <libby> petermika: does good software depend on stability?

13:11:21 <libby> a: shouldn;t assume the doc will not change but shold invest more in stable

13:11:26 <libby> sable: shouldn;t change much

13:11:41 <libby> testing: don;t make changes lightly

13:11:45 <libby> instable: may change

13:11:53 <libby> s/sable/stable/

13:12:37 <libby> ...translating the spec to different languages is very important, including translating anglo-centric properties e.g. schoolhomepage

13:13:01 <libby> q: what's the process for deciding about stability? people will want things to move to stable...

13:13:38 <libby> a: we can get a sense of things leading to stability: being used: being generated and consumed; need not be application-specific

13:14:12 <libby> q:different namespace? move stuff over to a good one?

13:14:26 <libby> a: DC problem....

13:14:38 <libby> (Ian was the questionner there, and doesn;t seenm convinced)

13:15:02 <libby> q: moving terms form one ns to another isn;t as bad as changing the ns.

13:15:18 <libby> q: experimentation *outside* the foaf ns

13:15:30 <libby> a: sorta tried this, using stats to see deployment

13:16:03 <libby> ...really don;t think that people change their code enough to make changes to ns a good idea

13:16:23 <libby> petermika: deprecated as well as stable etc? sounds useful

13:16:33 <libby> q: yeah some stuff ion owl about this, makes sense

13:16:40 <libby> a, rather

13:17:02 <libby> ...we need a better decision-trail, record trail of what's happened. doesn;t mean we need a w3c process

13:17:35 <libby> ..even if not a democracy, people need to know why properties are the way they are, have their objections acknowldged

13:17:53 <libby> ...current suggestion is to use the IssueTracker on the rdfweb.org wiki

13:18:22 <libby> ...we don;t really have time to spend time doing this ourselves, onus is on others to do it

13:18:34 <libby> q:use a mailing list of more fopmal issue tracker?

13:18:48 <libby> a: xquery got 2000 commnets, can;t deal with that

13:19:33 <libby> ....important to have regular public meetings for updates, making sure they are moving, not promising done by a certain date. make sure that people coming to foaf can find stuff, see that it's ongoing

13:19:56 <libby> ...basic stuff, datestamped logs about what's happened

13:20:13 <libby> ...also it is in cvs, thoughh raraely use that

13:20:39 <libby> ...would prefer w3c-type thing with datestamped older versions of the html and rdf specs

13:21:06 <libby> ...howevere if we do this we'll end up with old, datestamped namepaces; should we strongly dicourage people from using them?

13:21:17 * ldodds preferred it when the issues were in bugzilla

13:21:19 <libby> q: can we make sure the old ones can;t be used?

13:22:15 <libby> ian's suggestion: make sure that the old versions of the spec have relative uris

13:22:33 <iand> errm. absolute URIs

13:22:37 <libby> ...like's ian's suggestion

13:22:54 <libby> ...foafcamp is different things to dan at different times

13:23:20 <iand> (in rdf schema, make sure terms are defined using rdf:about rather than rdf:ID)

13:23:26 <libby> ...need to improve the information about what foaf is, why needs mixed namespaces, why mix of people, business tsuff etc

13:23:43 <libby> ...mailing lists: rdfweb.org, foaf-corp

13:23:52 <libby> sorry Ian, can;t think and type!

13:24:02 <libby> ...new list just for spec issues?

13:24:15 <crschmidt> (like the OWL thing recently?)

13:24:23 <libby> [maaaybe from the audience]

13:24:33 <libby> didn;t see taht Chris

13:24:57 <libby> ...foaf-talk, foaf-spec?

13:25:21 <libby> ...any objections?

13:25:54 <libby> morten: often people feel that they are missing properties, and think it should go in foaf

13:26:04 <libby> a:need better docs for this

13:26:36 <libby> [discussions: no partiocular objections, but not sure if it will work]

13:26:51 <libby> libby: rss feed for changes to spec would be v useful

13:27:11 <libby> danbri: SKOS usess testing/stable/unstable; he has an rss feed for the spec changes

13:27:18 <libby> ...thiunk we'll do taht

13:27:47 <libby> ...all this stuff will maybe get written up be semnatic web best pracices WG; ontology managment is not simple

13:28:57 <libby> ...people ask: can we have postal addresse in RDF?

13:29:16 <libby> ...danbri puts w3c hat on, suitable W3C WG fodder

13:29:43 <crschmidt> ooh, a w3c hat? I want one!

13:30:02 <libby> heh

13:30:47 <libby> ...things do trickle through, though peopele may not realize

13:31:18 <libby> ...dan connolly can;t be with us toay, - he suggested testcases; thse have been great in rdf core wg and owl, much better than before

13:31:25 <libby> hey jibberjim!

13:31:55 <libby> ...we're not sure how this related to things like foaf: not formal but 'is this the better way'

13:32:03 <libby> ...todo: have some examples

13:32:15 <JibberJim> hello galwayers, how's it going?

13:32:25 <mortenf> libby: ical/rdf experiences

13:32:34 <mortenf> ...wanted conversion

13:32:46 <mortenf> tried generating schema from converted data

13:32:54 <mortenf> ...now manually edited

13:33:19 <mortenf> ...lightweight process: change schema, let people scream within a week

13:33:39 <mortenf> ...except didn't work re timezone changes, people screamed, didn't back out

13:33:53 <mortenf> ...works well for minor changes

13:33:56 <dan-uk_> JibberJim:

13:34:22 <mortenf> ...difficult for larger changes

13:34:38 <mortenf> ...running code helped, also the examples

13:35:03 <mortenf> danbri: perhaps same for e.g. vcard/foaf conversions

13:35:13 <libby> jibberjim, good yeah, shame you're not here

13:35:24 <libby> !pic morten, in galway

13:35:26 * mortenf waves to jim

13:35:27 <JibberJim> yeah, I drunk lots last night in your honour

13:35:32 <mortenf> :)

13:35:43 <libby> aw, thanks

13:36:02 <libby> danbri: worries about namespaces disappearing, the socila aspects of managing namespace and ontologuies

13:36:03 <wh4experimental> Galway, Ireland http://swordfish.rdfweb.org/photos/2004/08/31/2004-08-31-Images/19.jpg Morten Frederiksen

13:36:04 <wh4experimental> Galway, Ireland http://swordfish.rdfweb.org/photos/2004/08/31/2004-08-31-Images/24.jpg Morten Frederiksen

13:36:10 <mortenf> heh

13:36:25 <libby> ..putting it in your will etc; paying for it for lots of years

13:36:51 <libby> todo list:

13:37:06 <libby> - regular meetings in irc: but what time?

13:37:34 <libby> - european afternoons?

13:37:50 <libby> - wednesdays

13:37:54 * crschmidt thinks that noon-ish eastern time, late afternoon BST might be best

13:38:27 <libby> ian worries what if it isn't peoples' day job?

13:39:01 <libby> dirk argued against them....

13:39:23 * libby thinks they are useful deadline

13:39:25 * crschmidt points out that it isn't his day job, but he can usually at least keep an eye on IRC during the day

13:39:42 <crschmidt> which I can't do most times at home, since children tend to demand more attention than my boss ;)

13:41:06 <crschmidt> with a worldwide population, you can't find a perfect time. with a distributed development base, you need interactive communication, and mettings lead to showing off, wanting to get things done in time, etc.

13:41:32 <crschmidt> (in my opinion.)

13:41:58 <libby> danbri suggests wednesday 15th sept 17:00 bst

13:42:14 <libby> .time

13:42:29 <libby> what's that in utc?

13:42:40 <JimH> 16:00

13:42:44 <libby> ta

13:43:13 <libby> --new topic: extension and cmpanion namespaces - linking to them, anything else?

13:43:17 <libby> companion

13:43:19 <crschmidt> 11:00am eastern, 8am pacific

13:43:28 <libby> jan would like more 'standardization; around this

13:43:36 <libby> thanks chris. is that any good fo you?

13:43:42 <libby> s/jan/jen/

13:44:09 <crschmidt> yeah, i like that time, that'd be my suggestion, seem to get a good amount of people online around then

13:44:37 <libby> danbri: wiki page with companion nsmaepace in?

13:44:43 <libby> jan: yep, trust one

13:44:47 <libby> jen even!

13:44:58 <libby> ian: bio ns, vocab.org

13:45:17 <crschmidt> I think that wikis are fine for deciding what should be on a page, but not good for "recommendation" if you want uptake, maybe. Better to say "here's our list" so people know it's a bit more stable, wikis have the feeling of "it's never done"

13:45:52 <libby> ian would like soem apps to help with the ontology management stuff: has ideas but no time

13:46:03 <crschmidt> (And I'm sure others will disagree, lots of projects only use wikis for docs, etc. but PHP for example has very good non-wiki documentation, same with mysql)

13:46:19 <libby> morten: benjamin nowack has some thing: owl owlchestra thingy

13:46:38 <libby> yeah re wikis chris. is tricky

13:46:57 <libby> danbri: foaf is just the first one facing these problemns

13:47:10 <libby> ian: translation - often just can ask for help

13:47:17 <libby> danbri: yeah e.g. foaf-a-matic

13:47:45 <libby> ian: deifinitely get stable stuff translated

13:48:14 <libby> danbri: for a long time masahide's documentation waas better than the foaf spec :)

13:48:14 <crschmidt> (foaf explorer also, i think?)

13:48:40 <libby> danbri: thinks that is one way of seeing if it's stable: have translations

13:48:45 <libby> ian: and localization issues

13:49:01 <gromgull> Look at the sad king: http://kasei.us/pictures/events/20040901-FOAF_Galway/15

13:49:17 <libby> danbri: namespace itself should ahve one line bio!

13:49:46 <libby> hehehe gromgull

13:50:12 <libby> oh dear

13:50:30 <libby> libby: I think people get frustrated because we don;t have a list of things to ick off when it's unstable

13:50:52 <libby> danbri: we have this list, but is intuitive.... :)

13:51:10 <libby> ...doesn;t think it's a solvable problem

13:51:30 <libby> ...but make sure we have a documented visible process/.paper trail

13:51:52 <xavier> hey, random question: foaf:dateOfBirth - what happned to it?

13:52:23 <libby> ian: very useful to people to show how can use foaf terms e.g. in rss; e.g. use the same name in xml as we do in rdf, e.g. like XFN

13:52:34 <libby> that came up on the list recently xavier

13:53:12 * libby looks

13:53:15 <xavier> libby: i think i saw something, but many foaf files got in the way

13:53:47 <danbri> loggerf, pointer?

13:53:47 <danbri> See http://www.ilrt.bris.ac.uk/discovery/chatlogs/foaf/2004-09-02#T13-53-47

13:54:03 <crschmidt> dateOfBirth doesn't really exist, but that didn't stop me from using it! :) libby, you may be thinking of the discussion in #wh4 with balbinus re: it

13:54:03 <libby>http://rdfweb.org/pipermail/rdfweb-dev/2004-August/thread.html#13566

13:54:09 <crschmidt> oh, or not :)

13:54:12 <libby> heh

13:54:22 <libby> yeah that too though

13:54:24 <xavier> crschmidt: ah. yeah. i was figuring that's what happened.

13:54:47 <crschmidt> there's a lot of it out there though, cause i didn't understand it wasn't in the spec and ended up implementing it for lj

13:54:50 <libby> jon/ian: lots of education to go

13:54:50 <crschmidt> (oops)

13:55:05 <xavier> crschmidt: that's one way to add a term to a vocabulary: make a large population use it

13:55:05 <libby> heh, it was meant to be biut we never fgot the details ironed out

13:55:21 <libby> chris: might be a bit dodgy - are there lots of kids in LJ?

13:55:34 <libby> ian: more examples, more time

13:55:35 <crschmidt> libby, it's only displayed if listed publically in the html version

13:55:42 <crschmidt> (same with all foaf stuff)

13:55:42 <libby> ah, just checkin'

13:55:48 <libby> you still might run into issue though

13:55:53 <xavier> libby: and the details that are omitted in the HTML ver, are still omitted.

13:55:55 <crschmidt> and it's off by default, unlike most things

13:57:11 <crschmidt> Like all data in FOAF on LJ, it's more likely to be an issue with the HTML version than the RDF version

13:57:18 <xavier> i'm just wondering, as my little toy i've been working on will display rdfs:label for things it knows, and it didn't have a label for that one ;-)

13:57:29 <crschmidt> if data is public, it's public, damnit :P

13:57:35 <xavier> agreed!

13:58:00 * crschmidt fought so long over issues like that with people on LJ FOAF data. "this makes it easier for people to $foo" "The data is already in html"

13:58:41 <xavier> crschmidt: i found it sorta scary that even some techies thought that security through obscurity was any obscurity at all

13:59:00 <libby> --coffeee

13:59:04 * crschmidt s/any obscurity/any security/

13:59:16 * xavier chuckles and thanks crschmidt

13:59:17 <jsled> well, certainly it's both.

14:03:50 <gromgull> Are there pubplans later?

14:24:40 * mortenf notes ldodds and danja have been busy the last days

14:24:50 <mortenf> should've spent the time in galway instead ;)

14:24:56 <ldodds> heh

14:25:27 * ldodds has some other code to release soon too

14:25:38 <mortenf> oh, and thanks for the seemingly global "plug morten's stuff" day :)

14:26:06 * ldodds grins

14:26:55 <danbri> morten++

14:30:21 <davidoc> Jean-Marc Seigneur on Mobile FOAF:

14:30:47 <davidoc> ... Review of existing work where foaf is/could be used in mobile settings

14:30:53 <davidoc> ... Plink

14:31:03 <davidoc> ... BlueFOAF (proximity based)

14:31:27 <davidoc> ... different approaches in different countries

14:31:37 <davidoc> ... proximity based comms cheaper in europe

14:32:05 <davidoc> ... mobile internet might get cheaper in europe eventually

14:32:24 <davidoc> ... important thing is that user is in control of their FOAF info

14:32:35 <thnorkus> thnorkus is now known as paolillo

14:32:42 <davidoc> ... location info in FOAF would change as you change location

14:33:07 <davidoc> ... passive interaction: space customised for you based on FOAF preferences

14:33:09 * crschmidt points to menow:located

14:33:16 <paolillo> ldodds you there?

14:33:18 <davidoc> ... active interaction: such as handing over a business card

14:33:43 <davidoc> ... would need some software to update FOAF info

14:34:04 <paolillo> ldodds, I was looking at your proposal for foaf:owns

14:34:09 <davidoc> ... Add FOAF:NetworkProvider for roaming/cost purposes

14:34:21 * davidoc gives up for now

14:35:08 <paolillo> have you thought at all about type/token ambiguity in "owns"? something of a problem: I own a CD vs Warner Bros owns a CD (title)

14:35:28 <dan-uk_> dd

14:39:03 <ldodds> paolillo: yes that did occur to me

14:39:35 <ldodds> technically Warner Bros owns some rights over a CD

14:39:47 <ldodds> so was hoping that could be modelled as a different URI

14:39:49 * mortenf discovers foafspace.com <- anyone here responsible?

14:40:27 <mortenf> or perhaps a different kind of ownership?

14:40:40 <ldodds> yes, that would be better

14:40:46 <paolillo> ldodds, yes, you could argue that sense of own is different, although I think that the type/token ambiguity remains

14:40:58 <dan-uk_> surely it depends if you are talking about the cd or the music on the cd.

14:41:08 <dan-uk_> you own the cd. wb owns the music.

14:41:19 <sh1m> mortenf, http://www.stackframe.com/people/

14:41:25 <Fides> Newbie question: Does anyone know if there are any papers around comparing trust through digital passports and trust through social network webs? Thanks

14:41:28 <sh1m> mortenf, people who did it

14:42:04 <ldodds> paolillo, why don't you think a different kind of ownership would deal with it?

14:42:05 * sh1m is not sure he has seem those seemingly very recognisable individuals

14:43:05 <ldodds> isn't there always some ambiguity, i.e. the "thing" and the page describing the "thing"

14:43:28 <elijahwright> hi john (paolillo)

14:43:38 <paolillo> I own rover != forall(x) [dog(x) & name(x,rover) -> own(me,x)]

14:43:39 * mortenf thanks and looks

14:44:35 <crschmidt> off

14:44:38 <dan-uk_> that's annoying, freenode's clone detection kills the chat 1 hour before we finish.

14:44:44 <elijahwright> Fides, I hven't seen a paper like that, but would think that one would be pretty interesting.

14:45:47 <dajobe> I'm still there

14:46:11 <dajobe> stefan decker

14:46:24 <dajobe> how to continue the semantic web effort

14:46:57 <dajobe> asks for feedback

14:47:09 <dajobe> marcC

14:47:23 <dajobe> what is the summary of the wiki rewrite?

14:47:37 <dajobe> looking at FOAFCommunityProcess

14:48:00 <dajobe> danbri - previously most of the foaf people new each other, open process and we knew what each otehr thought, didn't need to write it all down

14:48:22 <dajobe> but now want to move to something slithgly more responsible

14:48:32 <dajobe> and slightly more heavyweight

14:48:39 <dajobe> but not a fulls tandards org at this point

14:48:54 <dajobe> doc discusses the foaf approach, why changes in place rather than frozen versions

14:49:07 <dajobe> MarcC has asked why not versioned

14:49:31 <dajobe> danbri - experience from dublin core has shown that DC V1.0 to V1.1 changes wasn't working so well, code still has it to fix that

14:49:39 <crschmidt> (k-line for conference has been removed, however, not everyone can rejoin at this point)

14:50:08 <dajobe> crschmidt: thanks

14:50:34 <dajobe> so the updating live is there for a reason

14:50:57 <dajobe> now the spec is in cvs, you can rollback, but that is too geeky

14:51:08 <dajobe> danbri - changed the spec last night "foaf galway edition"

14:51:38 <golbeck> i'm not the only one who just got k-lined and kicked off, tight?

14:51:43 <libby> nope

14:51:44 <libby> :)

14:51:44 <crschmidt> no, the whole conference did

14:51:47 <sh1m> ugh

14:51:48 <golbeck> s/tight/right/

14:51:49 <golbeck> nice

14:51:49 <sh1m> stupid kline

14:51:50 <dajobe> ... danbri unstable and testing

14:51:53 <libby> is whatsit about?

14:52:19 <dan-uk_> you're all ssh tunnelling now?

14:52:22 <shellac> yep

14:52:22 <dajobe> Q: suggestion fon irc for a galway core, namespace, joke or not

14:52:28 <libby> yeah

14:52:28 <shellac> they've banned us :-(

14:52:35 <dajobe> A: more about process and documenting

14:52:35 <imajes> i fixed the kline

14:52:39 <golbeck> i'm using cgiirc

14:52:40 <imajes> you should be able to get in

14:52:40 <libby> thanks!

14:52:43 <dajobe> haven't agreed the galway core set of elements

14:52:44 <shellac> cheers!

14:52:48 <imajes> it's just because too many of you were signed in

14:52:49 <libby> tell john b

14:52:51 <shellac> bye!

14:52:51 <imajes> can someone try it out?

14:52:56 <dajobe> (I see others are still not getting in, such as JohnB)

14:53:04 <libby> try again?

14:53:18 <dajobe> danbri - can more peole do translation work?

14:53:23 <dajobe> ... please type your names into irc

14:53:28 <mortenf> i'll try some translation to danish

14:53:38 <libby> yay mortenf

14:53:53 <dajobe> danbri - my issues are cleanups for foaf f

14:54:17 <dajobe> maybe everyone who has a nmaespace has an rss feed for their namespace changes

14:54:42 <crschmidt> mortenf, might poke Freso re that as well ( http://www.livejournal.com/users/freso/info )? I can't remember what languages he speaks, but i think danish *might* be one of them

14:54:44 <dajobe> ? - yeah keepiung up with the changes

14:54:48 <dajobe> ^- plaxo

14:55:03 <dajobe> danbri - points at ontaria

14:55:05 <mortenf> it is

14:55:07 <dajobe> by sandro hawke

14:55:19 <dajobe>http://www.w3.org/2004/ontaria/

14:55:31 <dajobe> similar to schemaweb but a slightly different approach

14:55:48 <dajobe> funded by daml work in the US

14:56:07 <dajobe> prefer to look at the w3c for scalability and stabiliity that you might want

14:56:36 <dajobe> There is a possibility of ontology hosting for namespaces that are stable and won't go away.

14:56:39 <dajobe> the message is tricky

14:56:55 <dajobe> the point of rdf was to get away from centralising

14:57:11 <dajobe> MarcC - list of namespaces?

14:57:14 <dajobe> danbri - schemaweb

14:57:23 <mortenf> uldis volunteers to do latvian translation - yay!

14:57:26 <dajobe> MarcC - maybe a list of namespace on the foaf site, please

14:57:34 * jsled sighs

14:57:58 <dajobe> danbri shows sir' tim's head on schemaweb

14:58:04 <dajobe> does a schemaweb demo

14:58:34 <libby> nice one uldis!

14:58:51 <Cloud> we're back

14:58:58 <libby> schemaweb.info

14:59:14 <Cloud> thanks for the tunnel d

14:59:19 <dajobe> danbri - one issue with rdf depolyment was ..

14:59:37 <crschmidt> lilo says everything should be set now

14:59:38 <dajobe> ... there is no way to do hypertext documetnation in the schema language

15:00:04 <dajobe> documenting such things is one thing the w3c will get into, infrastructure for that

15:00:21 <shellac> Cloud: no problem. Watch rdfweb get k-lined :-)

15:00:21 <dajobe> danbri shows DOAP on www.schemweb.info

15:00:34 <dajobe> the foaf style of deployment

15:00:42 <levin> hi all.... sorry for the conflict with our clonebot detection, and welcome to freenode :)

15:00:54 <dajobe> hi lilo, er levin

15:01:05 <dajobe> again, we should have thought ahead :)

15:01:21 <levin> we're hoping the interim release of the server code will help solve this one 8)

15:01:33 <shellac> levin doesn't realise we're all clones

15:01:35 <libby> yeah thanks for unbanning us levin :)

15:01:45 <Fides> you mean we aren't?

15:01:49 <dajobe> danbri scrolls down past the foafcamp pics on schemaweb

15:01:49 <levin> libby: glad to help 8)

15:02:03 <dajobe> danbri [heart] schemaweb

15:02:22 <levin> oh, I already have a client here 8)

15:02:23 * levin waves

15:02:55 <dajobe> Marc: please make the site simpler

15:02:59 <dajobe> and clearly state the process

15:03:10 <dajobe> that somebody who wants toa dd something, what they exactly have to do

15:03:13 <dajobe> a Process Guide

15:03:32 <dajobe> maybe, a case study of what you have to do

15:03:33 <jen> jen is now known as golbeck

15:04:04 <dajobe> aharth - what should be the sequence to change the spec then make ppl use it, or see how people use thigns, then change the spec to match?

15:05:06 <dajobe> danbri, in foaf we deploy stuff and then see what's used out there, then add to spec. not entirely working since it wasn't in the spec

15:05:11 <dajobe> ^- historically

15:05:43 <dajobe> danbri schedules new foaf website for Oct 2004

15:05:57 <dajobe> danbri - the url for namespace is not likely to change

15:06:07 <dajobe> we are doing versioning at the level of terms

15:07:00 <dajobe> danbri points at planetrdf.com

15:07:10 <dajobe> which he says was simple; but actually was rss messy stuff

15:07:31 <dajobe> danbri might do something for foaf similar

15:07:45 <davidoc>http://www.ldodds.com/foaf/foaf-a-matic.html

15:07:46 <dajobe> also might add foaf-amatic into the foaf site

15:07:52 <dajobe> 10 language versions - nice!

15:08:01 <dajobe> fantastic that these foaf gadgets are appearing

15:08:08 * ldodds remembers he needs to add another

15:08:38 <ldodds> chinese in case anyone is wondering

15:08:38 <dajobe> danbri a fewthings - making into a cosnistnet web site

15:09:04 <dajobe> TomC - somebody has made an autodiscovery thing that uses google to help make foaf

15:09:16 <dajobe> sh1m, sh1mmer - please add the url

15:09:31 <sh1m>http://www.tecknik.net/autofoaf/

15:09:33 <dajobe> danbri - RDF DAWG - query language and network protocol

15:09:42 <dajobe> use cases have foaf

15:10:01 <dajobe> trying to ground in useful stuff

15:10:16 <dajobe> want it to be small (!XQuery)

15:10:29 <dajobe> and please comment on this doc

15:10:52 <dajobe> other W3C group is Semantic Web best Practices AND DEPLOYMENT

15:11:38 <dajobe> SWBPD has published a note with some modelling choices for semweb vocabs

15:12:11 <dajobe> danbri - think the w3c will do more things that are friendly to foaf

15:12:51 <dajobe> help with the wiki gardening

15:12:51 <libby> http://rdfweb.org/topic/DataSources has various companies etc in it

15:13:14 <dajobe> if you have a weblog on foaf, send feed url to danbri (or make a wiki page?)

15:13:46 <gromgull> So what pub are we going to?

15:13:47 <dajobe> MarcC: next yera?

15:14:00 <dajobe> danbri - yes, not sure if it's foaf, wider

15:14:11 * libby suggests same pub as last time for later: Tigh Neachtain, Corner of Cross Street and Quay Street

15:14:11 <dajobe> there are semweb confs, web confs on annual schedules

15:14:22 <dick_> let's talk about the pub destination

15:14:27 <libby> as in: http://rdfweb.org/topic/FoafGalway

15:14:32 <dajobe> tomC - go to 3 days?

15:14:37 <teefal>http://photobingo.com

15:14:41 <teefal> is up

15:14:43 <libby> god, I don't think I could take it

15:14:45 <teefal> tread lightly

15:14:53 <Cloud> I wonder do we need somewhere bigger :)

15:15:08 <dajobe> libby: evaluation forms will be web based, afterwards. we'll email you

15:15:24 <gromgull> teefal : how do I participate?

15:16:31 <dick_> time for a Guiness!!!

15:16:40 <dajobe> stefanD - closing remarks

15:16:42 <teefal> bingo?

15:16:51 <gromgull> yes

15:17:11 <dajobe> calls down JohnB, Ina and AndreaH for thanks

15:17:17 <dajobe> bnch of flowers for all three

15:18:08 <shellac> <----- aharth

15:18:20 <shellac> he moved :-(

15:18:53 <libby> stefan! don;t forget about foaf bingo!

15:18:55 * sh1m wonders about danbri in #frenchies on some irc network

15:19:01 <dajobe> stefanD - add more things in the wiki, about sessions

15:19:23 * ldodds plays foaf bingo. very cool

15:19:36 <dajobe> foaf bingo - teefal, Timothy Falconer

15:19:50 <dajobe> the best thing is it has all of you

15:19:57 <dajobe> ... photos of ...

15:20:13 <libby> photobingo.com

15:20:18 <dajobe> shows http://photobing.com/

15:20:23 <dajobe> eek, what libby said

15:20:33 <dajobe> shows the tags page

15:20:34 <libby>http://www.w3.org/2001/sw/Europe/events/foaf-galway/papers/pp/foaf_to_the_people/

15:20:44 <dajobe> at thist ime 90 2004-09-01 photos

15:20:56 <shellac> dajobe: photograph the corpse of bing crosby?

15:21:08 <dajobe> storymill creates rss feeds that photobingo reads

15:21:24 <libby> (lots of photos here btw: http://del.icio.us/tag/foaf-galway)

15:21:24 <dajobe> shows the /photos/ url

15:21:34 <dajobe> showing the tags, annotations below

15:21:55 <dajobe> you can do regions in tidepool, not showing now but will do

15:22:05 <dajobe> MarcC - can you do co-depiction?

15:22:55 <dajobe> teefal shows playing the game, who what where when

15:23:38 <dajobe> the way to win is to get the most points; by taking pics, uploading htem an annotating them

15:23:42 <dajobe> people who play will get a t-shirt

15:23:50 <dajobe> (the universal foaf currency)

15:24:30 <dajobe> danbri and others rush to get points

15:25:02 <dajobe> "mulligan" - a golf term, for harder things to get. costs you points though

15:25:36 <dick_> :( site is down

15:25:51 <dajobe> we noticed

15:27:16 <shellac> we broke it :-(

15:27:26 <iand> foafdotted

15:27:34 <dick_> it is confusing

15:27:47 <dick_> I just clicked links and got Bingo twice before it crashed

15:28:00 <gromgull> what stops me from just clicking like crazy?

15:28:41 <dick_> that is what I was doing :)

15:29:09 <dajobe> heh foafdot

15:29:18 <dick_> cool -- i got a servlet exception now

15:29:28 <Cloud> I got some source code

15:29:36 <crschmidt> heh

15:29:40 <sh1m> hah

15:29:42 * ldodds is happily playing

15:29:44 <crschmidt> not quite load tested :)

15:29:46 <gromgull> 500 Servlet Exception

15:29:46 <gromgull>

15:29:46 <gromgull> java.lang.NullPointerException

15:29:46 <gromgull> at com.photobingo.card.BingoTest.hasBingo(BingoTest.java:32)

15:29:46 <gromgull> at com.photobingo.action.PbBingoAction.perform(PbBingoAction.java:34)

15:29:46 <gromgull> at org.apache.struts.action.ActionServlet.processActionPerform(ActionServlet.java:1787)

15:29:49 <gromgull> at org.apache.struts.action.ActionServlet.process(ActionServlet.java:1586)

15:29:51 <gromgull> at org.apache.struts.action.ActionServlet.doGet(Action.......

15:29:53 <dajobe> teefal says possible all the tags have been called, will be reset

15:30:12 <dajobe> after discussion with libby at www2004 for making fun games for foaf and rdf

15:30:35 <dick_> we can email in the exception :)

15:30:39 <dajobe> cruel people take pictures of the crashes on the web pages ;)

15:31:01 <libby>http://moblog.nicecupoftea.org/

15:31:36 <dajobe> danbri - help make things fun. use the collecting instinct, rewards

15:32:19 <libby> sorta works: marc to jfk: http://swordfish.rdfweb.org/discovery/2002/02/paths/byname.jsp?name=canter&name2=kennedy

15:32:30 <dajobe> StefanD

15:33:01 * dajobe dies

15:33:13 <Cloud> www.KingsHead.ie

15:33:16 <dajobe> a pub

15:33:19 <Cloud> But not much there apart from a logo

15:33:22 <Cloud> or maybe www.thekingshead.ie

15:33:31 <Cloud> It's on Shop Street, the main street

15:35:01 <gromgull> Who is going to ISWC?

15:35:11 <shellac> danbri is

15:35:53 <dajobe> - - - End of Foaf Workshop Galway - - -

15:36:02 <gromgull> When are we meeting in the pub?

15:36:05 * crschmidt cheers.

15:36:07 <dajobe> asap

15:36:14 <ndw> Congrats, folks!

15:38:35 <libby> thanks ndw :)

15:39:18 <libby> ...and thanks for being on our PC :)

15:39:39 * crschmidt is looking forward to regular irc meetings

15:39:40 <ndw> thanks, libby. I wish I'd been able to do more :(

15:39:56 * ldodds wishes he could go and help them drink beer

15:40:12 <ndw> heh, that too

15:41:17 <JibberJim> so what was the conclusion from the conference?

15:43:37 <JibberJim> danbri leaving was the conclusion - eek!

15:45:45 <libby> yep that was it

15:45:49 <dajobe> conclusion? how about "danbri - help make things fun. "

15:45:57 <libby> wish you coulda come all!

15:46:00 <libby> later

15:49:07 <crschmid1> crschmid1 is now known as crschmidt

16:06:39 <crschmidt> Hey Freso

16:07:22 <crschmidt> we were talking earlier about translating the FOAF terms into languages (at least the stable ones) and mortenf mentioned he would give Danish a try (I think) -- you might want to talk to him if you're interested in helping out

16:08:31 <Freso> Hey Chris, and I'll talk to him :)

16:08:42 <inkel> Hi crschmidt , I could help with Spanish also

16:09:02 <ldodds> why not contact all the helpful folk that translated the foaf-a-matic

16:09:10 <inkel> :)

16:09:12 <ldodds> inkel and mortenf were among them

16:14:49 <inkel> How many stable terms are in foaf ?

16:18:37 <crschmidt> hm.

16:18:39 <crschmidt> one.

16:18:40 <crschmidt> homepage.

16:21:44 <crschmidt> So, I guess we need to mark some more terms as stable :)

16:22:55 <Freso> Hehe, yeah :p

16:25:13 <crschmidt> ^query select ?term where (?term vs:term_status "stable") using vs for <http://www.w3.org/2003/06/sw-vocab-status/ns#>

16:25:14 <redlandbot> [http://xmlns.com/foaf/0.1/homepage] ,

16:25:22 <crschmidt> ^part

16:28:12 <Freso> What is the requirement for being classified as stable?

16:28:38 <ldodds> I think generally that there's running code using it

16:28:59 <ldodds> but as crschmidt has already discovered the statuses are in dire need of updating

16:30:44 <Freso> Hm. Thought.

16:31:26 <Freso> How about an "InternetCode" or "DescriptionCode" or something to supclass GeekCode and would be there for other codes to subclass?

16:45:22 <Cloud2> Cloud2 is now known as Cloud

17:11:38 <balbinus_> balbinus_ is now known as balbinus

19:20:53 <dmwaters> {global notice} Hi all! Freenode is in the process of building a new ircd to replace the current dancer1.0. This needs to be done soon, and the code does need to be clean. If anyone is interested on helping code on this project, please stop by #newircd and talk to us. Have a nice day, and thank you for using freenode!

19:50:32 <CaptSolo> hi all :)

19:50:48 <crschmidt> Hi CaptSolo

19:51:09 <CaptSolo> hi chris :)

19:51:32 <CaptSolo> i've returned to irc after long silence

19:51:43 <CaptSolo> i did not have a laptop with me

19:51:47 <crschmidt> heh

19:51:51 <crschmidt> You were at Galway, right?

19:51:54 * CaptSolo chatting from the Dublin airport

19:51:58 <CaptSolo> yep

19:52:02 <crschmidt> SOmeone mentioned you volunteerign to do a translation of foaf terms

19:52:02 <CaptSolo> was fun :)

19:52:13 <crschmidt> (Of course, we later discovered that there's only one stable term in the schema)

19:52:20 <CaptSolo> lots of interesting things, people to meet, ... well - oyu saw the channel logs didn't you?

19:52:34 <crschmidt> some of them, haven't had a chance to read things from when I was asleep

19:52:42 <CaptSolo> heh, i do not mind translating the only stable term - saves work for me

19:52:47 <CaptSolo> it was morten who mentioned that

19:52:50 * crschmidt nods

19:53:47 <CaptSolo> photobingo - service temp unavailable

19:53:53 <CaptSolo> how are you chris?

19:54:01 <CaptSolo> you didn't make it to galway?

19:54:15 <crschmidt> No, no dinero, + needed to work

19:55:10 <CaptSolo> you were on the program comittee, how could you not come? :)

19:55:15 <crschmidt> Heh

19:55:16 * CaptSolo just joking

19:55:16 <EtherealShifter> So, while there actually seems to be someone here, I'll throw in my thought from earlier again...

19:55:17 <crschmidt> No money!

19:55:22 <CaptSolo> what did you do instead?

19:55:22 <EtherealShifter> EtherealShifter is now known as Freso

19:55:35 <CaptSolo> ether: go for it!

19:57:06 <CaptSolo> freso: so... your thoughts?

19:59:52 <Freso> Sorry, sec.

19:59:57 <Freso> (Gah.)

20:01:38 <Freso> How about adding InternetCode or DescriptionCode to supclass geekCode, which would also be subclassable if people want to add similar codes to other namespaces (or perhaps even if people decide on adding another code to the standard FOAF vocab)?

20:01:54 <CaptSolo> photobingo.com seems to have a) scalability / load problems or b) bugz that crash it

20:02:24 <crschmidt> I think someone mentioned that it has a certain number of tokens to give out

20:02:30 <crschmidt> and needs to be reset afterwards, possibly

20:02:33 <CaptSolo> freso: why not?

20:03:28 <Freso> CaptSolo: Well, I would love to see it :p I was just wondering whether there'd be any compelling reasons not to do it.

20:03:32 <CaptSolo> freso: not sure that would make into the foaf spec - you have to persuade danbri and libby that its needed, but i think you can use it. make your own subclasses of geekCode and use them

20:04:37 <Freso> Yes, well, this would *super*class geekCode (as in, geekCode being a subclass of this) and would also allow other codes to be subclassed from it.

20:05:24 <ronwalf> Is geekCode a uri or a literal?

20:06:09 <ronwalf> Do you mean sub property, perhaps?

20:06:22 <Freso> Literal

20:07:00 <Freso>http://xmlns.com/foaf/0.1/#term_geekcode

20:08:01 <Freso> (The text about geekcode really needs updating, btw. It's a great mix of cases...)

20:10:32 <Freso> 'sub property'?

20:10:49 <Freso> Ooh! Right.

20:11:45 <Freso> geekcode would be a subproperty of InternetCode/DescriptionCode/FoobarCode

20:16:33 <Freso> Hm...

20:19:33 <Freso> Something like this:

20:19:36 <Freso> <rdf:Property rdf:about="http://xmlns.com/foaf/0.1/InternetCode" vs:term_status="testing" rdfs:label="InternetCode" rdfs:comment="Some form of coded information about the object.">

20:19:36 <Freso> <rdfs:domain rdf:resource="http://xmlns.com/foaf/0.1/Agent"/>

20:19:36 <Freso> <rdfs:range rdf:resource="http://www.w3.org/2000/01/rdf-schema#Literal"/>

20:19:36 <Freso> <rdfs:isDefinedBy rdf:resource="http://xmlns.com/foaf/0.1/"/>

20:19:36 <Freso> </rdf:Property>

20:19:38 <Freso> <rdf:Property rdf:about="http://xmlns.com/foaf/0.1/geekcode" vs:term_status="testing" rdfs:label="geekcode" rdfs:comment="A textual geekcode for this person, see http://www.geekcode.com/geek.html">

20:19:41 <Freso> <rdfs:domain rdf:resource="http://xmlns.com/foaf/0.1/Person"/>

20:19:43 <Freso> <rdfs:range rdf:resource="http://www.w3.org/2000/01/rdf-schema#Literal"/>

20:19:45 <Freso> <rdfs:isDefinedBy rdf:resource="http://xmlns.com/foaf/0.1/"/>

20:19:47 <Freso> <rdfs:subPropertyOf rdf:resource="http://xmlns.com/foaf/0.1/InternetCode"/>

20:19:49 <Freso> </rdf:Property>

20:19:51 <Freso> I think.

21:21:48 <crschmidt> I've added http://rdfweb.org/topic/IrcMeeting to the wiki

21:22:01 <crschmidt> In an ttempt to spur on continued involvment playing off Galway and FOAFcamp

21:22:11 <crschmidt> with a short explanation of why, where, and what

21:22:15 <crschmidt> Now all I need is a who :)

21:22:57 <crschmidt> I'll email rdfweb-dev when i get home

21:24:56 <ear1grey> crschmidt: thanks for putting that up, it's a great idea that shouldn't be allowed to fizzle.

21:27:06 <crschmidt> I think a lot of people are excited about FOAF right now

21:27:20 <crschmidt> So I want to keep them excited with scheduled discussions