Friend of a Friend (FOAF) IRC Chat Logs for 2003-12-11

01:06:43 <Davey> Davey is now known as D[a]vey

10:13:35 <oPless> what a grotty day today :/

10:44:42 <mattb> so when did ecademy stop doing FOAF?

10:53:39 <mattb> ah, they didn't

10:54:04 <mattb> but compare http://www.ecademy.com/module.php?mod=network&op=foafrdf&uid=38506 to http://www.ecademy.com/module.php?mod=network&op=foafml&uid=38506

10:54:20 <mattb> and see the foafml.com strapline "FOAFml - FOAF just grew up"

10:54:34 <libby> !

10:55:06 <mattb> ! indeed

10:56:44 <mattb> did a scutter run last night, got ~1m triples

10:56:51 <mattb> think it crashed though, there will be more

10:57:27 <mattb> lots of predicates in http://foafml.com/0.5/

11:00:08 <danja> foafml? when did that appear? any more info?

11:00:21 <mattb> <mattb> but compare http://www.ecademy.com/module.php?mod=network&op=foafrdf&uid=38506 to http://www.ecademy.com/module.php?mod=network&op=foafml&uid=38506

11:00:21 <mattb> <mattb> and see the foafml.com strapline "FOAFml - FOAF just grew up"

11:00:28 <mattb> foafml.com registered 11th august 2003

11:00:46 <mattb> by Sentient Ltd

11:00:48 <mattb> isn't that thomas power?

11:01:00 <mattb> Graham Briggs admin contact, firevision.co.uk

11:01:53 <nmg> Erm, could be

11:02:16 <nmg> sentient do/did Freddie

11:02:59 <nmg>http://tellfreddie.com/

11:03:42 <mattb> "There is no easy way to explain RDF!"

11:03:58 <nmg> so...FOAFml : FOAF :: RSS2 : RSS1.0 ?

11:04:15 <mattb> i assume so

11:04:18 <mattb> a syntactic profile

11:04:42 <mattb> actually maybe it's more like FOAFml : FOAF :: RSS 1.0 : the RDF model of RSS 1.0

11:05:16 <mattb> since http://www.ecademy.com/module.php?mod=network&op=foafml&uid=38506 still parses as rdf

11:06:31 <danja> it just looks like FOAF with Person having a URI

11:07:36 <mattb>http://rdfweb.org/pipermail/rdfweb-dev/2003-August/011899.html

11:07:50 <mattb> seems to be the only real reference to it

11:14:40 <danja> a mail to the address given just bounced

11:25:58 <danja> the XML schema-compatible profiling may be worthwhile

11:26:08 <danja> but it don't need a new namespace

11:26:30 <mattb> i don't really see the benefit myself

11:26:34 <mattb> if it's to be an rdf app

11:26:47 <mattb> if you assume 'foafml' and process as xml, you get only x% of the foaf world

11:26:48 <danja> XML apps can produce it

11:26:59 <mattb> whereas rdf processors will get the foafml and everything else too

11:27:19 <mattb> for output, xml apps can produce any xml profile of foaf they like

11:27:35 <mattb> unless you're working in a specific point-to-point interchange where you control the producer and the consumer

11:27:38 <danja> oh yeah

11:28:34 <danja> but there are loads of non-RDF RSS 1.0 readers

11:29:33 <mattb> yeah, but the real benefit of foaf comes from vocab mixing

11:29:48 <mattb> whereas you can get a lot out of rss with a few tags

11:30:00 <danja> indeed - same with RSS

11:30:02 <mattb> i can see the foafml thing being interesting to someone producing a large closed-world app

11:30:04 <mattb> like ecademy

11:30:14 <mattb> because it's oriented around foaf-style info

11:30:56 <mattb> rss was xml based partly because of backwards compatibility

11:31:00 * danja a little out of sync

11:31:14 <mattb> to make adapting rss0.9 readers to 1.0 easy

11:31:18 <mattb> foaf doesn't have that legacy

11:31:33 <mattb> except that working with rdf is 'harder' than xml in some people's eyes

11:31:47 <danja> yep

11:31:54 <mattb> basically i don't see what problem foafml would be solving

11:31:58 <mattb> apart from tool support and lack of coder knowledge

11:32:07 <mattb> i don't see an actual *benefit* from it

11:32:10 <mattb> it's just a workaround

11:32:21 <mattb> rant off :)

11:33:40 <danja> there's certainly the perception that XML is easier

11:34:16 <danja> but I don't think a lot of folks that take that angle

11:34:40 <danja> can have played much with graph data

11:34:59 <danja> (which is strange, given the web)

11:36:40 <danja> my guess is ecademy already have there own fields

11:37:17 <danja> and are trying to tie FOAF to them

11:38:41 <danja> but I don'

11:39:00 <danja> (heh) 't know why they've opted for another ns for core terms

11:39:18 <danja> unless they really *want* to make an island of themselves

11:43:03 * danja wonders why they didn't use OPML...

11:54:23 <gromgull> Hi all, does anyone have a perl snippet for generating sha1sums?

11:55:14 <mattb> use Digest::SHA1 qw(sha1_hex);

11:55:19 <mattb> print sha1_hex("mailto:foo");

11:55:31 <gromgull> excellent - cheers

11:56:06 <mattb> hmm

11:56:12 <mattb> except that produces something different from what's in my oaf

11:58:01 <mattb> but it does seem to be what people on the web have used

11:59:48 <mattb> looks like it's my problem

12:00:37 <mattb> ah yes, my mistake

12:00:44 <mattb> stupid, wasn't escaping an @ in a perl "" string

12:01:13 <gromgull> :)

12:01:38 * mattb wonders if anyone made a sha1sum code snippets page on the wiki in the end

12:01:38 <gromgull> Now all I have to do is to convince our sysadmin to install the Digest module...

12:01:48 <mattb> you could also shell out to commandline sha1sum :)

12:02:19 <mattb> but if you do, make sure you don't pass it a newline at the end and alter the sum

12:15:44 <adr> It seems from reading a bit of the FOAF specification that a lot of it is still pretty unstable...is this so?

12:19:51 <danja> some terms are pretty solid - e.g. foaf:knows

12:20:25 <adr> What about the naming ones?

12:20:32 <adr> I see there's quite a few...

12:20:51 <danja> they're flagged 'unstable' etc

12:20:56 <adr> I mean what's the difference between firstName and givenname?

12:21:33 <danja> given name might come second...

12:21:50 <danja> naming is one of the tricky areas

12:22:05 <danja> internatiolization, cultural differences etc

12:22:14 <adr> You mean like if a person were to use their "second" name as their first?

12:22:23 <adr> Oh I see

12:22:35 <libby> there are lots and lots of problems with internationalization and names. we didn;t realise at first. given and family are I think better than first and second

12:23:37 <adr> And plain 'name' describes a person's full name? First, middle and last?

12:23:48 <danja> yep - even here in Italy you'd generally use Miller, Libby

12:24:02 * danja still out of sync :-(

12:24:08 <adr> heh

12:24:52 <adr> Should I organize my 'name' like that then? eg. Smith, Joe William

12:24:54 <libby> here's some of the issues to give you a flovour of the issues: http://www.wasab.dk/morten/2003/04/foaf-names-proposal-1.html

12:25:13 <adr> Ah thanks

12:25:28 <libby> write you name as you'd like i to be written, like a human-reeadable label

12:25:37 <libby> sorry about tha typing

12:25:44 <libby> like 'it' to be written

12:26:19 <adr> I see, since theoretically a FOAF 'browser' will just be organizing the data and not parsing it?

12:26:46 <libby> I think we decided that we'd probbaly keep name, givenname and familyname, and leave the rest to another namespace

12:27:01 * libby gotta run. this name stuff is high on the list

12:27:19 <adr> Okay, thanks (:

12:27:35 <libby> wrt unstable etc, many of the unstable ones can be fairly quickly changed to stable, to...

12:27:39 <libby> seeya

12:28:22 <oPless> libby.speling++

13:06:16 <MrDave> foafbot, MrDave's name

13:06:18 <foafbot> Can't identify 'MrDave', sorry.

13:08:28 <MrDave> foafbot, foafbot's name

13:08:29 <foafbot> Can't identify 'foafbot', sorry.

13:08:37 <MrDave> foafbot, workbench's name

13:08:38 <foafbot> Can't identify 'workbench', sorry.

13:08:45 <MrDave> foafbot help

13:08:51 <MrDave> foafbot

13:32:03 <oPless> foafbot, edd

13:52:45 <danja> can anyone see http://62.11.123.210 ?

13:56:45 <kota> i see "Lists"

13:57:50 <danja> thanks Kota

13:58:04 * danja is too insecure for his shirt

14:07:19 <oPless> danja, I see your problem, thats not a shirt you're wearing. It's a big girls blouse ! ;-)

14:07:42 * oPless hopes that joke translates to italian

14:11:25 <danja> try that address again

14:11:38 <danja> - oPless

14:13:04 <oPless> I don't wear dresses...

14:13:11 <oPless> oh, the url, RIIGHT :)

14:13:23 <danja> dear oh dear...

14:13:34 <oPless> thats cool

14:14:02 <oPless> sorry, I'm in a positivly odd mood today

14:15:12 <danja> I've been wandering around without a firewall

14:15:30 <danja> very undressed

14:15:44 <oPless> musta been a bit chilly

14:16:11 <danja> aah, that's better - found a rug

14:16:58 <oPless> you have a wall on fire!

16:01:25 <Davey> really, the FOAF stuff is quite recursive :/

16:02:10 <libby> eh?

16:02:14 <libby> why d'you say that?

16:02:36 <Davey> the fact that <foaf:knows> can contain full FOAFs if you were that way inclined.

16:03:32 <libby> hm, that's just the way rdf works - it's not a document-based format, it's about graphs, which arbitrarily may be stored in whatever files you like

16:03:48 <danbri> davey, http://rdfweb.org/pipermail/rdfweb-dev/2002-October/010439.html

16:04:18 <libby> that's why much better to use an rdf parser than an xml one

16:04:35 <libby> though of course if you're controlling the format (if) then you could parse it how you like

16:05:22 <Davey> but I'm *creating* the RDF/XML and because RDF parsers don't care about the way its created (i.e I can choose any number of valid ways to create RDF/XML) so I can use XML tools for this.

16:06:18 <libby> so what's the problem? you can just choose a nice format to output the data in...?

16:06:20 <Davey> danbri: not entirely sure what you're pointing out with that.

16:06:38 <danbri> just re recursion, depth, sillyness

16:06:41 <Davey> libby: but no matter the format, there is recursion

16:07:11 <Davey> code wise, I mean

16:07:23 <Davey> not necessarily in the XML itself

16:08:26 <Davey> http://pixelated-dreams.com/blog/foaf.rdf <--- I can create that entire FOAF with my class already (except the seeAlso, haven't coded that yet) but it requires me to pass around an object. Which is nasty :/

16:11:39 <libby> so why not limit what people can say about other people, e.g. just name/sha1mbox/foaffile, like foaf-a-matuc does?

16:12:01 <Davey> because I'd still have repeat code. :/

16:12:40 <Davey> I might need to have a XML_FOAF::knows and XML_FOAF::endKnows

16:12:51 <Davey> that would work... then I know where to add the data.

16:16:10 <Davey> ok, I know how this is going to work. :)

16:17:04 <libby> hurrah!

16:18:05 <Davey> or... not :/

16:18:14 * Davey bends his mind around a pole

16:38:24 <Davey> w00t!

17:00:04 <Davey> Hmm... shouldn't FOAF have a version string somewhere?

17:00:33 <Davey> so when you change it as planned, both FOAF 0.1 (??) and FOAF whatever can be parsed still

17:00:51 <Davey> wait... I guess the FOAF xmlns resolves that :)

17:01:17 <libby> hm, yeah, but that was a miosake. we then decided to do it property by property

17:01:19 <danbri> we keep the ns uri stable, edit it in-place

17:01:32 <danbri> 0.1 is 1.0 backwards

17:01:52 <Davey> ok, thats confused me

17:02:12 <Davey> I have XML_FOAF's theory out the way now, just need to make it add all the elements in FOAF 0.1 properly :)

17:02:57 <Davey> in any event, it does all that foaf-a-matic does right now (I believe, I need to check that I didn't miss anything out when creating my FOAF with it)

17:04:50 <Davey> ok, most of it :)

17:08:35 <libby> cool Davey. are you using first and last or given/family? latter woudl probbaly be better I think

17:08:57 <Davey> libby: all will be available soon :)

17:09:51 <Davey> I have in place the framework (functions with no content yet) to add all element of foaf:Person, theres only a few others needed for foaf:Group or foaf:Organization

17:10:04 <libby> cool

17:10:19 <Davey> then I'm going to abstract it in XML_FOAF_Lite :D

17:12:15 <Davey> I need to create XML_FOAF_Parser too

17:12:32 * adr waves to Davey

17:12:44 <Davey> Hey adr :)

17:13:05 <Davey> gah! this class is going to take FOREVER to add phpdoc too

17:14:11 <Dorward> Davey: Well .. if you documented as you coded :)

17:14:35 <Davey> Dorward: Heh

17:14:50 <libby> davey, is there not a php rdf parser you can use?

17:15:00 <Davey> libby: indeed, RAP :)

17:15:06 <libby> ah, cool

17:15:15 <libby> just didn;t want you to start form scratch

17:15:15 <Davey> which is going to be ported to PEAR and cleaned up :)

17:15:23 <Davey> libby: nah, not doing that ;)

17:15:52 <libby> good :)

17:16:05 <Davey> but I will abstract it so the user is clueless :)

17:38:24 <Davey> Davey is now known as D[a]vey

17:57:42 * mattb goes to the pub

18:01:06 <oPless> <Davey> but I will abstract it so the user is clueless :)

18:01:16 <oPless> isnt that always the case ?:)

18:07:37 <JimJibber> JimJibber is now known as JibberJim

18:17:35 <JibberJim> !picpath jim to chris

18:17:35 <whwhwhwh> try !paths name1, name2 to find images linking people. Use the fullest form of name possible for best results

18:17:40 <JibberJim> !picpath jim, chris

18:18:56 <libby> whwhwhwh isa bit tired today for some reason

18:19:11 <whwhwhwh> Jim Hendler to Chris Lilley via Eric Miller, Dan Connolly

18:19:12 <whwhwhwh> in 3 steps

18:19:13 <whwhwhwh> Jim Hendler to Chris Mitchell via Dave Beckett, Jan Grant

18:19:14 <whwhwhwh> in 3 steps

18:19:15 <whwhwhwh> Jim Ley to Chris Mitchell via Jim Ley, Damian Steer, Martin Poulter

18:19:16 <whwhwhwh> in 4 steps

18:19:17 <whwhwhwh> Jim Ley to Chris Lilley via Jim Ley, Dave Beckett, Dan Connolly

18:19:18 <whwhwhwh> in 4 steps

18:19:19 <whwhwhwh> Jim Ley to Chris Mitchell via Damian Steer, Martin Poulter

18:19:20 <whwhwhwh> in 3 steps

18:19:20 <JibberJim> ah, a heavy night last night maybe?

18:19:22 <whwhwhwh> Jim Ley to Chris Lilley via Dave Beckett, Dan Connolly

18:19:24 <whwhwhwh> in 3 steps

18:19:26 <whwhwhwh> Jim Ley to Chris Mitchell via Jim Ley, Damian Steer, Martin Poulter

18:19:26 <mortenf> :)

18:19:28 <whwhwhwh> in 4 steps

18:19:30 <whwhwhwh> Jim Ley to Chris Lilley via Jim Ley, Dave Beckett, Dan Connolly

18:19:32 <whwhwhwh> in 4 steps

18:19:34 <whwhwhwh> Jim Hendler to Chris Lilley via Dan Connolly

18:19:36 <whwhwhwh> in 2 steps

18:19:38 <whwhwhwh> Jim Hendler to Chris Mitchell via Dan Brickley, Martin Poulter

18:19:40 <whwhwhwh> in 3 steps

18:19:42 <whwhwhwh> see http://swordfish.rdfweb.org/discovery/2002/02/paths/byname.jsp?name=jim&name2=chris

18:25:52 <D[a]vey> ok, I have foaf-a-matic down :)

18:26:45 <mortenf> nice!

18:27:05 <D[a]vey> is the foaf:logo derived of foaf:depiction? (as foaf:img is)

18:27:16 <JibberJim> foaf:logo ?

18:27:17 <D[a]vey> D[a]vey is now known as Davey

18:27:42 <Davey> JibberJim: believe thats it, lemme check

18:28:02 <Davey> hmmm... wonder where that came from.

18:28:06 <mortenf> no, not related

18:28:13 <mortenf> but it's in there

18:28:31 <mortenf> supposed, i think, to be like foaf:img for foaf:Organization's

18:28:34 <Davey>http://xmlns.com/foaf/0.1/#term_logo

18:29:07 <mortenf> didn't someone suggest a foaf:avatar predicate?

18:29:22 <JibberJim> Yeah.

18:29:46 <JibberJim> A few people, I wanted something generic whereby you could say "image of me for service X"

18:30:14 <mortenf> so, it should have a domain of Account instead of Person?

18:30:53 <JibberJim> No, is that what foaf:logo is, I've never heard of it?

18:31:10 <JibberJim> but what I wanted was headshots for a foafnaut like thing where the blubs contained images.

18:31:17 <mortenf> but if you want to tie it to a service?

18:35:54 <JibberJim> Erm, yeah, got disconnected there, and loads of stuff I wrote never made it :-(

18:36:05 <mortenf> :(

18:36:36 <JibberJim> Yes, I'd like it attached to both a person and a service, which gives us the usual problem of foaf:imageForFoafnaut being dead easy, and the generalised solution being complicated.

18:37:17 <mortenf> which is why i suggested hanging it off of an OnlineServiceAccount

18:37:29 <mortenf> (the "complicated" version)

18:37:34 <JibberJim> Ah, it seems this end of the cafe is dodgy signal area (which is odd as the guy outside is even further in this direction)

18:37:59 <JibberJim> Yeah, but it's not just for online services really... - you could use it to denote your passport photo too.

18:38:21 <mortenf> yeah, ok, i see that - but that doesn't retract from the usefulness of the online types

18:38:54 <JibberJim> no

18:39:46 <mortenf> but you are right that OnlineAccount is too strict for general use

19:47:09 <edd> foafbot, analyse edd and danbri

19:47:43 <edd> foafbot, analyze edd and danbri

19:47:45 <foafbot> Edd Dumbill (ENFJ) and Dan Brickley (INTP) could learn a lot from each other.

19:47:45 <foafbot> Computed using information from Edd Dumbill; and anonymous sources.

19:48:31 <Davey> Davey is now known as D[a]vey

20:03:33 <edd> foafbot, picture of nori?

20:03:34 <foafbot> Can't identify 'nori', sorry.

20:03:38 <edd> oh, shame

20:03:44 <edd> foafbot, picture of danbri?

20:03:45 <foafbot> Picture at http://photos.heddley.com/albums/www2002/hawaii_017.sized.jpg -- according to Edd Dumbill

20:03:45 <foafbot> Caption: Dan Brickley

20:03:45 <foafbot> Picture at http://photos.heddley.com/albums/www2002/hawaii_028.sized.jpg -- according to Edd Dumbill

20:03:45 <foafbot> Caption: Libby Miller and Dan Brickley getting ready before DevDay

20:04:00 <edd> foafbot, picture of danbri and libby

20:04:01 <foafbot> Picture at http://photos.heddley.com/albums/www2002/hawaii_028.sized.jpg -- according to Edd Dumbill

20:04:01 <foafbot> Caption: Libby Miller and Dan Brickley getting ready before DevDay

20:04:49 <danbri> :)

20:05:08 <danbri> foafbot, picture of Masahide Kanzaki

20:05:08 <foafbot> Can't identify 'Masahide Kanzaki', sorry.

20:05:09 <edd> just ensuring stuff works for the demo, danbri :)

20:05:14 <edd> it's only on a partial crawl

20:05:21 * danbri not actually demoing, just slides

20:05:22 <edd> so just the Obvious Suspects have their data in

20:05:23 <danbri> chicken

20:05:40 <danbri> if i had ruby on this laptop, might've shown some crawler demos, but stuff scrolling up screen is kinda dull

20:05:54 <edd> i'd have loved to have demo'd dashboard

20:05:55 <danbri> s/ruby/ruby and all the xml gunk etc working/

20:06:04 <edd> but obviously not worken on powerpc

20:06:12 <danbri> obviously.

20:07:37 * edd goes to speaker ready room to see if laptop might just work with the projector

20:28:38 <D[a]vey> Hmm, can anyone give me an example of foaf:Agent being used (i.e. for a bot/dog/plant)

20:28:50 <D[a]vey> I'm still bothered about it :/

20:30:20 <mortenf> it's simply a replacement for foaf:Person

20:30:27 <mortenf> (in the RDF/XML version)

20:31:12 <D[a]vey> ok, I thought so!

20:31:29 <D[a]vey> mortenf: thank you for being so clear about that :)

20:31:33 <mortenf> np :)

20:32:41 <mortenf> of course, there's more to it than that, since a foaf:Person is also a foaf:Agent, and some properties imply foaf:Person (depending on the range/domain) - but you can probably ignore those issues for now

20:34:14 <danbri> I am not aware of any properties which apply only to non-Person foaf:Agents, though that is conceivable

20:34:34 <danbri> eg. dogs:pedigree might apply to a Dog, which is a non-person Agent.

20:34:42 <danbri> maybe some will emerge in time...

20:34:49 <mortenf> true, i was refering to the ones that have a range or domain of foaf:Person

20:36:43 <danbri> yup

21:19:41 <D[a]vey> Hmm, <foaf:fundedBy rdf:resource="http://example.org/funder" /> is correct?

21:20:00 <mortenf> could be

21:20:28 <D[a]vey> http://xmlns.com/foaf/0.1/#term_fundedBy is the spec for it, but I'm afraid I simply don't understand the domain/range stuff :/

21:20:41 <D[a]vey> well, I think I *do*

21:20:50 <mortenf> well, as the prose says, it's underspecified

21:21:06 <mortenf> in other word, you can have anything as the subject and object

21:21:16 <D[a]vey> Domain is the "parent", the Range is the values it can have, in this case, its values are defined by the rdf:resource values.

21:21:40 <D[a]vey> mortenf: hmm. I wonder if I should leave foaf:fundedBy out

21:21:44 <mortenf> yeah, kind of

21:21:49 <D[a]vey> libby: Welcome back :)

21:21:50 <mortenf> that's probably wise

21:21:59 <libby> heya

21:22:05 * libby just off down pub

21:22:28 <D[a]vey> enjoy libby ;)

21:22:53 <libby> I will. bye!

21:23:37 <mortenf> so when do we get to see and nitpick at your tool, Davey? ;)

21:23:51 <D[a]vey> mortenf: well, you can do that... now. The basics work :)

21:24:11 <mortenf> ah, where?

21:24:46 <D[a]vey> well, you want a form front end? or are you happy to take the PHP and use it? :)

21:24:56 <mortenf> php is fine

21:25:52 <D[a]vey> I'm documenting the source right now :)

21:26:09 <mortenf> ok, will check it out

21:28:22 * mortenf wonders if such classes-as-objects source could be auto-generated

21:29:07 <danbri> hmm is shellac's xquery foaf crawler public?`

21:29:16 <mortenf> dunno

21:30:40 <D[a]vey> btw, the comments at the bottom show which foaf:* elements are in each Agents domain

21:31:08 <D[a]vey> but if you look at isAllowedForAgent() you'll see which are common, and which are only for a specific agent.

21:32:41 <mortenf> yep, saw that.

21:33:27 <mortenf> have you thought of doing a single "function set($property,$value)" instead (combined with a list of properties and their range, resource or literal)?

21:34:16 <mortenf> btw, typo in setFundedBy (uses foaf:depiction)

21:36:13 <mortenf> the serialization of foaf:knows/foaf:Person is wrong, did you read the striping doc?

21:37:05 * D[a]vey will check

21:38:14 <mortenf> seems you should just drop the "isset" test in "knows", and always perform the "else" part

21:39:05 <mortenf> oh, and the range of foaf:mbox is a resource, not a literal (unlike mbox_sha1sum)

21:39:08 <D[a]vey> oops, yes, I see that :)

21:39:18 * D[a]vey edits

21:40:35 <D[a]vey> both fixed :)

21:41:35 <mortenf> ah, much nicer :)

21:43:50 <D[a]vey> :)

21:44:27 <mortenf> does the comment "Although we use another instance of XML_FOAF, we could re-use" really hold?

21:44:36 <D[a]vey> yes

21:44:45 <mortenf> i mean, the code uses addChild

21:45:08 <mortenf> so if you set the name twice, you end up with two names, right?

21:45:18 <D[a]vey> pardon?

21:46:16 <mortenf> if you change from $libby to $matt, you'll get a different result

21:46:22 <D[a]vey> mortenf: from my mimimal testing, any changes to the object doesn't get changed in the XML once its added using XML_FOAF::knows() - so if you just do $matt->newAgent() you can start over with a new (clean) agent.

21:46:39 <mortenf> ah, ok, but that's not reuse :)

21:46:47 <D[a]vey> yes it is

21:46:55 <D[a]vey> in that you don't instantiate a new copy of XML_FOAF

21:46:59 <mortenf> no, you create a new object

21:47:08 <D[a]vey> no you don't

21:47:17 <D[a]vey> you use the same object with a new XML tree in it ;)

21:47:21 <mortenf> hmm

21:47:32 <mortenf> ah, ok, yeah, sorry

21:47:58 <D[a]vey> :)

21:49:31 <mortenf> but it looks good, should be quite usable for newcomers

21:49:47 <mortenf> hmm, based_near should work for all agents

21:49:57 <D[a]vey> mortenf: Actually, this is the powerful not-for-newbies class :)

21:50:03 <D[a]vey> D[a]vey is now known as Davey

21:50:33 <mortenf> hmm, ok ;)

21:51:37 <Davey> theres going to be a XML_FOAF_Lite which is basically do the same as foaf-a-matic

21:52:00 * Davey rejoices over the new IE bug

21:53:14 <danbri> http://swordfish.rdfweb.org/discovery/2001/08/codepict/scutterplan.rdf has only 4 links, hmm

21:53:58 <mortenf> hmm, well, /DataSources has more

21:55:13 * danbri saves http://www.w3.org/2001/sw/Europe/talks/xml2003/Overview-3.html

21:55:26 <danbri> need add more of y'all's stuff...

21:58:36 <mortenf> i'd add an "automagically" or so to slide 9 in "identity reasoning: our tools figure out when two files talk about the same thing"

21:59:11 <mortenf> i realise you don't want to enter the specifics

22:23:29 * danbri fends off an xml<->mainframes salesguy looking for IBM contacts

22:23:30 <epeus> epeus is now known as Kevin_Marks

22:23:47 <mortenf> heh

22:24:12 <mortenf> did you see: <mortenf> i'd add an "automagically" or so to slide 9 in "identity reasoning:? our tools figure out when two files talk about the same thing"

22:24:39 <Davey> danbri: Hey... do you have any IBM contacts? *waiting on his to get back to him*

22:24:45 <Davey> (I'm joking btw...)

22:24:48 <danbri> :)

22:24:54 <danbri> go ask guha :)

22:25:38 <Davey> I'd kill to get IBM to sponsor PaWS, but I think its a little... off tilt for them.

22:27:08 <adr> Davey, talked to anyone at the W3C?

22:27:14 <danbri> getting ibm to do anything is a big job. they're large...

22:27:18 <Davey> adr: Sent a mail, no reply :/

22:27:23 <Davey> danbri: yeah, one of the many problems.

22:29:11 <Davey> danbri: The only company that didn't have in-house PR though

22:29:17 <Davey> even Apple PR is in-house

22:31:10 <danbri> do you have sponsors, currently?

22:34:09 <Davey> danbri: not something I want to answer in public, especially where its logged ;)

22:34:49 <danbri> ok

22:35:10 <Davey> If you want to query me on that, feel free though :)

22:59:58 * danbri resaves http://www.w3.org/2001/sw/Europe/talks/xml2003/all.htm

23:01:47 <Davey> something tells me that foaf:logo should be in the domain of foaf:image

23:01:56 <Davey> foaf:Image sorry

23:55:23 <Davey> what on earth is foaf:page?

23:59:30 <danbri> a bit like foaf:homepage, but not so exclusive

23:59:36 <danbri> imagine i wrote a page about you

23:59:51 <danbri> there's a relationship wee call foaf:page that links from you to it;

23:59:59 <danbri> inversely, there's a foaf:topic relation from the page, to you.