Friend of a Friend (FOAF) IRC Chat Logs for 2004-09-03

00:08:28 <crschmidt> redlandbot is mostly quiet, and works a lot like wh4, only he uses ^ for commands. still under development, if he gets noisy, just say ^part

00:08:54 <Xavier> if you pet him, he'll purr

00:09:40 * Talliesin strokes redlandbot behind the ears

00:10:04 <crschmidt> he won't remember to stay parted the next time he restarts, but i'm working on that

00:10:27 <Xavier> perhaps you just need a ^bitchslap for that.

00:10:44 <crschmidt> heh

00:23:19 <dustym> What's redlandbot written in crschmidt?

00:24:02 <crschmidt> Python

00:25:44 <dustym> BTW, I've been using FOAF files gleaned from lj to get my hands dirty with redland, I appreciate the work you have done over there.

00:26:09 <crschmidt> Heh

00:26:18 <crschmidt> It'd be nice if I had some kind of contorl over the files/format now

00:26:25 <crschmidt> There's so many things I would change if I could do it over

00:28:10 <crschmidt> But thanks! :()

00:28:12 <crschmidt> :)*

00:31:01 <dustym> np. What kind of storage is redlandbot using?

00:31:06 <crschmidt> mysql

00:31:18 <Talliesin> crschmidt, remember how you were saying that one of the advantages of LJs FOAF was that it would stop people doing naïve gra

00:31:25 <Talliesin> grab-everything spider?

00:31:30 <crschmidt> Talliesin: yep

00:32:04 <Talliesin> sh1mmer wants to do deliberately naïve grab-everything spidering of it, with a fat pipe and a big ol' lump of storage :)

00:32:26 <crschmidt> yeah, i know, he mentioned it to me

00:32:32 <crschmidt> I think he's in for more than he's bargained for.

00:32:43 <Talliesin> It could make wh4's claim of "I know... everything" closer to being true :)

00:33:23 <crschmidt> Basically, you have to be getting a new file every... 4.3 seconds.

00:33:36 <Talliesin> Heh, he seems to want to take the google approach; do something inherently unscalable and then throw unfeasible amounts of hardware at it :)

00:34:03 <Talliesin> If nothing else it would certainly be an interesting thing to try to do, even if it didn't quite work.

00:34:13 <crschmidt> yep.

01:16:27 <pjz> is there a schema that lets an agent specify what other agents can see?

01:22:17 <pjz> so maybe I don't want to share foaf:mbox with everyone, I should be able to do soemthing like <foaf:Group><foaf:name>All</foaf:name><auth:deny rdf:nodeID="me"><foaf:mbox /></auth:deny></foaf:Group>

01:22:47 <pjz> hrm, but instead of 'me' it would be the URL to my foaf resource

01:23:06 <crschmidt> Right, but how to guarentee that anyone respects that?

01:23:12 <pjz> well, that's the fun but

01:23:17 <pjz> er, fun bit

01:23:38 <pjz> that's what FoafCheck or whatever other schemes we come up with are for

01:23:41 * crschmidt still needs to set the bot to ping something, the nat connection dies after 10 minutes

01:24:06 <pjz> and also, some things are enforcable by law if nothing else

01:24:38 <pjz> oh, hrm, it's really like a robots.txt

01:24:55 <crschmidt> Right, and you can't really trust that.

01:25:11 <pjz> and if there *is* an auth mechanism, it would/could use this internally to store its perms

01:25:19 <pjz> what to share out to whom

01:26:39 * crschmidt nods

01:26:54 <crschmidt> I suppose the real question is "is there a way to store authentication levels or something of the sort in RDF"?

01:27:13 <crschmidt> The authentication still needs to be done seperately by an application delivering the data, i think

01:27:28 <crschmidt> but you might be able to say 'this is viewable to all,' 'friends' 'only me' or something

01:27:41 <pjz> sure

01:27:58 <pjz> so.. 'is ther a way to store authentication levels or something of the sort in RDF?'

01:28:02 <crschmidt> the answer is, "I don't know" :)

01:28:05 <pjz> :P

01:28:38 <pjz> I'm new to this rdf thing

01:28:59 <crschmidt> ah! good. welcome to RDF. It's all triples, triples all the way down, always remember that.

01:29:01 <pjz> the only hits I can find for authorization ar ethe wot thing, which has a different focus

01:29:18 <crschmidt> Yeah, I think that you're talking about something which hasn't really been done yet

01:31:48 <pjz> cuz I've got an idea on a way to extend FOAFCheck to make it reasonable, so the first thing I thought of was using it to let you define different access levels to your foaf info

01:31:54 <crschmidt> which isn't to say that it shouldn't be done or couldn't be done

01:34:21 <pjz> right

01:34:37 <crschmidt> might be a good thing to bring up on the foaf mailing list

01:34:40 <crschmidt> .g rdfweb-dev

01:34:42 <phenny> rdfweb-dev: http://rdfweb.org/mailman/listinfo/rdfweb-dev

01:36:22 <pjz> hrm, yah, I guess I need to subscribe

01:37:10 <pjz> for all I know there's some part of the OWL spec or something that can be leveraged

01:37:17 <pjz> (though I looked)

01:39:54 <crschmidt> Yeah, I've been doing this for a while, and haven't seen anything that i recognize as easily answering your questions.

01:46:02 <pjz> how do you use rdf syntax to impose ordering?

01:46:06 <pjz> or can you?

01:46:40 <crschmidt> ordreing of?

01:46:59 <pjz> well, I was thinking about this allow/deny stuff

01:47:06 <pjz> I need.. hrm, dont' need ordering, really

01:47:11 <pjz> just need a way to say 'everything'

02:14:31 Disconnected from irc.freenode.net (Connection reset by peer)

02:15:09 Topic now http://rdfweb.org/topic/FoafGalway -- -- Leverage your relationship assets with World Wide FOAF - http://www.foaf-project.org/ - http://rdfweb.org/ - [set your IRC client charset to Unicode/UTF-8 please!] - chat logs http://www.ilrt.bris.ac.uk/discovery/chatlogs/foaf/ (may be offline till 2004-09-01) Summer of FOAF events (do come! tell all your friends!): http://rdfweb.org/mt/foaflog/archives/2004/07/05/11.49.19/

02:15:09 Users on #foaf: loggerf eaon KjetilK edd eggzip antont shaunm kota redlandbot pjz eikeon crschmidt balbinus_ bitsko Xavier dustym MacIntire Netminder ronwalf chrisc joeldg WSBot TrustBot imajes gritzko foafgwy dngor kasei wh4experimental lion CaptSolo Arnia ericP bskahan mattmcc Jibbler Jipp Grantbow Mutiny YaroslavVB phenny Treenaks sh1mmer workbench iwaim lilo azz BenjaFallenstein SlowByte_ optika

02:15:10 <NickServ> This nickname is owned by someone else

02:15:10 <NickServ> If this is your nickname, type /msg NickServ IDENTIFY <password>

02:15:10 <NickServ> Password accepted - you are now recognized

02:15:10 <MemoServ> You have no new memos

02:32:12 <crschmidt> crschmidt is now known as julie

02:32:25 <julie> julie is now known as crschmidt

02:35:31 <crschmidt> everyone, meet julie. She's like redlandbot, with personality.

02:37:27 * CLoCkWeRX loox @ I-Neighbors

03:11:39 <pjz> hrm, I don't think I'm getting this quite right.

03:12:40 <pjz> is <a><b>c</b></a> the same semantically as <a b="c"/> ?

03:36:13 <pjz> ?

03:50:26 <jsled> pjz: in RDF/XML?

03:51:07 <jsled> yes, in some cases.

03:55:15 <pjz> 'some cases' ?

03:55:45 <pjz> so, like, is <foaf:Group foaf:name="All"> valid?

04:02:12 <jsled> hmm. probably. lemme find the serialization rule.

04:02:45 <jsled> http://www.w3.org/TR/rdf-syntax-grammar/#section-Syntax-property-attributes

04:02:51 <jsled> Specifically...

04:03:02 <jsled> "When a property element's content is string literal, it may be possible to use it as an XML attribute on the containing node element."

04:03:27 <jsled> So, I'm assuming foaf:name is a string literal, and thus: yes, that construction is valid.

04:03:47 <pjz> got it.

04:03:56 <pjz> (and thanks for the pointer to where it says that, by the way)

04:04:11 <jsled> [of course, there should probably be a rdf:about or rdf:ID in the <foaf:Group ...> fragment...]

04:07:18 <pjz> right

04:08:11 <pjz> okay, another question: is it valid to put a ns on the right side of an = ?

04:08:26 <pjz> or does that break some kind of rule?

04:08:34 <jsled> ex?

04:10:22 <pjz> <deny rdf:resource=foaf:name>

04:10:41 <pjz> to deny an agent access to my name

04:10:49 <jsled> well, all xml attributes must be quoted... so that's one. ;)

04:10:51 <pjz> oh, rhmr, that should be a real resources, not a namespace. nevermind.

04:11:10 <jsled> yeah ... it's not so much that it breaks a rule as it's not allowed anywhere.

04:11:23 <jsled> Most people "compress" that part with XML entities...

04:11:29 <pjz> ?

04:11:38 <jsled> <deny rdf:resource="&foaf;name" />

04:11:50 <jsled> where &foaf; is defined earlier in the XML.

04:12:08 <jsled> But tooling support for entity decls is weak, so I wouldn't recommend it.

04:12:17 <pjz> ahh... exactly what I was looking for, I think.

04:12:19 <pjz> oh, hrm.

04:13:13 <pjz> I'm trying to write a syntax to allow me to declare what groups of agents I authorize to look at some of my subresources (ie.

04:13:28 <pjz> (ie. "&me;mbox")

04:13:42 <jsled> &me; = ?

04:14:03 <pjz> &me; = "http://myserver/~myhomedir/foaf.rdf"

04:14:55 <jsled> hmm. literally, "&me;mbox" => "http://myserver/~myhomedir/foaf.rdfmbox"

04:15:00 <jsled> which might not be what you meant.

04:15:13 <pjz> yeah, you're right

04:15:45 <pjz> what I mean is 'the foaf:mbox property of the person described by that foaf.rdf file"

04:15:58 <jsled> Oh ... I see the backscroll, now.

04:16:08 * jsled reads

04:18:22 <jsled> Hmm. yeah ... enforcement is the real problem.

04:18:53 <jsled> But you could certainly come up with another vocabulary which talks about rdf Properties.

04:19:08 <jsled> But I don't know what the semantics of "deny" are ... what's being denied?

04:19:18 <pjz> access to my mbox

04:19:35 <jsled> define "access".

04:20:00 <pjz> so my agent that serves out the foaf.rdf to you won't include my mbox property if you've been denied access to it

04:20:23 <jsled> and when you say "mbox", do you mean the mailbox itself, or the rdf:Object of a { ?person foaf:mbox ?mbox. } triple?

04:20:35 <pjz> I mean the triple

04:20:39 <jsled> I see.

04:21:57 <pjz> it's a resource too, right? there's some way to point to ?person:foaf:mbox ?

04:21:58 <jsled> So, you're really defining a set of rules for a transformation of a base set of RDF triples, given some retreival function and user credentials.

04:22:30 <jsled> to point to the statement itself?

04:22:46 <pjz> yes

04:22:59 <jsled> kind-of.

04:23:03 <pjz> ?

04:23:27 <jsled> there is something called "reification", which is generally the process of describing things within themself.

04:24:06 <jsled> But a) reifying [sp] a statement doesn't assert the statement. and b) doesn't automagically give it a resource.

04:24:31 <jsled> so, not really.

04:30:58 <jsled> but ... you basically want to say ... if 'user' requests the data, then in/ex-clude the following properties...

04:31:11 <pjz> mostly

04:31:18 <jsled> possibly, "... the following properties based on some pattern of their content"

04:31:31 <pjz> more than likely I want to say if user-from-group ...

04:31:53 <pjz> ie. define the restriction as part of a foaf:Group definition

04:32:58 <jsled> Sure ... so something like ... [ a :Rule; :userInGroup <eg:someGroup>; :include ( foaf:mbox ); :exclude ( foaf:phone ) ].

04:34:18 <pjz> right

04:35:09 <jsled> you might be able to do something fancy with N3 rules, but I'm not quite sure. I actually intuit not, but couldn't defend it.

04:40:48 <pjz> I need to read the namespace spec I think

04:43:44 <pjz> I don't understand how foaf:Person maps to http://xmlns.com/foaf/0.1/Person"

04:44:20 <jsled> hmm. somewhere, somehow, the namespace name "foaf" is mapped to the URI "http://xmlns.com/foaf/0.1/"

04:44:28 <pjz> right

04:44:33 <pjz> that part I get

04:44:39 <jsled> Thus, "foaf:Person" is literally the string-concatenation of the URI with the "Person" part.

04:44:53 <jsled> In RDF/XML this is done with XML namespaces.

04:45:11 <pjz> but there's no <Person> tag in the rdf pointed to by the namespace

04:45:12 <jsled> So, somewhere there will be the string xmlns:foaf="http://xmlns.com/foaf/0.1/"

04:45:17 <pjz> right

04:45:44 <pjz> but I'm looking for what creates valid :foos inside one of those

04:45:55 <jsled> Anything that exists when you dereference the URI is helpful, but not required.

04:46:32 <jsled> [FTR, I believe that valid RDF does come back if you GET http://xmlns.com/foaf/0.1/, though ... including a definition of the rdfs class foaf:Person...]

04:46:52 <pjz> ah, okay, so that's not required tho

04:46:53 <jsled> The "creation of valid :foos" ... well ... there's two takes on that.

04:47:01 <jsled> One is: if there's agreement, then it's valid. :)

04:47:13 <jsled> The other is: if it's formally or informally declared then it's valid.

04:47:47 <jsled> That is ... if some RDFS [or OWL or DAML+OIL or ...] states that eg:SomeClassName has certain semantics, then it does.

04:48:06 <pjz> right

04:48:31 <jsled> No, it's not required, but it's Really Good Practice. :)

04:48:37 <pjz> sure

04:50:05 <pjz> I was looking for the right set of semantics to define allow and deny

04:50:26 <jsled> I don't believe they exist as generically as you want them. I haven't seen them as such.

04:50:37 <pjz> hrm

04:50:44 <pjz> they don't seem that complex

04:50:45 <jsled> you can check the usual places [google, schemaweb.info, ...]

04:59:56 <jsled> anyways, gotta run; take care pjz.

05:00:14 <pjz> thanks for your help

08:12:15 <JimH_> JimH_ is now known as JimH

09:55:08 * ldodds discovers he's missed out on a several months worth of rdfweb-dev

09:55:17 <ldodds> I thought it had gone oddly quite

09:55:21 <ldodds> quiet even

09:57:27 <Cloud_> Morning

09:57:46 <ldodds> morning

09:58:35 <Cloud_> A filter gone bad ldodds?

09:58:53 <ldodds> I think I must have been unsubscribed

09:59:12 <ldodds> I've just sent two mails to the list, so maybe I'm in a blacklist somewhere

10:02:03 <__> __ is now known as Cloud__

10:25:55 <[GNU]> morning all

10:26:12 <ldodds> morning

10:27:08 <danbri> hiya

10:27:12 * danbri back in brizzle now

10:27:18 <Cloud__> Hey Dan

10:27:22 <danbri> woke up: 5.40am

10:27:23 <Cloud__> That was good time

10:27:30 <danbri> closed eyes for a second: 5.41am

10:27:41 <danbri> woken up by simon phone call: 6.26am

10:27:52 <danbri> arrived airport: 6.45am

10:27:59 <Cloud__> eek!

10:28:02 <danbri> board plane: 6.56am

10:28:08 <danja> hooo

10:28:10 <danbri> scheduled take off: 7am

10:28:28 <danbri> it was delayed a little anyway, but that's was pretty tight?

10:28:37 <Cloud__> Woh, close one - cheers to Simon for the wakeup

10:28:46 <danbri> and what lesson do I learn from all this? :)

10:28:55 <danbri> "don't worry I'll catch it somehow"

10:28:58 <Cloud__> Don't go to sleep?

10:29:03 * ldodds grins

10:29:06 <Cloud__> Hehehe

10:29:17 <danbri> That's my usual method, but I'm flying again on Sunday and had little sleep this week already...

10:29:33 <ldodds> only sleep on planes?

10:29:47 <danbri> was a great conferencey-workshop-thing! thanks again Cloud for all your efforts :) :)

10:30:18 <danja> any write-ups planned?

10:30:23 <danbri> yes

10:30:27 <danbri> we need to do a workshop report

10:30:29 <Cloud__> Thanks! It was fun, and I'm glad things worked out pretty well, my anti-rain-dancing went okay too, I forgot a few steps but there you go

10:30:43 <danbri> and i want to write something for a w3c context re possible new activities in foaf-related area

10:30:54 <danbri>http://moblog.nicecupoftea.org/

10:31:01 <danbri> i couldn't moblog all week, sent a few pics on way home

10:31:06 <danja> mmm, that sounds promising

10:31:10 <ldodds> indeed

10:31:18 * Talliesin is currently doing a "Don't worry be crappy" version of a way of documenting archived ontologies.

10:32:37 <danbri> oh cool

10:32:43 <danja> heh, you got chaals in church?

10:32:49 <[GNU]> danbri: received any answers what this is?? http://moblog.nicecupoftea.org/archives/000879.html

10:32:53 <danbri> yyyyyeah....

10:33:08 <danja> Talliesin, cool

10:33:08 <danbri> not yet. OK folks, what is http://moblog.nicecupoftea.org/archives/000879.html ?

10:33:18 <danbri> its pretty obvious in that pic

10:33:21 <danbri> some others are better

10:33:23 * danbri looks

10:33:39 <[GNU]> ahm... the turbo prop not gotten right by your cam?

10:33:57 <danbri> weird strobing effect from airplane propellers

10:33:59 <danbri> old fashioned plane

10:34:10 <danja> aaah

10:34:27 <ldodds> not aliens then

10:34:45 <[GNU]> maybe a giant sail of a oil tanker...

10:34:51 <ldodds> thought you'd trumped the SETI@home people :)

10:35:06 <danja> dolphins?

10:35:16 <ldodds> next-gen ducks?

10:36:40 <danja> heh, yeah, carbon-fibre stealth ducks

10:37:13 <Talliesin> Danbri, are the datestamps you put on the FOAF ontology localtime (BST in summer) or GMT?

10:37:31 <danbri> yes

10:38:19 <danbri> danbri@snowball2:~$ date

10:38:19 <danbri> Fri Sep 3 11:38:09 BST 2004

10:38:28 <danbri> that's from the same server, so however CVS does it...

10:38:47 <danbri> You just mean the CVS $Date$'d bit?

10:39:10 <Talliesin> Yes. I might as well have accurate data in my example.

10:39:50 <Talliesin> And I'd prefer to use an 8601 timestamp (like all good people, grrr CVS) and timezone info is nice.

10:40:07 <danbri> +1

10:40:27 <Talliesin> Did you make that last change at half past three or half past two?

10:40:39 <danbri> yesterday, i forget

10:40:47 <danbri> in coffee break

10:41:08 <danbri> "Namespace Document 2 Sept 2004 - FOAF Galway Edition"

10:41:13 <Talliesin> Heh, I forget when that was too.

10:41:36 <Talliesin> Cloud, when did we drink coffee yesterday. half-past three or half past two?

10:41:51 <Cloud__> 2 I think

10:42:26 <Talliesin> Right.

10:42:32 <danbri> cool, check out the 2nd propeller photo, http://moblog.nicecupoftea.org/

10:42:49 <danja> aw, adding CC license to RSS feed is a depressing amount of effort

10:43:03 <Talliesin> Oh wait. I got muddled about BST. Really it's either half-three or half-four

10:43:35 <Talliesin> So it must have been half-three. So the stamps must be BST.

10:44:03 <danbri> http://moblog.nicecupoftea.org/archives/images/default/2004-09-03_114302_31243_0.jpg :)

10:44:23 <danja> danbri, where are you off to on Sunday?

10:44:30 <danbri> I stowed my business card in luggage eventually ('cos of the rush), but it survived ok

10:44:32 <danbri> tokyo

10:44:37 <Cloud__> Cool :)

10:44:37 <danbri> svgopen

10:44:46 <Cloud__> Haha

10:44:48 <danbri> must...finish... paper...

10:44:56 <Cloud__> Did they stick something on it

10:45:04 <danbri> yes. came off tho.

10:45:35 <ldodds> discovered why I'm not getting mail from rdfweb-dev. IP is blacklisted by SORBS apparently

10:45:45 <danja> ouch

10:45:57 <danbri> uhoh, what's caused that?

10:46:02 <ldodds> your dynamic ip

10:46:05 <danbri> it is a domestic cable account

10:46:06 <danbri> ah yeah

10:46:10 <danbri> feck

10:46:36 <ldodds> our sysadmin says standard advice is to route mail through ISP'smail server (smarthost) or get static ip

10:46:59 <danbri> ok, will look into that

10:47:01 <ldodds> and whole of 82.32.0.0/16 is blacklisted

10:47:20 <ldodds> we've added to exceptions locally, so I can get mail again

10:47:50 <ldodds> as my earlier mails aren't in the archive, I assume I've ended up in the moderation bin

10:50:59 <iand> hi danbri

10:51:15 <iand> thanks for the galway thing

10:51:23 <iand> good fun

10:51:45 <iand> got a question though

10:52:20 <iand> noticed yesterday that foaf:Person rdfs:subClassOf wgs84:SpatialThing

10:52:39 <iand> this sort fo rules out applying foaf:Person to dead people

10:53:45 <iand> which maybe a bit dodgy anyway

10:54:44 <danbri> hmm lemme think about that

10:54:54 <danbri> where's heaven? ;)

10:55:17 <danbri> can you scribble that question in the FAQ or IssueTracker please? in wiki...

10:55:21 <[GNU]> aint heaven a concept of christians?!

10:55:41 <iand> will do

10:57:52 <ldodds> why is using foaf:Person for dead people dodgy?

10:58:00 <ldodds> (aside from spatial issue)

10:58:30 <iand> just a time related thing. can you say a dead person knows anyone? they certainly _knew_ someone

10:59:01 <Cloud_> And can dead pople know dead people in the afterlife that they didn't know while alive?

11:00:14 <ldodds> I hadn't considered knows, more "this person wrote that, or composed this"

11:03:02 <iand> lots of those types of properties are valid of course. foaf:made works well conceptually

11:04:25 <ldodds> so foaf:knows works fine for living and fictional folk, but not dead folk

11:05:34 <JimH> foaf:knew ? ;-)

11:06:44 <Cloud_> Does that work for amnesia too?

11:06:56 <Cloud_> Sorry, just being silly

11:07:03 <danbri> I forget

11:07:12 <Cloud_> lol

11:07:34 <iand> we have no temporal aspect to RDF so it seems to me that merging graphs asserting facts from different periods of time is on shaky ground

11:07:53 <iand> of course we do that every day and hope for the best :)

11:08:55 <iand> foaf:made works well because it's true for all time, i.e. it has a start time but no end

11:09:05 <iand> foaf:knows has a duration

11:09:22 <[GNU]> why has it a duration?

11:09:39 <[GNU]> i mean, what is the difference of knowing or makeing a thing?

11:10:03 <iand> it's a grammatical tense thing

11:10:16 <iand> made is something that happened in the past

11:10:18 <danja_> grr, dead connection

11:10:21 <iand> knows is something happening now

11:10:30 <[GNU]> but if i start knowing a persone... why should it end?

11:10:35 <danja_> <danja> this reminds me

11:10:35 <danja_> <danja> I've been meaning (for ages) to sort out something for the Burke KnowledgeWeb people

11:10:35 <danja_> <danja> to RDFify their data on historical figures

11:10:35 <danja_> <danja> remember James Burke's connections?

11:10:35 <danja_> <danja> mostly dead people

11:10:42 <iand> if they die, do you still know them?

11:10:49 <[GNU]> ja

11:10:55 <[GNU]> but only cant talk to him

11:11:13 <danja_>http://www.k-web.org/

11:11:58 <danja_> seems a reasonable use case for keeping foaf:knows post mortem

11:12:11 <iand> yes. interesting.

11:12:26 <iand> actually I'm not suggesting that knows doesn't apply to dead people

11:12:28 <danja_> Napolean knows Josephine etc

11:12:50 <iand> just that it might be dodgy to write it :)

11:12:52 <[GNU]> i mean.. if you take it a level more philosophical... what you "know" of a person is the idea of the person

11:13:00 <[GNU]> and that idea is immortal

11:13:04 <ldodds> understood. the angle hadn't occured to me

11:13:50 <iand> but the SpatialThing subclass does seem to limit usage for fictional or dead people

11:14:07 <iand> and come the singularity we'll all be in trouble...

11:14:41 <ldodds> makes for a simple graph :)

11:14:53 <danja_> .

11:14:58 <iand> singularity - the great smush!

11:15:05 * ldodds LOL

11:16:35 <danja_> I guess the Big Bang was the original spring layout algorithm

11:17:13 <iand> [GNU] what about if you've yet to know someone?

11:17:20 * danbri checks definition of SpatialThing

11:17:30 <iand> my foaf today doesn't decribe people I don't yet know

11:17:57 <iand> but in a year it will be different

11:18:12 <iand> so the facts at a point in time are different

11:18:16 <danbri> <rdfs:Class rdf:about="http://www.w3.org/2003/01/geo/wgs84_pos#SpatialThing">

11:18:16 <danbri> <rdfs:label>SpatialThing</rdfs:label>

11:18:16 <danbri> <rdfs:comment>Anything with spatial extent, i.e. size, shape, or position. e.g. people, places, bowling balls, as well as abstract areas like cubes.

11:18:16 <danbri> </rdfs:comment>

11:18:16 <danbri> </rdfs:Class>

11:18:36 <ldodds> is a dead person an abstract area?

11:18:37 <danbri> Anywhere-and-nowhere...

11:18:47 <danbri> s/is/has/ ?

11:19:00 <danja_> sounds a reasonable kludge

11:19:23 <[GNU]> iand: is that "i know someone in the future" ??

11:20:07 <iand> for a historical figure do you write foaf describing all the people they ever knew?

11:20:09 <danbri> their area could be the entire universe

11:20:12 <iand> see also http://www.cyc.com/cycdoc/vocab/spatial-vocab.html

11:20:15 * danja_ recognises lyrics to 'Hi Ho Silver Lining'

11:20:32 <iand> First, a #$SpatialThing may be #$PartiallyTangible (e.g. #$Texas-State) or wholly #$Intangible (e.g. #$ArcticCircle ...

11:21:12 <[GNU]> ahm... iand: i think so... we sould describe whom the hist. person knew

11:21:30 <[GNU]> iand: maybe pepsys (?) diary is a good thing to learn from

11:21:41 <crschmidt> danbri: there's a slight problem with the vs: terms in the foaf schema:

11:21:43 <crschmidt> ^query select ?term where (?term vs:term_status "stable") using vs for <http://www.w3.org/2003/06/sw-vocab-status/ns#>

11:21:43 <[GNU]> there is this guy topic mapping the diary

11:21:44 <julie> [http://xmlns.com/foaf/0.1/homepage] ,

11:21:50 <crschmidt> only one is stable :)

11:22:17 <[GNU]> not even name ?

11:22:30 <danbri> yeah, we never defined stability

11:22:44 <danbri> hence the FoafCommunityProcess rewrite, to get some standards in place for that

11:22:50 <iand> we ought to propose a definition and then a candidate list on rdfweb-dev

11:22:53 <danbri> no point in saying thy're stable till we have some associated definitions

11:23:04 <danbri> its more like a draft of some lighhtweight process issues

11:23:28 <iand> can they be defined by mean time between changes?

11:23:33 * ldodds last two postings were process related

11:23:53 <iand> stable - no changes expected over next 2 years

11:23:56 <crschmidt> ^query select ?label where (?term rdfs:label ?label) (?term vs:term_status "unstable") using vs for <http://www.w3.org/2003/06/sw-vocab-status/ns#>, rdfs for <http://www.w3.org/2000/01/rdf-schema#>

11:23:58 <julie> reading , Project , Online Gaming Account , based near , read , Organization , Online Account , toRead , theme , sha1sum (hex) , publications , account service homepage , Online E-commerce Account , membershipClass , isbn , favourite , holds account , account name , Group , Agent , funded by , DNA checksum , Online Chat Account , member ,

11:24:12 <crschmidt> hm.

11:24:28 <crschmidt> apparently someone has some non foaf things in there with vs:term_status, heh

11:24:32 <[GNU]> iand: stable: not changed by need but only by process

11:24:48 <danbri> cool

11:25:12 <crschmidt> also, if anyone hasn't seen it yet, http://rdfweb.org/topic/IrcMeeting

11:25:20 <iand> i think implementors need some idea of how hard it will be to change something stable

11:26:01 <iand> i wouldn't want to see changes to stable terms fast-tracked through

11:26:34 <[GNU]> so... maybe we jus tneed to look up the definition of "change management" in some software developers handbook?!

11:26:50 <iand> :)

11:27:06 <iand> an oft-talked about subject, rarely implemented :)

11:27:16 <[GNU]> :/

11:27:26 <ldodds> crschmidt my reading schema has vs terms in there

11:27:33 <crschmidt> ldodds: ah, that would be it

11:27:34 <ldodds> thats probably what you're picking up

11:27:55 <crschmidt> not upset or anything, the bot has an open interface to add, i just didn't realize it :)

11:28:13 <ldodds> whats it got in it?

11:28:54 <crschmidt> ^smush

11:28:55 <julie> Smushing data. Currently 10705 triples.

11:28:59 <julie> Data smushed: 10705 statements to 10705 statements.

11:29:00 <crschmidt> lots :)

11:29:03 <crschmidt> It's like wh4

11:29:09 <ldodds> oh ok

11:29:11 <crschmidt> only redland based

11:29:25 <crschmidt>http://crschmidt.net/rdfpython/python.cgi?all=1

11:30:15 <crschmidt> ^info

11:30:15 <julie> I'm a Redland/Python based RDF query bot. Source in SVN at <http://crschmidt.net/svn/rdfpython/trunk/>. Commands are ^add <url>, which adds statements, ^newcommand <name> is <RDQL Query>, ^runcommand <commandname> <arguments> . ^commandlist lists current commands, ^commandlist <command> shows info on a command, ^part will have me part a channel, ^join <#chan> will have me join a channel. Talk to crschmidt for more. @@ More here

11:30:38 <crschmidt> hi ina

11:30:47 <Cloud_> Morning Ina

11:31:03 <danbri> 'morning' Ina :)

11:31:12 <Ina> Morning John

11:31:14 <Ina> Hi Dan :-)

11:31:38 <danbri> Ina, I didn't believe you when you said I could get a taxi at 6am for a 7am flight

11:31:57 <Ina> and...

11:32:01 <danbri> Then I woke up at 6.26am and it was ALL TRUE. I'll never doubt you again!

11:32:12 <Ina> :-)

11:32:24 <danbri> was a mite close, i'll admit...

11:32:27 * iand notes SpatialThingIssue as resolved on IssueTracker

11:32:34 <danbri> thanks Ian!

11:32:43 <Ina> who is Ian?

11:32:50 <crschmidt> ian davis

11:32:56 <crschmidt> i think?

11:32:56 <danbri> = iand

11:33:08 <iand> yes

11:33:16 <crschmidt> oh, heh, i though you said iand. *tired, needs to go to work*

11:33:20 <crschmidt> i'll bbl

11:33:23 * danja_ thinks julie a great name for a bot

11:41:16 <Talliesin>http://www.hackcraft.net/ontHistory/

11:53:02 <Talliesin> Insane approach?

12:01:54 <[GNU]> ahm... aint a new version kind of an inheritance of the old one?!

12:02:13 <[GNU]> i mean... couldnt we subclass all classes in the new version?

12:03:10 <[GNU]> erhm... need to get food to regulate blood sugar level... :)

12:03:37 <Talliesin> Well, the decision to keep the same URI for later versions wasn't made my me.

12:04:34 <Talliesin> In theory I agree with you. In practice I agree with danbri, relying on entailment to make new versions work with old apps doesn't really work that well (with current apps anyway).

12:06:23 <[GNU]> i have something in mind like a base version, and new "features" could be kind of aspects of that base version

12:06:34 <[GNU]> or maybe a version negotiation stuff is needed?!

12:07:02 <[GNU]> this way you alway have a fallback base version all apps can work with

12:07:04 <iand> i think the namespace uri persistence issue has come from the XML world rather than the RDF world

12:07:16 <iand> RDF has the mechanisms to do the inheritance and equivilece

12:07:34 <iand> but XML doesn't so changing the NS URI in XML is a _big_ deal

12:07:40 <[GNU]> ah, ok, gotcha

12:08:03 <iand> danbri mentioned pain experienced by dublin core when changed from 1.0 to 1.1

12:08:22 <iand> i'm not convinced that it necessarily translates to RDF vocabularies

12:08:49 <Talliesin> Well in terms of what you describe about a base version and fallback, FOAF *is* the base version.

12:09:01 <[GNU]> right

12:09:16 <Talliesin> What we want to deal with is the way it keeps changing.

12:09:44 <[GNU]> from a technical point of view

12:09:44 <ldodds> loggerf: pointer

12:09:44 <ldodds> See http://www.ilrt.bris.ac.uk/discovery/chatlogs/foaf/2004-09-03#T12-09-44-1

12:10:21 <Talliesin> iand, some of the pain mentioned was just with Dublin Core as used in RDF.

12:10:59 <danbri> yeah, eg. see Annotea codebase migration woes as went from dc 1.0 to dc 1.1

12:11:04 <iand> but before OWL

12:11:30 <iand> was Annotea codebase RDFS/OWL aware?

12:11:45 <Talliesin> Also, do we want to rule out apps which do no entailment, or only poor-man's entailment (as in having the knowledge of which FOAF properties are IFPs and can be smushed on hard-coded in).

12:12:01 <iand> that's the issue

12:13:06 <Talliesin> Because if we want an app to cope with http://xmlns.com/foaf/0.2/ then it has to cope with entailment, OWL discovery, authenticating the OWL (so we can't play silly buggers by getting it to parse an incorrect OWL doc) and maybe other issues.

12:13:08 <iand> Think Jon's document makes a good point though: " Storing the old version in the state it was in in a publicly accessible location would mean that triples from the old ontology could (almost certainly would) be mixed in with triples from the new one."

12:14:01 <Talliesin> Ironically though, that would only be a problem for the more powerful processors that could deal with the difficulties I just mentioned with changing namespace names.

12:14:13 <danbri> if the ontology was crawled, you mean?

12:14:19 <Talliesin> Yes.

12:14:20 <danbri> or you mean triples that _use_ the old ontology?

12:14:29 <danbri> OK. I've a cheapass answer: robots.txt

12:14:38 * danbri not convinced himself, tho ;)

12:14:50 <danbri> a context-happy store shouldn't mix things that naively

12:15:23 <danbri> I think archiving index.rdf as index.20040431.rdf and using absolute URIs to the living namespace within that is a plausible thing to explore

12:15:37 <Talliesin> No. Even if robots.txt is followed the protocol doesn't require a bot that was "told" to go to a particular URI from disobeying robots.txt, so it woudn't rule out problems.

12:16:06 <iand> Often helps to look at the extremes to clarify thinking: in 50 years time will terms have significantly changed meaning? If so then they really ought to have a new URI

12:16:09 <danbri> I think its still safe...

12:16:27 <danbri> most terms key of natural language

12:16:32 <danbri> ...and those don't change quite that fast

12:16:46 <Talliesin> Some do, and it's hard to predict which.

12:16:54 <danbri> Agent

12:16:56 <Talliesin> seeAlso foaf:schoolHomepage

12:17:03 <Talliesin> ;)

12:17:14 <danbri> yeah :)

12:17:51 <iand> danger is changing meanings over long periods. Historical data subtly becomes different.

12:18:03 <Talliesin> Yes, we need nice descriptions.

12:18:17 <Talliesin> But with the meaning "nice" had 200 years ago ;)

12:19:06 <Talliesin> .w nice

12:19:09 <phenny> nice 1. a city in southeastern France on the Mediterranean; the leading resort on the French Riviera

12:19:13 <phenny> nice 2. pleasant or pleasing or agreeable in nature or appearance; "what a nice fellow you are and we all thought you so nasty"- George Meredith; "nice manners"; "a nice dress"; "a nice face"; "a nice day"; "had a nice time at the party"; "the corn and tomatoes are nice today"

12:19:17 <phenny> nice 3. socially or conventionally correct; refined or virtuous; "from a decent family"; "a nice girl" [...]

12:19:26 <Talliesin> Not any of those meanings.

12:19:34 <iand> foaf:weblog - will it mean anything in 10 years, let alone 50?

12:19:50 <Talliesin> If it does, I doubt it will be the same as now.

12:20:05 <danbri> it'll be a relationship between a person and a document, stil

12:20:12 <Talliesin> The definition of "weblog" is disputed territory.

12:20:36 <danbri> yup

12:20:48 <iand> we should cede definition to Father of Weblogs surely

12:21:11 <Talliesin> "Web sites built using Userland technology"?

12:24:05 <CaptSolo> hi all

12:24:36 * CaptSolo is back at work, direct from the galway

12:25:12 * Talliesin is to, though he seems to be writing up ideas about FOAF instead :)

12:26:23 <Talliesin> Oh, for the sake of anyone reading the logs who doesn't know; "nice" used to mean "admirably precise" around the 17th/18th century.

12:27:28 <CaptSolo> tall: galway was inspiring in terms of foaf and not only that. so - it's no wonder the mind wants to keep working on that

12:28:16 <Talliesin> CaptSolo, yes... once the blood-alcohol level's dropped sufficiently.

12:28:36 <CaptSolo> tall: you got it pubped up pretty efficiently :)

12:28:49 <CaptSolo> hi inkel!

12:29:04 <inkel> hi CaptSolo, how are you?

12:30:09 <Talliesin> Yeah, and chaals can't hold his drink!

12:31:09 <Talliesin> (Or possibly just looked less capable of doing so in the morning because I didn't touch the bottle of rum).

12:31:56 <Talliesin> I have a strange craving for the kind of nasty burgers you get on trains.

12:32:50 <danbri> fight it with all your strength!

12:33:03 <CaptSolo> inkel: fine. yourself?

12:33:10 <danbri> hey inkel

12:33:25 <CaptSolo> hey danbri

12:33:35 <danbri> you're home now, captsolo? or teleworking?

12:33:35 <inkel> I'm fine, thanks

12:33:36 <CaptSolo> Tall: i wonder how is chaals now

12:33:41 <inkel> hi danbri!

12:33:57 <CaptSolo> danbri: at work already - straight from plane+bus

12:34:15 <CaptSolo> i did drop into irc yesterday from dublin airport for a moment

12:34:31 * crschmidt saw him then!

12:34:34 <CaptSolo> danbri: where are you? partying in galway?

12:34:41 <danbri> I'm home already

12:34:44 <CaptSolo> chris - hi :)

12:35:08 <CaptSolo> danbri: and what are you up to? something new inspired by the workshop?

12:35:16 <CaptSolo> how did you like it overall?

12:35:28 <danbri> I thought it went great

12:35:58 <CaptSolo> i have the same opinion

12:36:00 <danbri> I feel bad not to have met more people, but there were 100+ and was trying juggle too many things, but i'm sure there'll be other such events...

12:36:06 <crschmidt> danbri: Do you have thoughts or opinions on the IrcMetting wiki-page?

12:36:34 <CaptSolo> was good to be there. matching faces to names / nicknames is also wonderful. suddenly there is more personality behind the texts and ideas

12:36:36 <danbri> I had a brief look, generally seems sensible. couple comments: check out the http://esw.w3.org/topic/ScheduledTopicChats (or similar name) page for #rdfig

12:36:43 <danbri> ...aslo i can't make nxt week, but that's fine

12:36:50 <crschmidt> To be honest, I'm trying to figure out if weekly or biweekly is better

12:37:02 <danbri> Ah maybe you missed the discussion yesterday

12:37:12 <danbri> I proposed monthly, starting Sept 15th

12:37:15 <crschmidt> ahh

12:37:18 <danbri> see FoafCommunityProcess

12:37:18 <crschmidt> Yeah, I missed that :/

12:37:35 <danbri> but people of course are very welcome to have other scheduled topic chats, in #rdfig or #foaf

12:37:39 * crschmidt thinks monthly is too long

12:37:40 <crschmidt> yeah

12:37:44 <danbri> the monthly proposal was specficially re the spec editing cycle

12:37:48 * crschmidt nods

12:37:52 <danbri> It depends on people's schedules

12:37:53 <crschmidt> that sounds good I think

12:37:58 <danbri> I'm going to be on the road all of October

12:38:01 <inkel> http://esw.w3.org/topic/ScheduledTopicChat is the URL

12:38:02 <danbri> so we need to be pragmatic

12:38:02 * crschmidt nods

12:38:07 <danbri> could have two in one week....

12:38:30 <crschmidt> yeah, the only thing i caught from scribing of irc meeting stuff was discussing a time

12:39:51 <crschmidt> I'm wondering if weekly topicchats for general topics + monthly spec chats sounds good to anyone else?

12:40:18 <[GNU]> i like that :)

12:40:31 <crschmidt> I'm building mostly off of what I picked up from drumm, re deanspace, which seemed to be that 1. biweekly confused people, they could never remember when it was, 2. weekly let people communicate and not lose track of what they were supposed to be doing amongst piles of other projects

12:40:49 <crschmidt> (this is what i felt like i learned by participating, not a list of any kind that he told me or anything like that.

12:40:53 <[GNU]> but aint we got rdfical for that?

12:41:08 <danbri> I think topic chats could well end up being better as an #rdfig thing

12:41:13 <danbri> can have plenty foaf stuff of course...

12:41:24 <danbri> but trust, storage, query, calendar, etc etc all a bit more general

12:41:29 * crschmidt nods

12:41:30 <crschmidt> true enough

12:42:49 * crschmidt sighs, forgot my keys, no one at work yet despite it being 30 minutes past time when people usually get here, standing in hallway on laptop

12:46:22 <crschmidt> My only concerns regarding using #rdfig: 1. it's got a lot of people in there who may not be FOAF-oriented, taking over the channel might be bad manners (see http://esw.w3.org/topic/ScheduledTopicChat, "Finding Upcoming" section)

12:46:56 <crschmidt> 2. Some poeple are in #foaf that I'd like to hear from who might not be #rdfig-ers // not willing to wander over

12:47:01 <CaptSolo> chris: rdfig is quite silent - some action should not harm it

12:47:24 <CaptSolo> and as danbri points out - the topics go beyond foaf and into generic topics as well

12:47:43 <crschmidt> yeah

12:47:58 * crschmidt is scared of #rdfig, too many RDF people there ;)

12:48:00 <CaptSolo> chris: re 2 - a simple announcement on topic change on #foaf would help - "come to rdfig for a scheduled chat"

12:48:12 <crschmidt> yeah, i think so

12:49:18 <crschmidt> Okay, I'm going to reply to ldodds in a bit (once i get sitting down in the office, darn coworkers and forgetting my keys)

12:50:56 * CaptSolo launched compilation on linux and aix and can rest a bit waiting for results

12:51:08 <CaptSolo> (that is work, not foaf thing)

12:51:12 <crschmidt> heh

12:51:33 <CaptSolo> takes some time

12:53:06 <danbri> naptime! cu...

12:53:27 <inkel> bye danbri

12:55:24 <CaptSolo> cu

12:55:33 * crschmidt gets inside

12:58:01 <danbri> ooh, #rdfig 75 members, #foaf 65 ...

12:58:04 <danbri> we're catching up!

12:58:11 <danbri> same face in large part, but still...

12:58:14 <danbri> ok really offline now

12:58:55 <crschmidt> heh

12:59:01 <crschmidt> okay, so those of you still around

12:59:11 <crschmidt> First IrcMeeting this week, or next week?

12:59:22 <crschmidt> (this week meaning 5 days, next week being 12)

13:02:15 <[GNU]> topic less one?

13:03:50 <crschmidt> No, this would be what I was talking about at http://rdfweb.org/wiki/IrcMeeting, so a guided topic discussion

13:04:29 <[GNU]> 404

13:04:41 <crschmidt> er

13:04:51 <[GNU]> the wiki is gone :)

13:04:52 <crschmidt>http://rdfweb.org/topic/IrcMeeting

13:05:01 <crschmidt> no, it'ss just at /topic/ , not /wiki/ :)

13:05:04 <[GNU]> a wiki producing a 404 would be senseless

13:05:22 <crschmidt> heh

13:05:35 <crschmidt> "that term odesn't exist, you can't create it, it's banned" :)

13:07:55 <[GNU]> man, its 15:00 local time, and i am completly unable to work or read websites

13:08:07 <[GNU]> were is the fscking edit button?!

13:08:23 <crschmidt> You have to login first

13:08:31 <crschmidt> click userpreferences in the corner

13:08:34 <crschmidt> (upper right)

13:09:40 <[GNU]> ahh :) ok

13:10:06 <[GNU]> wow, now edit really edits

13:15:52 <rawles> Does anyone have a FOAF dataset which is reasonably large and has a lot of interlinking? I am doing some mining but the current data I have is not really up to the job of supporting the learning enough at this stage.

13:16:02 <MacIntire> hehehe

13:16:18 <crschmidt> rawles: how big is reasonably large?

13:16:38 <crschmidt> And what kind of data are you looking for? lots of livejournal, lots of stuff with full contact details, lots of...

13:16:41 <MacIntire> LJ is perfect....though woefully limiting in what it supports foaf-wise

13:16:58 <rawles> well i am only debugging right now so i'm working on only 40 foaf files. maybe 1000 foaf files would be nice

13:17:03 <rawles> yes, livejournal, of course.

13:17:15 <rawles> it doesnt matter that it doesnt use the full foaf spec, in fact that's probably better.

13:17:17 <crschmidt> it's actually a relatively good use of the FOAF core, but not much beyond that

13:17:31 <crschmidt> it doesn't use rel, it doesn't use geo, it doesn't use bio - but those aren't foaf :)

13:17:31 <MacIntire> cool

13:17:40 <rawles> that is much better

13:17:42 <MacIntire> you know me, crschmidt.....I want it all

13:17:47 <rawles> i dont use geo, rel, etc.

13:17:56 <rawles> it's beyond the scope of wha ti want to do

13:18:16 <rawles> i just need a bunch of files which are from some kind of close knit group of otherwise different people

13:18:33 <crschmidt> rawles: do you want it for graphing or something like that?

13:18:39 <crschmidt> (or if you tell me do you have to kill me?)

13:18:50 <rawles> no its for some experiments in data mining

13:19:00 <rawles> people always go "oooh evil" when i say that

13:19:07 <rawles> but i'm a good miner

13:19:20 <rawles> as opposed to evil anyway

13:19:23 <crschmidt> rawles: to be honest, i'dd say to just graba relatively well linked file (grab http://jessical.livejournal.com/data/foaf ) and write a quick scutter

13:19:32 <rawles> i have a scutter written

13:19:36 <crschmidt> ah

13:19:39 <crschmidt> start at that link

13:19:40 <rawles> i could just scut within the lj domain

13:19:52 <rawles> is the load an issue? i guess not

13:19:58 * crschmidt thinks, jess's file will link to... hm.

13:20:03 <crschmidt> on lj? not really

13:20:06 <bru> hallo!

13:20:08 <rawles> ok

13:20:13 <crschmidt> with 700 dynamic page loads/second you'll be a drop in the bucket

13:20:20 <rawles> ok just asking :-)

13:20:23 <rawles> this sounds like a good idea

13:20:39 <crschmidt> But yeah, that'll get you a good data for graphing

13:20:41 <rawles> you have communities of people on there dont you?

13:20:49 <[GNU]> hmm..

13:20:52 <crschmidt> yeah, but they aren't linked from peoples foaf files

13:21:11 <crschmidt> we create a foaf:Group for comms, but we don't have an easy enough way of saying "member of"

13:21:24 <[GNU]> rawles: maybe you'll have a look at the rdf file of theyrule??

13:21:31 <rawles> what would be really good is to have all foaf files from community A, and all from a different community B, and i could mine how to differ them?

13:21:42 <rawles> [GNU]: I think my friend Simon Price has been looking at that.

13:21:52 <[GNU]> ok

13:21:53 * crschmidt thinks

13:22:46 <rawles> or some otherwise highly interlinked cluster of lj users

13:22:57 <rawles> well, 2 clusters

13:22:58 <crschmidt> rawles: http://www.livejournal.com/users/blinkies/data/foaf

13:23:11 <crschmidt> use that, and then use http://crschmidt.livejournal.com/data/foaf

13:23:37 <rawles> okay, you mean scut from each one?

13:23:41 <crschmidt> yep

13:23:49 <crschmidt> that'ss probably as distinct of a population as you can get on livejournal

13:24:06 <rawles> who is blinkies?

13:24:32 <crschmidt> it'ss a community of people who make animated graphics to be used as livejournal icons

13:24:36 <rawles> ok

13:24:54 <rawles> do you have control of the lj database?

13:25:00 <crschmidt> Nope

13:25:06 <crschmidt> I'm just a guy on LJ

13:25:18 <rawles> okay so you couldn't for example, easily give me a list of all foaf files from two communities of users?

13:25:22 <rawles> cos that would be best of all

13:25:24 <crschmidt> don't work there, don't have any relations to anyone that works there

13:25:47 <crschmidt> i just gave you two files, oen with 100 seeAlsos, one with 78

13:25:56 <crschmidt> scutter from each, and you should get a relatively distinct set

13:26:01 <rawles> okay thanks buddy

13:26:04 <crschmidt> np

13:26:14 <rawles> sorry if i was cheeky just then

13:26:18 <crschmidt> oh, no problem

13:26:30 <[GNU]> .w cheeky

13:26:32 <phenny> cheeky 1. offensively bold; "a brash newcomer disputed the age-old rules for admission to the club"; "a nervy thing to say"

13:26:37 <crschmidt> you have to do some work on your own though :)

13:26:50 <rawles> yeah, it wont be hard to scut that data

13:26:56 <rawles> i wrote some foaf scutter code in prolog

13:27:01 <rawles> it's proper bo.

13:27:22 * crschmidt nods

13:27:29 <crschmidt> just make sure you markk which URLs you've grabbed.

13:27:41 <crschmidt> you'll see a *lot* of interconnection

13:28:35 <crschmidt> and grabbing the same files every time is a very nice way to get your spider banned

13:28:41 <crschmidt> oh, you might want to read http://www.livejournal.com/bots/

13:28:51 <rawles> we cache everything, use robots.txt and have delays between fetches.

13:29:00 <crschmidt> it suggests putting URL/mailto: in User-Agent

13:29:03 <crschmidt> so you're recognized as a bot

13:29:12 <rawles> umm we havent registered the bot with LJ

13:29:14 <rawles> oh i see

13:29:18 <rawles> all the same sentence

13:29:49 <rawles> we have a url there as well

13:30:01 <rawles> we spent about a week coding in all the "be a nice bot" stuff

13:30:05 * crschmidt nods

13:30:07 <crschmidt> keedokee

13:30:17 <crschmidt> just tossing in the little lj-specific knowledge that'ss good to know :)

13:30:26 <rawles> yeah, i dont want to get banned

13:30:30 <crschmidt> Even though I've been asked to not spend any more time developing on livejournal *sigh*

13:30:33 <rawles> btw, my friend is using lj and loves it.

13:30:40 <[GNU]> anyone has an idea how to use this one: http://www.cyc.com/cycdoc/vocab/time-vocab.html#Event

13:30:44 <rawles> she is chinese and using it to get her english better.

13:30:51 <crschmidt> rawles: good deal

13:31:17 <rawles> so what is your day job crschmidt

13:31:42 <crschmidt> Developer at http://www.wedu.com

13:32:39 <rawles> i love the graphic design very nice

13:33:05 <crschmidt> We're in the process of redesinging our site to be completely flash based *sigh*

13:33:07 <[GNU]> seems to me like a how-to-beeing-not-cheeky

13:33:42 <rawles> i dont understand [GNU]

13:34:08 <[GNU]> ja, no problem its my confused mind

13:34:49 <rawles> i was being cheeky cos i basically asked crschmidt for a database dump which would require him to do work

13:34:59 <crschmidt> heh

13:35:12 <crschmidt> the biggest thing is that lj can't really do a "database dump" in any easy way

13:35:18 <crschmidt> with 4 million files...

13:35:29 <rawles> being cheeky is asking something beyond your relationship with another person

13:36:09 <[GNU]> ah... thats not getting clear to me from the wn definition

13:36:23 * crschmidt curses at english :)

13:36:26 <crschmidt> .mw cheeky

13:36:29 <phenny> cheeky: definition not found

13:36:34 <rawles> hehe

13:36:37 * crschmidt curses at phenny

13:36:44 <rawles> sounds like uk english

13:36:56 <rawles> where are you from [GNU]

13:37:08 <[GNU]> bloody german

13:37:14 <rawles> OED: Characterized by .cheek.; insolent or audacious in address; coolly impudent or presuming.

13:37:25 <rawles> I think my german teacher called me "frisch" when i was being cheeky to her.

13:37:59 <rawles> aha.. "frech", "unverschaemt"

13:38:07 <[GNU]> ok :) that makes sense

13:38:32 <[GNU]> alles frisch is more like "hey silly"

13:39:47 <rawles> where are you in germany?

13:39:54 <rawles> i have been there a few times

13:40:02 <[GNU]> the so called "Bundesstadt Bonn"

13:40:23 <rawles> Are you at Uni Bonn?

13:40:43 <[GNU]> no, have been there... in... 94?!

13:40:49 <[GNU]> or 95-97

13:41:09 <[GNU]> have you been there?

13:41:15 <[GNU]> at the uni?

13:41:35 <rawles> I have friends who work at Schloss Berlinhoven in Sankt-Augustin

13:41:43 <[GNU]> ah, that is GMD

13:41:43 <rawles> And have visited S-A a number of times.

13:41:48 <[GNU]> or Frauenhofer today

13:42:32 <rawles> yeah

13:42:37 <rawles> and they have links to uni-bonn

13:42:53 <rawles> i have visited bonn and koeln a number of times

13:43:12 <[GNU]> ja, they have close relation

13:43:22 <[GNU]> drop a line if you come here again

13:43:46 <[GNU]> cs at bristol?!

13:43:56 <[GNU]> thats why you use prolog :)

13:44:02 <rawles> yes

13:44:05 <rawles> you know peter?

13:44:11 <rawles> yes i may come later in the year

13:44:22 <[GNU]> i know arround 5 peters

13:44:23 <rawles> are you near Sankt-Augustin, the town?

13:44:31 <[GNU]> ne, bonn central

13:44:45 <rawles> ok

13:44:49 <[GNU]> but its about 20min by train

13:44:53 <rawles> yes i'll let you know if i am in the area again

13:45:09 <[GNU]> so you can teach me prolog :)

13:45:23 <rawles> you're a sadist!

13:45:39 <[GNU]> oh... happily not a masochist :)

13:45:41 <rawles> okay speaking of the prolog, i'd better get back to it

13:45:54 <[GNU]> have fun!?! :)

13:45:56 <rawles> if you want help or chat, i'm on rawles@cs.bris.ac.uk

13:46:06 <[GNU]> that is email?

13:46:39 <[GNU]> we should put foaf file url in the realname field of irc

13:46:59 <crschmidt> You can always dump a foaf file with nick into wh4/julie

13:47:06 <crschmidt> then do:

13:47:09 <crschmidt> ^aim crschmidt

13:47:10 <julie> cr5chmidt ,

13:47:16 <crschmidt> ^mbox crschmidt

13:47:16 <julie> [mailto:crschmid@uiuc.edu] ,

13:47:24 <[GNU]> ^jabber crschmidt

13:47:31 <crschmidt> not programmed in

13:47:34 <crschmidt> plus i don't use it :)

13:47:34 <[GNU]> :)

13:47:52 <[GNU]> how to feed my url in julie... which is not of any sexual background...

13:48:03 <crschmidt> type ^add url

13:48:17 <rawles> ^add url http://www.cs.bris.ac.uk/home/rawles/foaf/foaf.rdf

13:48:17 <crschmidt> ^addcommand select ?jabber where (?p foaf:nick "%s") (?p foaf:jabberID ?jabber) using foaf for <http://xmlns.com/foaf/0.1/>

13:48:17 <[GNU]> ^add http://B4mad.Net/FOAF/goern.rdf

13:48:18 <julie> Adding that to my database...

13:48:18 <julie> Adding that to my database...

13:48:22 <julie> Added 171 statements from http://B4mad.Net/FOAF/goern.rdf. Model size is 10876.

13:48:44 <[GNU]> ^jabberID [GNU]

13:49:00 <[GNU]> argh... not programmed :)

13:49:02 <crschmidt> ^addcommand jabber is select ?jabber where (?p foaf:nick "%s") (?p foaf:jabberID ?jabber) using foaf for <http://xmlns.com/foaf/0.1/>

13:49:08 <crschmidt> ^newcommand jabber is select ?jabber where (?p foaf:nick "%s") (?p foaf:jabberID ?jabber) using foaf for <http://xmlns.com/foaf/0.1/>

13:49:09 <julie> New command jabber is select ?jabber where (?p foaf:nick "%s") (?p foaf:jabberID ?jabber) using foaf for <http://xmlns.com/foaf/0.1/>

13:49:15 * crschmidt can't remember commands for his own bot

13:49:19 <[GNU]> :)

13:49:25 <[GNU]> ^jabber [GNU]

13:49:25 <julie> goern@jabber.org ,

13:49:26 <crschmidt> rawles: can you try your url again?

13:49:36 <[GNU]> that bot is way to cool for earth

13:49:38 <crschmidt> for some reason it didn't seem to go in :/

13:49:52 <crschmidt> oh wait

13:49:59 <crschmidt> don't type "url" in there

13:50:09 <crschmidt> just ^add http://foo

13:50:09 <[GNU]> ^add http://www.cs.bris.ac.uk/home/rawles/foaf/foaf.rdf

13:50:10 <julie> Adding that to my database...

13:50:18 <julie> Added 103 statements from http://www.cs.bris.ac.uk/home/rawles/foaf/foaf.rdf. Model size is 10979.

13:50:23 <crschmidt> ^smush

13:50:23 <julie> Smushing data. Currently 10979 triples.

13:50:27 <julie> Data smushed: 10979 statements to 10979 statements.

13:50:31 <crschmidt> heh

13:50:38 <crschmidt> apparently you don't know anyone already in the graph :)

13:50:53 <rawles> no that would make sense

13:50:57 <rawles> it's only a little file

13:51:03 <[GNU]> apparently i am very unshure about whom i know or not

13:51:07 <rawles> also i dont know how to qualify who i know

13:51:13 <crschmidt> rel! :)

13:51:22 <crschmidt> .g relationship rdf schema

13:51:22 <rawles> hehe of course

13:51:48 <[GNU]> ^add http://b4mad.net/datenbrei/index.php?author=2/foaf.rdf

13:51:49 <julie> Adding that to my database...

13:51:51 <julie> Added 178 statements from http://b4mad.net/datenbrei/index.php?author=2/foaf.rdf. Model size is 11157.

13:51:53 <phenny> relationship rdf schema: http://www.w3.org/TR/NOTE-rdf-uml/

13:51:57 <crschmidt> hm

13:52:00 <[GNU]> ^smush

13:52:01 <crschmidt> not what i was looking for

13:52:01 <julie> Smushing data. Currently 11157 triples.

13:52:05 <julie> Data smushed: 11157 statements to 11150 statements.

13:52:11 <crschmidt> woot :)

13:52:18 <crschmidt> ^xfn_met [GNU]

13:52:18 <julie> Till Wille , Daniel Riek , Stephan Dühr , Marcel Hild , Nicolai Schleifer , Stefan Zosel ,

13:52:33 <[GNU]> hmm.. utf-8 problem?!

13:52:41 <[GNU]> or is that my irc client?!

13:52:44 <crschmidt> not sure

13:52:45 <rawles> ^mbox rawles

13:52:46 <julie> Query select ?mbox where (?p foaf:mbox ?mbox) (?p foaf:nick "rawles") using foaf for <http://xmlns.com/foaf/0.1/> returned no results

13:52:54 <rawles> hehe

13:53:03 <[GNU]> ^mbox [gnu]

13:53:04 <julie> Query select ?mbox where (?p foaf:mbox ?mbox) (?p foaf:nick "[gnu]") using foaf for <http://xmlns.com/foaf/0.1/> returned no results

13:53:07 <rawles> i probably summed it

13:53:15 <[GNU]> of couse :)

13:53:15 <crschmidt> .pc interrobang

13:53:19 <phenny> 0069: LATIN SMALL LETTER I (i)

13:53:22 <phenny> 006E: LATIN SMALL LETTER N (n)

13:53:25 <phenny> 0074: LATIN SMALL LETTER T (t)

13:53:25 <crschmidt> er, dang it

13:53:31 <crschmidt> .cp interrobang

13:53:33 <phenny> 203D: INTERROBANG (‽)

13:53:57 <[GNU]> doesnt look good to me

13:54:09 <crschmidt> yeah, it's delivered in utf-8

13:54:17 <crschmidt> Stephen looks fine to me

13:54:22 <crschmidt> so your client is b0rk :)

13:54:30 <[GNU]> ah... iso gaim

13:54:37 <[GNU]> doofes ding

13:54:49 <crschmidt> [GNU]: searches are case sensitive

13:54:55 <crschmidt> so you have to do ^mbox [GNU]

13:55:04 <[GNU]> so... its been 3 hrs since my plan of having lunch...

13:55:07 <[GNU]> ^mbox [GNU]

13:55:08 <julie> Query select ?mbox where (?p foaf:mbox ?mbox) (?p foaf:nick "[GNU]") using foaf for <http://xmlns.com/foaf/0.1/> returned no results

13:55:08 <crschmidt> (I'll figure out how to make that not true one of these days)

13:55:16 <crschmidt> ^sha [GNU]

13:55:17 <julie> 0a9d600329b781dcdd710bf97c8ad16c44cf8dc9 ,

13:55:19 <crschmidt> yep

13:55:23 <[GNU]> gotcha

13:55:35 <[GNU]> maybe i love julie

13:55:46 <crschmidt> "I think I love you, so what are you so afraid of"

13:55:47 <[GNU]> julie: like to go out to lunch wiht me?

13:56:15 <[GNU]> autsch... got a appmt for lunch at 1600 local

13:58:41 <rawles> does anyone use bitchx?

13:58:49 <crschmidt> I have in the past, use irssi now

14:00:09 <[GNU]> crschmidt: osx and irssi??

14:00:45 <crschmidt> [GNU]: Well, you can get it, you can use colloquy (quite nice on os x) or like me, you can ssh to your linux box and use irssi there

14:00:49 <crschmidt> ;)

14:01:19 <[GNU]> haha :)

14:01:24 <rawles> any idea how to switch bitchx into utf-8?

14:01:39 <[GNU]> google for a bitchx.conf ?

14:02:07 <[GNU]> off to lunch

14:02:41 <crschmidt> not a clue

14:02:44 <rawles> okay

14:02:59 <crschmidt> use irssi :)

14:03:01 <rawles> let me try something, i'll log out and then connect again and maybe someone could type me some utf-8

14:05:24 <rawles> hmm

14:05:28 <rawles> looks the same...

14:06:33 <rawles> hit me with some utf-8

14:06:37 <crschmidt> .cp interrobang

14:06:40 <phenny> 203D: INTERROBANG (‽)

14:06:57 <rawles> okay it just came up with a square, so i reckon that's working, just not in my font

14:07:02 <crschmidt> yep

14:07:06 <rawles> .cp LAMBDA

14:07:08 <phenny> 019B: LATIN SMALL LETTER LAMBDA WITH STROKE (ƛ)

14:07:11 <phenny> 039B: GREEK CAPITAL LETTER LAMDA (Λ)

14:07:14 <crschmidt> .cp diaresis

14:07:14 <phenny> 03BB: GREEK SMALL LETTER LAMDA (λ) [...]

14:07:17 <phenny> Sorry, no results found for 'diaresis'.

14:07:19 <rawles> yup, works okay

14:07:21 <crschmidt> .cp dia

14:07:24 <phenny> 00A8: DIAERESIS (¨)

14:07:27 <phenny> 00C4: LATIN CAPITAL LETTER A WITH DIAERESIS (Ä)

14:07:30 <phenny> 00CB: LATIN CAPITAL LETTER E WITH DIAERESIS (Ë) [...]

14:07:34 <rawles> yup, lots of diaeresises

14:07:39 <crschmidt> cool

14:10:12 <rawles> .cp thai

14:10:14 <phenny> 0E01: THAI CHARACTER KO KAI (ก)

14:10:17 <phenny> 0E02: THAI CHARACTER KHO KHAI (ข)

14:10:20 <phenny> 0E03: THAI CHARACTER KHO KHUAT (ฃ) [...]

14:10:24 <rawles> ooo

14:10:52 <rawles> .cp arabic

14:10:54 <phenny> 0600: ARABIC NUMBER SIGN (؀)

14:10:58 <phenny> 0601: ARABIC SIGN SANAH (؁)

14:11:00 <phenny> 0602: ARABIC FOOTNOTE MARKER (؂) [...]

14:11:15 <crschmidt> you can also do that in PM

14:11:24 <crschmidt> so that you don't mess with everyone else's display, and don't flood the channel :)

14:14:06 <rawles> sorry hehe

14:14:08 <rawles> i got carried away

14:34:15 * dajobe waves

14:34:19 <crschmidt> morning dajobe

14:34:26 <crschmidt> or afternoon, as the case may be

14:34:53 <dajobe> 4 foafers at dublin airport, enjoying the free wifi

14:35:10 <crschmidt> very cool

14:35:23 <crschmidt> i wish airport here had free wifi, costs 7/hr 15/day

14:35:50 <dajobe> swh just got boarding call

14:38:26 * crschmidt is still upset he didn't get to go to anything

14:42:06 <davidoc> dajobe: it's free now? cool!

14:42:26 <dajobe> in celebration of Ireland's presidency of the EU, it says

14:42:44 <crschmidt> ooh, that's nice

14:42:45 <davidoc> Anyone in Dublin this evening?

14:43:14 <Talliesin> That's especially cool given that we haven't held the presidency for a couple of months now.

14:43:17 * Talliesin is.

14:43:20 <dajobe> indeed

14:44:07 <Talliesin> Indeed I'm just opposite Dublin Castle where most of the EU shindigs and meetings were held. This meant that this street was sealed off and I was one of the few non-EU people allowed on it.

14:44:25 <Talliesin> Which was great for keeping the number of people shooting up in our carpark down.

14:44:31 <davidoc> Talliesin: ah, so you're a local!

14:44:40 <dajobe> !

14:44:53 <Talliesin> I am.

14:44:57 <Talliesin> You here ATM?

14:45:07 <davidoc> I'm working in TCD

14:45:17 <Talliesin> Cool.

14:45:32 <dajobe> I was amused watching people try to get on the trams

14:45:43 <Talliesin> We aren't used to them yet.

14:45:46 <dajobe> and about 2 people nearly killed by ignoring buses and trams

14:46:17 <Talliesin> Ah well, we don't really have a concept of "jay-walking" as something you shouldn't do, just a concept of "get out of my f***ing way".

14:46:55 <Talliesin> Though it is acknowledged that being hit by a car has a best-case scenario of making you look really stupid, and hence it is a Bad Thing.

14:47:12 <Talliesin> ^EU Sealed-off street.

14:47:42 <Talliesin> One week though when the street was sealed of for 4 out of 5 working days on the other day one of our cars got ripped off.

14:47:51 * Talliesin rolls his eyes.

15:15:19 <crschmidt> for the ScheduledTopicChats that happened in the past ( http://esw.w3.org/topic/ScheduledTopicChat ) does anyone know how often they usually go for?

15:37:28 <ronwalf> Anyone have feedback on the foaf:Group stuff I mentioned on rdfweb-dev?

15:37:33 <pjz> is there a schema for the whole set of friendster/orkut stuff? ie. hometown, do you smoke/drink, own pets, what sports, what activities, what books you like/are currenty reading ... ?

15:38:44 <ronwalf> specifically, that there is no need for foaf:membershipClass

15:39:05 <iand> ronwalf: I found it very interesting. just haven't had a chance to think coherently about it

15:39:57 <crschmidt> pjz: the menow schema (sorely underdocumented, I'm working on that, i promise) covers what you are reading/listeningto/mood, etc. as for the rest, there isn't a lot, although hometown can be described with something i think - perhaps bio?

15:40:34 <crschmidt> hm

15:40:36 <crschmidt> perhaps not

15:40:51 <pjz> I was just looking at orkut

15:41:05 <pjz> and thinking of writing a little CGI with approx. the same interface

15:41:14 <pjz> that would generate a foaf.rdf

15:41:30 <crschmidt> well, orkut is going to be generating its own foaf "real soon now"

15:41:37 <crschmidt> so you may want to hold off and see what that does first

15:41:39 <pjz> right, but that defeats the point

15:41:46 <pjz> b/c I want control over it

15:41:46 <crschmidt> what's the point?

15:41:56 <pjz> ie. I don't want it stored in a central place.

15:42:32 <pjz> which is to me the point of foaf in the first place - decentralized social networking

15:42:40 <crschmidt> would this proposed CGI script be run by you?

15:42:49 <pjz> yah, it would live in my webspace

15:42:59 <crschmidt> Then isn't it still centralized? :)

15:43:05 <pjz> it only holds my info

15:43:16 <pjz> consider MT vs. LJ

15:43:30 <crschmidt> Right, but the fact that LJ offers a FOAF file doesn't mean you only have to use it on LJ

15:43:41 <crschmidt> http://crschmidt.net/foaf.rdf is based almost entirely off LJ FOAF files

15:43:53 <crschmidt> however, it's not on LJ, it's on my machine

15:44:11 <pjz> right, but right now there's no standalone software

15:44:15 <pjz> so I'm writing it

15:44:38 * crschmidt nods

15:44:41 <pjz> I'd like to give it some of the scutter functionality that, say, orkut displays

15:44:57 <crschmidt> i was just trying to save you some effort, since I know orkut is working on foaf

15:45:03 <pjz> ie. 'there's X many people in your network'

15:45:14 <pjz> .nod

15:45:17 * pjz nods.

15:47:51 <Talliesin> More post-Galway thoughts: http://www.hackcraft.net/FOAFNonIFP/

15:51:18 <pjz> Talliesin: hrm, interesting, it does seem to be a problem to figur out how to tell if two things are the same if they have the value of a property in common

15:51:38 <crschmidt> " This ontology defines 6 properties." <- which?

15:52:35 <pjz> Talliesin: but even for chatIDs (esp. jabberIDs) I could envision ways/places where there might be a 1->many mapping (consider a 'support' ID)

15:52:51 <crschmidt> Talliesin: and I completely agree with you. The foaf:homepage that LiveJournal offers is crap for smushing, because lots of people put (for example) http://www.livejournal.com/ (as advised by some documentation)

15:53:03 <Talliesin> !

15:53:08 <Talliesin> Why would documentation advise that?

15:53:14 <crschmidt> Because of the way the system works, some customization changes require a homepage.

15:53:37 <Talliesin> Oh yes, that reminds me.

15:53:45 <pjz> what about being self-referential? make up/use foaf:foafinfo

15:54:02 <crschmidt> (Basically, it's a hack where the %%webpage%% variable is replaced by some alternate customization, but it requires you to have a webpage defined)

15:56:04 * crschmidt wonders what Talliesin was reminded of.

15:56:20 <Talliesin> loggy, pointer

15:56:26 <crschmidt> loggerf: pointer

15:56:26 <crschmidt> See http://www.ilrt.bris.ac.uk/discovery/chatlogs/foaf/2004-09-03#T15-56-26

15:57:45 * Talliesin makes a slight edit.

15:58:10 <crschmidt> hey, that's my name!

15:59:24 <Talliesin> :)

15:59:56 <crschmidt> LiveJournal data sucks so much

16:00:30 <crschmidt> in the dump jibberjim did a while ago, there's 12 people who list their jabberID as "What?"

16:00:37 <Talliesin> crschmidt, as a reviewer of version 1.0 you are mentioned in version 2.0.

16:00:52 * Talliesin does "release early release often" at breakneck speed.

16:01:00 <crschmidt> oh! the page itself has the schema

16:01:12 <crschmidt> that's cool!

16:01:47 <Talliesin> Yep, schema with XML to produce the human-readable bit, mainly from the dc:description, and with source in the human-readable bit.

16:01:56 * crschmidt nods

16:01:58 <Talliesin> I love the circularity of the approach.

16:02:13 <Talliesin> s/XML/XSL

16:02:16 * pjz mumbles something about Literate Programming.

16:02:24 <crschmidt> hm.

16:02:40 <crschmidt> My only complaint is that sometimes you can't know if there is more than one resource that shares a foaf:homepage or not

16:02:53 <Talliesin> Firefox isn't liking it :)

16:02:55 <Talliesin> :(

16:02:59 <crschmidt> works here

16:03:03 <crschmidt> is it delivered as rdf/xml?

16:03:15 <Talliesin> No, as text/xml.

16:03:20 <crschmidt> hm.

16:03:25 <crschmidt> is it just not styling it?

16:03:28 <crschmidt> Or does it not work at all?

16:03:30 <Talliesin> But in firefox it's treating it as HTML with nonsense.

16:03:36 <crschmidt> hm.

16:03:55 <Talliesin> See, if the XSLT works you also get summaries of each of the properties.

16:03:59 <crschmidt> looks fine here, in Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.6) Gecko/20040206 Firefox/0.8

16:04:18 <crschmidt> oh

16:04:20 <crschmidt> yeah, i don't get that

16:05:17 <crschmidt> anyway, my main complaint is that you say that by definition, if you use the not-IFP properties, there is neccesarily more than one person who uses that resource as, for example, a homepage

16:05:46 <crschmidt> but in large data sources, that doesn't work: ljuser crschmidt has a homepage which *is* an IFP, while ljuser joeblow uses lj.com

16:06:08 <pjz> 'homepage' is ill defined

16:06:43 <pjz> mbox_sha1sum is better, but again could still be shared

16:06:59 <crschmidt> FOAF allows a thing to have multiple homepages, but constrains foaf:homepage so that there can be only one thing that has any particular homepage.

16:07:24 <crschmidt> the definition of when you should use <http://xmlns.com/foaf/0.1/homepage> to describe something is well defined

16:07:29 <Talliesin> crschmidt, I have homepage which may or may not be for only one person (you don't know) shared_homepage that you know is shared (the not-IFP rather than merely non-IFP) and of course foaf provides the IFP version.

16:07:43 <crschmidt> Talliesin: ah, right right

16:07:57 * crschmidt read it and didn't understand it, got it now

16:08:16 <crschmidt> pjz: A lot of what IFPs are about is limiting the data you use as an input

16:08:48 <crschmidt> pjz: if two things can have a matching mbox_sha1sum, you shouldn't publish them as being attached to two different agents

16:08:55 <crschmidt> but rather to a foaf:Group with two members, for example

16:09:48 <crschmidt> I'm gonna run across the street and grab me a sandwich. Be back in 10 or so.

16:10:04 <Talliesin> D'oh I'm a moron.

16:10:17 <balbinus_> balbinus_ is now known as balbinus

16:10:18 <davidoc> Talliesin: mind if I play with your XSLT stuff?

16:10:27 * leighd ponders shared_mbox

16:10:35 <Talliesin> Also FireFox yet again 0wnz IE, because the XSLT shouldn't have worked. Fixed now.

16:10:39 <Talliesin> davidoc, course not.

16:10:40 <leighd> what about mailing list

16:10:47 <leighd> when is a shared_mbox not a mailing list

16:10:56 <leighd> should that be modelled

16:11:28 <Talliesin> There was someone in the TAG who made the point that he can be considered pretty hard-core 'net, what with him being in the TAG, but that he shares his main mbox with his wife.

16:11:53 <Talliesin> Hmm. Good point.

16:12:19 <danbri> "personal mailbox - A personal mailbox, ie. an Internet mailbox associated with exactly one owner, the first owner of this mailbox."

16:12:22 <danbri> ...doesn't rule out sharing

16:12:44 <Talliesin> davidoc, you'll notice that everything on that site has a Creative Commons license unless explicitly stated otherwise (and mostly if stated otherwise it's Public Domain).

16:14:56 <Talliesin> dan, but aliceandbob@example.net shares the association, not just the use.

16:15:21 <danbri> sure, some mail accounts won't fit, but i tried to word it to be pretty inclusive

16:16:07 <leighd> dan, so you were thinking single owner, but might be delivered to multiple people

16:16:23 <leighd> I don't think I'd picked up on that, I'd assumed single owner, single person at the end of it

16:16:27 <Talliesin> It is very inclusive. As inclusive as you can get and be an IFP. And it being an IFP is very, very useful. I just want to catch the cases where it simply isn't an IFP. (Quite rare with mbox, more common with blog, quite common with homepage).

16:16:47 * danbri nods

16:16:48 * Talliesin is in no way criticising the existing predicates.

16:17:01 <danbri> they're all kinda goofy, but i'm not ashamed of that!

16:17:07 <Talliesin> heh.

16:17:40 <leighd> how is shared_homepage really different to foaf:page?

16:17:46 <Talliesin> Well mine first three here are goofy and non-commital, and the next 3 are even more goofy.

16:18:18 <Talliesin> "Homepage" means something other than "page" in natural language. I'm not entirely sure what.

16:18:41 <Talliesin> But it does seem to have an effect when you put it in a question on a HTML form stuck in front of a user.

16:19:08 * leighd isn't criticising just thinking it over

16:19:17 <Talliesin> Course.

16:19:30 <leighd> it's the "shared" bit in the names of some properties that's nagging me

16:19:36 <leighd> like you're exposing half the semantics

16:19:49 <leighd> you're hinting that there's a foaf:Group or something behind it

16:19:55 <leighd> at least in my mind

16:20:03 <leighd> so I'm wondering how to state that

16:20:04 <Talliesin> Well my first thought were the first three, were you simply don't know they are a good IFP.

16:20:10 <Talliesin> I had thought of this some time back.

16:20:39 <pjz> to me it just hilights that the relation between the subject and the object might be one-to-many or many-to-one

16:20:49 <pjz> ie. not an IFP

16:21:10 <Talliesin> Then at Galway when there was talk of mining it occurred to me that these people were taking calculated risks in treating non-IFP data as IFP and I thought it might be useful to be able to mark cases were you knew for sure it was wrong to do so.

16:21:30 <leighd> it's certainly a good idea

16:21:48 <leighd> although whether I obey that depends on how much "risk" I might want to take with my data

16:22:05 <Talliesin> So you have them split into three groups: IFP, Not IFP (but further data may show that there is only one resource with this value) and Definitely not IFP.

16:22:15 <Talliesin> That's the case with any risk.

16:23:36 * Talliesin should probably make the distinction between the first three and second three clearer in the description.

16:30:30 <leighd> lets say I took a risk and smushed on non-IFP data, and then decided it was a bad idea. Using these terms I'd go back to my unsmushed data. I then have two options: Either it was bad to smush on that non-IFP property full stop (and I could take that out of my local schema) OR I discovered that it was only some triples that were dodgy. So then I rewrite the original data to change some statements to use your new properties, then resmush.

16:30:44 <leighd> aah, but you're also asking the users to more explicitly label the data

16:30:55 <leighd> so getting it correct at source

16:31:16 <Talliesin> Yes.

16:31:23 <crschmidt> LJ data is dodgy, so LJ should change their FOAF output to use non-ifp forms of homepage

16:32:11 <leighd> or just page which achieves the same end

16:32:14 <Talliesin> There's also a psychological thing. "An blog only used by you" a user might go "ah sure, close enough" Whereas 1. A blog, 2. Is this only used by you? would, I imagine, elicit a more accurate response.

16:32:15 <crschmidt> which saves you risk (or at least more accurately describes the risk)

16:32:15 <crschmidt> right

16:33:12 <Talliesin> I'm thinking the non-IFP homepage may be adequately served by foaf:page as you suggest. But I'm not sure.

16:33:49 <pjz> there would need to be an analogue for foaf:weblog, since it's also presented as an IFP though it may not be

16:33:52 <leighd> It reminds me of the 9 different types of NULL (or however many there were).

16:34:01 <crschmidt> pjz: john did that :)

16:34:02 <leighd> it's unknown what assumptions you can make about the data

16:34:17 * pjz . o O ( the 9 million faces of etc )

16:34:28 <Talliesin> Yes. The difference between planb.nicecupoftea and the moblog - both blogs of libby's but she shares one of them.

16:35:12 <pjz> right.

16:36:38 <Talliesin> That's actually the case I see this ontology proving most useful.

16:36:45 <Talliesin> Some of the best blogs are shared.

16:37:31 <pjz> sure

16:37:48 <leighd> but the domain of foaf:weblog is foaf:Agent. And a foaf:Group is a foaf:Agent

16:37:59 <leighd> so can't I already state that it's a shared weblog?

16:38:00 <danbri> home should fix that too

16:38:19 <danbri> FOAF Spec - Galway Edition made domain of homepage more mellow

16:38:26 <danbri> hmm

16:38:28 <danbri> nah seems ok

16:38:34 <danbri> having a weblog is more active than a homepage

16:38:41 <leighd> it's just wrong to say that libby -> foaf:weblog -> moblog.

16:40:07 <pjz> oh, you're saying you should say libby -> foaf:memberOf -> Mobloggers ; Mobloggers -> foaf:weblog -> moblog ?

16:40:16 <leighd> yes

16:40:42 <Talliesin> True.

16:40:47 <leighd> and also that that richer relationship is implicit in the "shared_weblog" property, so why not model it?

16:41:03 <Talliesin> Yes, I agree with that.

16:41:28 <Talliesin> Okay. That's killed the last 3 predicates. Do the first 3 live or die?

16:41:45 <leighd> So to follow Talliesin's example, my UI might say: is this your blog? no, then who else uses it, and collect data about the extra people and describe the group

16:41:51 * leighd suddenly feels guilty

16:41:55 <Talliesin> No.

16:42:12 <Talliesin> I want to solve a problem, not have another document with my name at the bottom.

16:42:29 <Talliesin> See, that's the advantage of being an amateur, it doesn't matter if I don't publish ;)

16:42:41 <Talliesin> You're solution is better. I'm happy.

16:42:58 <Talliesin> loggerf:pointer

16:42:58 <Talliesin> See http://www.ilrt.bris.ac.uk/discovery/chatlogs/foaf/2004-09-03#T16-42-58

16:43:06 * leighd checks time, realises late

16:43:12 <leighd> cya soon

16:45:04 <pjz> hey, can someone explain why its foaf:mbox_sha1sum instead of <foaf:mbox><foaf:sha1sum>920398423084</foaf:sha1sum></foaf:mbox> ?

16:45:25 <Talliesin> Good point.

16:45:29 <danbri> I did start experimenting w/ a sha1 property

16:45:32 <Talliesin> Though there's no harm in it being as it is.

16:45:36 <danbri> but you gotta be very careful what it is a property of

16:45:41 <Talliesin> Oh.

16:45:43 <pjz> danbri: why?

16:45:53 <Talliesin> Yes, <foaf:sha1sum> as an IFP of everything?

16:45:56 <danbri> Consider <Document><sha1>foo</...

16:46:00 <Talliesin> You'll find collisions soon enough.

16:46:15 <danbri> and two different 0-byte documents, with different owner and creationDate properties.

16:46:47 <pjz> define it as the sha1 of the document it's enclosed in.

16:46:49 <Talliesin> And it's awkward to define a property as inversely functional for a given domain only (can it be done?)

16:46:59 <pjz> danbri: then it's enclosed at the wrong level

16:46:59 <danbri> also if you stretch it out into two properties and an intermediate object, you get to deduce tha two mailboxes are the same, not that the owners of the mailboxes are the same.

16:47:15 <pjz> but we know an owner can only have one mailbox

16:47:19 <pjz> so the logic holds

16:47:34 <danbri> it is a selective piece of modelling that looks quirky but works ok with deployment pragmatics, i think.

16:47:56 <Talliesin> yeah, the IFP of the IFP will identify. But I don't like treating general sha1 of everything ever expressed as numers as an IFP.

16:48:04 <danbri> can't type much more now... rsi...

16:48:19 <danbri> but it isn't a sha1 of the 'mailbox' but of an URI of the mailbox

16:48:21 <Talliesin> With just emails it's arguably dicey, but in practice probably okay.

16:48:52 <danbri> mb1 ---uriname--> "mailto:danbri@w3.org" --hash-> "12321"

16:48:54 <Talliesin> More triples! Triples for the sake of triples!

16:49:01 <danbri> mb1 ---uriname--> "mailto:danbri@W3.org" --hash-> "543421"

16:49:05 <danbri> etc

16:49:08 <Talliesin> Let there be triples!

16:49:12 <Talliesin> Yes, it's yucky.

16:49:20 <danbri> and you can't have properties of strings, so you'd need yet another layer of indirection

16:49:28 <pjz> hasing always has to worry about canonicalization

16:49:57 <Talliesin> Yeah, I wonder if we should have rules about case in email addresses being hashed, but that's another question.

16:50:15 * crschmidt discussed that with danbri re:foaf way back when

16:50:26 <pjz> Talliesin: only matters if you want people to compare them

16:50:27 <crschmidt> decided that since it's user-input, you just have to trust the user, i think

16:50:43 * pjz laughs at the idea of trusting the user.

16:50:50 <crschmidt> i know.

16:50:52 <crschmidt> sad, isn't it?

16:51:18 <Talliesin> To compare this question to my shared_homepage. I agreed it wasn't necessary, but my predicate wasn't wrong, just not the nicest way of modeling things. In this case even if increasing the complexity of how mbox_sha is modeled isn't "wrong" it's probably nicer to model it as a property of the person.

16:51:53 <Talliesin> Can *Internet* email addresses be case-sensitive in the username portion?

16:52:10 <Talliesin> Otherwise I'd suggest we move to canoncialising to lower-case.

16:52:34 <Talliesin> <pjz> Talliesin: only matters if you want people to compare them

16:52:45 <Talliesin> I think we do. That's what you do when you smush after all.

16:52:51 <Aredridel> technically, the local part is insensitive. In practice, some small percentage of mailers care.

16:53:03 <Talliesin> Bah.

16:53:18 <Aredridel> I'd say follow the spec and munge to lowercase.

16:53:30 <Talliesin> The mail spec?

16:53:31 <Aredridel> And tell people with broken mailers to go back in their holes.

16:53:33 <Aredridel> Yup.

16:53:51 <crschmidt> Right, but now you've got a huge chunk of data out there that doesn't use that system :/

16:53:57 <Talliesin> Yes. On balance I think I would too. crschmidt put doubt in my mind on this earlier though.

16:54:17 <Talliesin> crschmidt, at this point we'll *always* have a huge chunk of data out there.

16:54:25 <Talliesin> We can still make moving to lower case a SHOULD

16:54:38 <Talliesin> will LJ not even let you change that? :(

16:54:43 <crschmidt> Nah.

16:54:50 <Aredridel> Nope.

16:54:51 <crschmidt> I don't think I"ll ever get a change into LJ again

16:55:03 <crschmidt> someone else might be able to

16:55:05 <pjz> What Aredridel said.

16:55:08 <Aredridel> Mmm.

16:55:14 <Aredridel> But I bet the folks over there would be receptive to changes.

16:55:17 <pjz> wait, no

16:55:24 <crschmidt> Over where?

16:55:24 <crschmidt> LJ?

16:55:27 <pjz> the spec says 'reserved local part requires preservation of case'

16:55:30 <Aredridel> Yeah, LJ.

16:55:37 * crschmidt snickers.

16:55:41 <Aredridel> 822 or 2822?

16:55:46 <pjz> 822

16:55:47 <pjz> oh, hrm.

16:55:49 <Aredridel> Mm.

16:56:01 <crschmidt> It took me four months of constant emails to LJ to get them to even look at my patch.

16:56:07 <Aredridel> Ouch.

16:56:19 <crschmidt> Then another 3 weeks of working on it for about 4 hours a day side by side with a dev to get everything to where it wanted to be.

16:56:36 <crschmidt> I have a 4 line patch I submitted 8 months ago awaiting review since february.

16:56:43 <Aredridel> 2822 omits that particular wording.

16:56:55 <Talliesin> Right. Knowing what the spec says for sure is an important start.

16:56:56 <Aredridel> Oh, man.

16:57:08 <Talliesin> I looked at this before and got contradictory info.

16:58:05 <Aredridel> Damnit. 2822 omits any reference to case in the address.

16:58:28 <pjz> Aredridle: no, it's a quoted-string

16:58:34 <crschmidt> So, LJ is not exactly the most receptive development environment.

17:00:03 <Aredridel> pjz: Yeah -- but I still find no reference to case.

17:00:38 <pjz> Aredridel: section 3.2.5 defines quoted-string

17:01:02 <Aredridel> pjz: Yes -- all case is allowable, but it doesn't say how to compare.

17:01:48 <LeoS> hi crschmidt

17:02:08 <Aredridel> Yeah...

17:02:31 <crschmidt> howdy LeoS

17:02:49 * Aredridel looks at 2821 for help.

17:03:02 <LeoS>http://www.leobard.net/foaf.xml

17:03:28 <LeoS> crschmidt: how to add this to your bot? I have a kiss in it

17:03:34 <Aredridel> 2821 says it's sensitive. §2.4

17:04:00 <eaon> ^info

17:04:01 <julie> I'm a Redland/Python based RDF query bot. Source in SVN at <http://crschmidt.net/svn/rdfpython/trunk/>. Commands are ^add <url>, which adds statements, ^newcommand <name> is <RDQL Query>, ^runcommand <commandname> <arguments> . ^commandlist lists current commands, ^commandlist <command> shows info on a command, ^part will have me part a channel, ^join <#chan> will have me join a channel. Talk to crschmidt for more. @@ More here

17:04:10 <eaon> LeoS: there's your answer :)

17:04:25 <LeoS> ^add http://www.leobard.net/foaf.xml

17:04:25 <julie> Adding that to my database...

17:04:27 <julie> Added 100 statements from http://www.leobard.net/foaf.xml. Model size is 11250.

17:04:32 <crschmidt> ^smush

17:04:33 <julie> Smushing data. Currently 11250 triples.

17:04:37 <julie> Data smushed: 11250 statements to 11240 statements.

17:05:03 <LeoS> does smushing mean that my triples about libby are now smushed with others?

17:05:05 <eaon> now we have a small problem leo

17:05:17 <eaon> you didn't specify a nick ;)

17:05:21 <eaon> and most commands are nick based

17:05:27 <eaon> ^like-pubs Arnia

17:05:28 <julie> Cuths Bar ,

17:05:37 <eaon> ^pub-address Cuths Bar

17:05:37 <julie> Durham United Kingdom 12 South Bailey DH1 3EE ,

17:05:42 <LeoS> ok, that is not a problem for the master of gnowsis

17:05:56 <LeoS> mom

17:05:59 <eaon> :)

17:06:05 <eaon> ^icbm eaon

17:06:06 <julie> 48.2538 16.4384 , 48.2538 16.4384 , 48.2538 16.4384 , 48.2538 16.4384 ,

17:06:15 <pjz> Aredridel: yeah, you're right

17:06:19 <jsled_> jsled_ is now known as jsled

17:06:53 <crschmidt> ^query select ?name where (?p foaf:mbox_sha1sum "eaa363e2b75a11db14e62afb8995340a198e8b9e") (?p foaf:name ?name) using foaf for <http://xmlns.com/foaf/0.1/>

17:06:53 <julie> Leo Sauermann ,

17:07:07 <crschmidt> ^query select ?name where (?p foaf:mbox_sha1sum "eaa363e2b75a11db14e62afb8995340a198e8b9e") (?p foaf:nick ?name) using foaf for <http://xmlns.com/foaf/0.1/>

17:07:08 <julie> Query select ?name where (?p foaf:mbox_sha1sum "eaa363e2b75a11db14e62afb8995340a198e8b9e") (?p foaf:nick ?name) using foaf for <http://xmlns.com/foaf/0.1/> returned no results

17:08:21 <pjz> ^query select * where (?p foaf:name "Leo Sauermann") using foaf for <http://xmlns.com/foaf/0.1/>

17:08:21 <julie> (r1094178929r39) ,

17:08:55 <crschmidt> ^query select ?nick1 where (?p2 kiss:kissed ?p) (?p foaf:name ?nick1) (?p2 foaf:name "Leo Sauermann") using foaf for <http://xmlns.com/foaf/0.1/>, kiss for <http://www.gnowsis.org/ont/kissology#>

17:08:55 <julie> Chris Schmidt , Ingrid Brunner-Sauermann ,

17:09:02 <crschmidt> apparently leo kissed me?

17:09:13 * crschmidt isn't sure how that happened

17:09:27 <pjz> heh

17:10:12 <eaon> hehe

17:10:37 <inkel> crschmidt, how does julie handle latin characters?

17:10:47 <crschmidt> inkel: no idea :)

17:11:01 <inkel> ha! Let's try then

17:11:08 <crschmidt> I know that it deals okay with utf-8

17:11:17 <inkel> ^add http://f14web.com.ar/inkel/rdf/foaf/inkel.xrdf

17:11:17 <julie> Adding that to my database...

17:11:26 <julie> Added 128 statements from http://f14web.com.ar/inkel/rdf/foaf/inkel.xrdf. Model size is 11368.

17:11:33 <LeoS> hehe

17:11:38 <inkel> what's next?

17:11:45 <eaon> LeoS: so how come you kissed crschmidt?

17:11:47 <crschmidt> run a query?

17:11:50 <inkel> ^smush

17:11:51 <julie> Smushing data. Currently 11368 triples.

17:11:55 <julie> Data smushed: 11368 statements to 11360 statements.

17:11:57 <LeoS> hm, that is the problem with trust.

17:12:02 <inkel> now yes, the query

17:12:10 <crschmidt> what do you want to know?

17:12:14 <LeoS> probably some sneaker said this to julie about me kissing crschmidt

17:12:15 <inkel> mi name

17:12:18 <inkel> my name

17:12:27 <eaon> ^query select ?name where (?p foaf:nick "inkel") (?p foaf:name ?name) using foaf for <http://xmlns.com/foaf/0.1/>

17:12:28 <julie> Leandro Mariano López ,

17:12:35 <inkel> great!

17:12:36 <eaon> yo leandro :)

17:12:45 <inkel> yeap, that's me :)

17:12:51 <crschmidt> tu nombre es Leandro, hombre!

17:13:03 <inkel> :)

17:13:23 <inkel> So it works ok with latin characters

17:13:40 <crschmidt> yep

17:13:41 <eaon> it works okay with utf-8 i guess :)

17:13:44 <crschmidt> I expected it would

17:13:52 <crschmidt> as I've seen proper utf-8 come out with it

17:13:58 <eaon> ^like-musicians eaon

17:13:59 <julie> Sigur Rós , Žagar , Autechre , Mogwai , Björk , Godspeed You! Black Emperor , B. Fleischmann , Múm , Arvo Pärt , Miraque I Myro ,

17:14:07 <crschmidt> i just wasn't sure how it would deal with taking in non-utf8

17:14:13 <eaon> thats all utf-8

17:14:18 <inkel> theres is a >julie, help ?

17:14:23 <[GNU]> crschmidt: is the src somewhere available?

17:14:23 <crschmidt> ^info

17:14:24 <julie> I'm a Redland/Python based RDF query bot. Source in SVN at <http://crschmidt.net/svn/rdfpython/trunk/>. Commands are ^add <url>, which adds statements, ^newcommand <name> is <RDQL Query>, ^runcommand <commandname> <arguments> . ^commandlist lists current commands, ^commandlist <command> shows info on a command, ^part will have me part a channel, ^join <#chan> will have me join a channel. Talk to crschmidt for more. @@ More here

17:14:28 <inkel> thanks

17:14:29 <crschmidt> yes, and yes, read the ^info :)

17:14:35 <[GNU]> :) yes

17:14:37 <Talliesin> non-utf8 should either be parsed at the XML parser level, or else it should raise an error.

17:14:54 <crschmidt> Talliesin: That's what happens. Although the error is silent atm.

17:14:54 <Talliesin> Unless the non-UTF8 is UTF-16, as XML parsers MUST accept that too.

17:15:18 <Talliesin> No idea what charsets it does support?

17:15:28 <[GNU]> #\^/usr/bin/python

17:15:33 <[GNU]> what sthat?

17:15:33 <crschmidt> it's redland and python. You can probably figure it out from there :)

17:15:46 <Talliesin> US-ASCII is easy, just pretend it's UTF-8 but b0rk on any octet > 0x80

17:15:58 <crschmidt> [GNU]: that's a broken #! line

17:16:08 <[GNU]> ah, ok :)

17:16:10 <crschmidt> which i've never noticed becaus ei always run it as "python redlandbot"

17:16:20 <LeoS> ^add http://www.leobard.net/foaf.xml

17:16:21 <julie> Adding that to my database...

17:16:22 <julie> Added 102 statements from http://www.leobard.net/foaf.xml. Model size is 11462.

17:16:27 <LeoS> ^mush

17:16:30 <LeoS> ^smush

17:16:30 <julie> Smushing data. Currently 11462 triples.

17:16:34 <eaon> mush there LeoS

17:16:35 <julie> Data smushed: 11462 statements to 11366 statements.

17:16:36 <eaon> ;)

17:16:48 <eaon> ^kissed LeoS

17:16:48 <[GNU]> i guess i have found something to play with for tonight :)

17:16:48 <julie> Query select ?nick1 where (?p2 kiss:kissed ?p) (?p foaf:nick ?nick1) (?p2 foaf:nick "LeoS") using foaf for <http://xmlns.com/foaf/0.1/>, kiss for <http://www.gnowsis.org/ont/kissology#> returned no results

17:16:50 <LeoS> so....

17:16:53 <LeoS> ^commands

17:16:57 <crschmidt> ^commandlist

17:16:58 <julie> Current commands: olb, like-pubs, like-books, img, sha, dob, pub-address, like-musicians, keywords, jabber, picOfA, contact, aim, like-same-music-as, like-musicalwork, based_near, icbm, xfn_met, kissed, mbox, phone,

17:16:59 <Talliesin> ^mush should be a command. Preferably triggering some sort of easter egg.

17:17:06 <LeoS> ^img leobard

17:17:07 <julie> [http://www.gnowsis.com/img/leo.jpg] ,

17:17:11 <LeoS> holla

17:17:18 <crschmidt> Easter eggs are kind of silly considering the source is open ;)

17:17:19 <LeoS> ^kissed leobard

17:17:20 <julie> SmileLoki , SmileLoki ,

17:17:23 <crschmidt> ^xfn_met leobard

17:17:23 <julie> Query select ?name where (?p xfn:met ?p2) (?p2 foaf:name ?name) (?p foaf:nick "leobard") using foaf for <http://xmlns.com/foaf/0.1/>, xfn for <http://gmpg.org/xfn/1#> returned no results

17:17:24 <Talliesin> ^img talliesin

17:17:25 <julie> Query select ?img where (?p foaf:img ?img) (?p foaf:nick "talliesin") using foaf for <http://xmlns.com/foaf/0.1/> returned no results

17:17:28 <crschmidt> bah, the kissed query is breaking again

17:17:29 <Talliesin> ^img Talliesin

17:17:29 <julie> [http://www.hackcraft.net/images/jon.png] ,

17:17:49 <crschmidt> leo: I think it's a bug in the way rasqal parses some queries

17:17:52 <LeoS> ^img crschmidt

17:17:53 <julie> Query select ?img where (?p foaf:img ?img) (?p foaf:nick "crschmidt") using foaf for <http://xmlns.com/foaf/0.1/> returned no results

17:17:57 <LeoS> °?

17:17:59 <crschmidt> There are no images of me :)

17:18:00 <Talliesin> Really that should search for foaf:img foaf:depicts and foaf:depiction queries.

17:18:03 <LeoS> gr

17:18:08 <eaon> ^img eaon

17:18:09 <julie> [http://zine.niij.org/photos/kaernten-march-2004/IM000578.JPG/variant/small] , [http://zine.niij.org/photos/general/IM000330.jpg/variant/small] ,

17:18:14 <crschmidt> Talliesin: If you can write it in RDQL, feel free

17:18:18 <eaon> there are TWO of me :)

17:18:26 <Talliesin> crschmidt, I tried before and failed.

17:18:28 <crschmidt> you can take botplay to #julie

17:18:41 <crschmidt> so as to not flood channel/logs with it

17:18:45 <Talliesin> You could entail foaf:depicts from the other two during the smush and then just look for that.

17:19:24 <inkel> ^add http://f14web.com.ar/inkel/rdf/foaf/inkel-depictions.xrdf

17:19:25 <julie> Adding that to my database...

17:19:26 <julie> Added 27 statements from http://f14web.com.ar/inkel/rdf/foaf/inkel-depictions.xrdf. Model size is 11393.

17:19:32 <inkel> ^smush

17:19:33 <julie> Smushing data. Currently 11393 triples.

17:19:37 <julie> Data smushed: 11393 statements to 11389 statements.

17:19:42 <inkel> ^img inkel

17:19:42 <julie> Query select ?img where (?p foaf:img ?img) (?p foaf:nick "inkel") using foaf for <http://xmlns.com/foaf/0.1/> returned no results

17:19:44 <LeoS> nick leobard

17:19:53 <LeoS> LeoS is now known as leobard

17:19:58 <eaon> ^kissed leobard

17:19:59 <julie> SmileLoki , SmileLoki ,

17:20:02 <inkel> great! there aren't pictures of me :)

17:20:03 <crschmidt> yeah, that's broken

17:20:10 <crschmidt> ^kissed crschmidt

17:20:11 <julie> SmileLoki , smileloki , SmileLoki , pie_is_good , jessical , littledownpour ,

17:20:14 <eaon> SmileLoki ?

17:20:14 <crschmidt> not sure why

17:20:22 <eaon> who is smileloki?

17:20:35 <crschmidt> smileloki is someone i've kissed ;)

17:20:46 <pjz> "multiple times, it seems

17:20:48 <eaon> aaaha

17:20:50 <eaon> ;)

17:20:52 <crschmidt> no, she just has multiple nicks

17:20:52 <eaon> hehe

17:20:54 <crschmidt> :)

17:20:59 <eaon> ^kissed jessical

17:20:59 <julie> smileloki , SmileLoki , SmileLoki , crschmidt ,

17:21:00 <inkel> who cares... it's better for your eyes if there are no pictures of me

17:21:10 <crschmidt> ^kissed raventhon

17:21:11 <julie> emmavescence , iniquity , damp_goddess , SmileLoki , SmileLoki , SmileLoki , SmileLoki , cyco , ozsager , skycrashesdown , agate456 , katieimp , matreia , sauri36 ,

17:21:13 <[GNU]> pretty resistent to channel changes

17:21:14 <crschmidt> hm.

17:21:23 <crschmidt> apparently loki has been kissed by everyone somehow

17:21:25 <crschmidt> must fix that

17:21:29 <pjz> heh

17:21:42 <leobard> one moment: didn'T we define kiss:kiss as being a loveing kiss?

17:21:52 <leobard> how could this be?

17:21:57 <leobard> thats orgiastic

17:22:09 <inkel> #foaf-porn ?

17:22:09 * crschmidt gestures to his wordpress

17:22:12 <eaon> haha

17:22:26 <eaon> leobard: make a better distinction then - new kissology version!

17:22:56 <eaon> kiss:relative kiss:loving kiss:fun kiss:orgiastic

17:23:21 <Talliesin> please no foaf-porn

17:23:31 <[GNU]> crschmidt: why would it always join #test ?!

17:23:36 <Talliesin> I did find some RDF metadata for porn recently.

17:23:39 <leobard> hm. well, lets decide the fun kissing, but not the orgiastic, they are not easy to diff

17:23:44 <inkel> yes, please, don't make me think of it... I'm at work!! ;)

17:23:59 <leobard> Talliesin: sounds interesting, finally RDF is commercially used?

17:24:04 <eaon> Talliesin: porn for rss?

17:24:09 <eaon> via, rss that is

17:24:13 <leobard> no rss for porn!

17:24:19 <pjz> hah, someone did solve what I was thinking about last night (restricting access to your foaf info)

17:24:21 <eaon> fleshbot has rss feeds

17:24:25 <eaon> not sure if rss 1.0 though

17:24:39 <leobard> what is the dreaded task of fleshbot??

17:24:49 <Aredridel> /me laughs

17:24:53 <Aredridel> Fleshbot++

17:25:04 <eaon> fleshbot is a weblog

17:25:08 <Talliesin> No, porn with RDF metadata.

17:25:21 <leobard> say it in RDF man: URL!

17:25:23 <eaon> wait?

17:25:24 <eaon> realyl?

17:25:45 <leobard> f*** the literal fleshbot, i want the url

17:25:58 <eaon> www.fleshbot.com

17:26:17 <eaon> ^website-title fleshbot :P

17:26:27 <leobard> (nooo, this link opened in my livestream burning man browser window and closed my stream)

17:30:33 <pjz> PEDAL is nice

17:31:14 <Aredridel> Focus follows click, not mind.

17:38:09 <pjz> .chuckle

17:38:28 <crschmidt> ^query select ?dep, ?title where (?p foaf:depiction ?dep) (?dep dc:title ?title) (?p foaf:nick "inkel") using foaf for <http://xmlns.com/foaf/0.1/>, dc for <http://purl.org/dc/elements/1.1/>

17:38:29 <julie> [http://usuarios.sion.com/zero/fotos/inkel.jpg] An ugly picture of an ugly guy , [http://usuarios.sion.com/zero/fotos/inkel.jpg] Una foto fea de un tipo feo , [http://purl.org/net/inkel/photos/mer-inkel-sillon.jpg] Mer y yo sentados en uno de los sillones del Hotel Bavaro H10 en Bavaro, República Domincana , [http://purl.org/net/inkel/photos/inkel-lancha.jpg] Yo sosteniendome en la lancha , [http://purl.org/net/inkel/photos/mer-inkel-lan

17:38:29 <julie> cha.jpg] Mer y yo en la lancha, en República Dominicana , [http://purl.org/net/inkel/photos/inkel-hamaca.jpg] Yo reposando muy tranquilamente en las hamacas del hotel en República Dominicana , [http://purl.org/net/inkel/photos/mer-inkel-palmeras.jpg] Mer y yo en la playa del hotel en República Dominicana , [http://purl.org/net/inkel/photos/mer-inkel-balcon.jpg] Mer y yo en el balcón de nuestra habitación en República Dominicana , [h

17:38:52 <crschmidt> inkel: and you said there was no pictures of you.

17:39:04 <inkel> eeeeeeerrrrr....... yes :)

17:39:27 <eaon> ^query select ?img, ?title where (?p foaf:img ?img) (?dep dc:title ?title) (?p foaf:nick "inkel")

17:39:28 <julie> Query select ?img, ?title where (?p foaf:img ?img) (?dep dc:title ?title) (?p foaf:nick "inkel") returned no results

17:39:31 <eaon> duh

17:39:36 <eaon> ^query select ?img, ?title where (?p foaf:img ?img) (?dep dc:title ?title) (?p foaf:nick "eaon")

17:39:36 <julie> Query select ?img, ?title where (?p foaf:img ?img) (?dep dc:title ?title) (?p foaf:nick "eaon") returned no results

17:39:41 <eaon> what!

17:39:45 <crschmidt> do you have dc:titles attached?

17:39:49 <eaon> yes

17:39:52 <crschmidt> to a foaf:img?

17:39:58 <crschmidt> you're not specifying namespaces, dork

17:40:12 <eaon> hehe

17:40:16 * Aredridel laughs

17:40:22 * eaon does too

17:40:37 <eaon> ^query select ?dep, ?title where (?p foaf:depiction ?dep) (?dep dc:title ?title) (?p foaf:nick "inkel") using foaf for <http://xmlns.com/foaf/0.1/>, dc for <http://purl.org/dc/elements/1.1/>

17:40:38 <julie> [http://usuarios.sion.com/zero/fotos/inkel.jpg] An ugly picture of an ugly guy , [http://usuarios.sion.com/zero/fotos/inkel.jpg] Una foto fea de un tipo feo , [http://purl.org/net/inkel/photos/mer-inkel-sillon.jpg] Mer y yo sentados en uno de los sillones del Hotel Bavaro H10 en Bavaro, República Domincana , [http://purl.org/net/inkel/photos/inkel-lancha.jpg] Yo sosteniendome en la lancha , [http://purl.org/net/inkel/photos/mer-inkel-lan

17:40:38 <julie> cha.jpg] Mer y yo en la lancha, en República Dominicana , [http://purl.org/net/inkel/photos/inkel-hamaca.jpg] Yo reposando muy tranquilamente en las hamacas del hotel en República Dominicana , [http://purl.org/net/inkel/photos/mer-inkel-palmeras.jpg] Mer y yo en la playa del hotel en República Dominicana , [http://purl.org/net/inkel/photos/mer-inkel-balcon.jpg] Mer y yo en el balcón de nuestra habitación en República Dominicana , [h

17:40:43 <eaon> gnah

17:40:45 <eaon> no

17:40:47 <eaon> crap

17:40:51 * eaon hides

17:40:52 <inkel> hey! ;)

17:40:54 <crschmidt> specify your username, goof :P

17:41:30 <inkel> ^query select ?dep, ?title where (?p foaf:depiction ?dep) (?dep dc:title ?title) (?p foaf:nick "inkel") using foaf for <http://xmlns.com/foaf/0.1/>, dc for <http://purl.org/dc/elements/1.1/>

17:41:36 <inkel> damn

17:41:46 <inkel> ^query select ?dep, ?title where (?p foaf:depiction ?dep) (?dep dc:title ?title) (?p foaf:nick "mer") using foaf for <http://xmlns.com/foaf/0.1/>, dc for <http://purl.org/dc/elements/1.1/>

17:41:57 <[GNU]> crschmidt: any special thing to do with the mysql storage for the first time the bot runs??

17:42:46 <[GNU]> it wouldnt fetch any url, neither store it :)

17:42:47 <crschmidt> wait, the bot is thinking

17:43:00 <crschmidt> [GNU]: yeah, you have to add a "new='true'" the first time

17:43:17 <julie> [http://usuarios.sion.com/zero/fotos/inkel.jpg] An ugly picture of an ugly guy , [http://usuarios.sion.com/zero/fotos/inkel.jpg] Una foto fea de un tipo feo , [http://purl.org/net/inkel/photos/mer-inkel-sillon.jpg] Mer y yo sentados en uno de los sillones del Hotel Bavaro H10 en Bavaro, República Domincana , [http://purl.org/net/inkel/photos/inkel-lancha.jpg] Yo sosteniendome en la lancha , [http://purl.org/net/inkel/photos/mer-inkel-lan

17:43:18 <julie> cha.jpg] Mer y yo en la lancha, en República Dominicana , [http://purl.org/net/inkel/photos/inkel-hamaca.jpg] Yo reposando muy tranquilamente en las hamacas del hotel en República Dominicana , [http://purl.org/net/inkel/photos/mer-inkel-palmeras.jpg] Mer y yo en la playa del hotel en República Dominicana , [http://purl.org/net/inkel/photos/mer-inkel-balcon.jpg] Mer y yo en el balcón de nuestra habitación en República Dominicana , [h

17:43:18 <julie> [http://usuarios.sion.com/zero/fotos/inkel.jpg] Una foto fea de un tipo feo ,

17:43:37 <[GNU]> but thats in the try-catch thing

17:43:50 <[GNU]> catch?! too much java

17:43:55 <crschmidt> [GNU]: i've never tried the try-catch, so i don't know if that actually works ;)

17:44:04 <[GNU]> :)

17:44:07 <[GNU]> it doesnt

17:50:40 * pjz looks into cwm.

17:50:53 * pjz kinda shudders.

17:52:52 <[GNU]> ah.. the mysql storage cant be created by redland

17:52:57 <[GNU]> but why?!

17:53:27 <crschmidt> did you create the redland database?

17:54:41 <[GNU]> using mysqladmin and the same user/passwd ja

17:54:47 <[GNU]> but it contains no tables

17:54:50 <[GNU]> is that ok?

17:55:37 <crschmidt> well, it should make them

17:56:25 <[GNU]> and spider.py should work?

17:56:35 <crschmidt> no, no

17:56:39 <crschmidt> spider.py is horribly broken

17:56:44 <crschmidt> and doesn't really work at all

17:57:01 <crschmidt> i haven't gotten finished with it yet

17:57:05 <[GNU]> but the Storage creation should work in spider?!

17:57:10 <crschmidt> dunno

17:57:15 <crschmidt> haven't even gotten far enough to check that

17:57:58 <[GNU]> ok, strings are equal in spider and redlandbot

17:58:06 <[GNU]> spider doesnt create the storage at all

17:58:24 <[GNU]> and redlandbot just says "Adding that..." and nothing else after a ^add

17:58:35 <crschmidt> did you look in the log that it prints?

17:58:43 <crschmidt> oftentimes there is more detailed error mesages there, although not always

17:58:49 <[GNU]> ja

17:58:50 <crschmidt> since i'm kind of ignoring redlanderrors at the moment, heh

17:58:52 <[GNU]> nothing good

17:59:11 <[GNU]> how to raise the level of redland debug output?

17:59:28 <crschmidt> you have to pull off a try-catch somewhere in the ircAsync code

17:59:45 <[GNU]> in put() i guess

18:00:16 <crschmidt> i think in:

18:00:17 <crschmidt> try:

18:00:17 <crschmidt> self.rxdMsg(args, text, origin)

18:00:17 <crschmidt> except:

18:00:17 <crschmidt> pass

18:00:21 <crschmidt> pull that off

18:01:23 <[GNU]> error: uncaptured python exception, closing channel <__main__.T connected at 0x402a9f6c> (RDF.RedlandError:'Creating Storage failed' [/usr/lib/python2.3/asyncore.py|read|69] [/usr/lib/python2.3/asyncore.py|handle_read_event|390] [/usr/lib/python2.3/asynchat.py|handle_read|136] [redlandbot.py|found_terminator|128] [redlandbot.py|rxdMsg|158] [redlandbot.py|put|259] [/usr/lib/python2.3/site-packages/RDF.py|__init__|1356])

18:01:26 <[GNU]> negative

18:01:41 <[GNU]> the storage creation is the initial problem

18:01:46 <inkel> I took me less than a minute to compile smusher.c, but I took me a lot to install the needed libraries :)

18:01:48 <[GNU]> need to cut that out

18:02:06 <[GNU]> inkel: and is redland.py running and adding urls?

18:02:13 <[GNU]> redlandbot.py

18:02:16 <inkel> anyway, crschmidt, perhaps you should know this: /usr/bin/ld: warning: libcurl.so.2, needed by /usr/lib/librdf.so, may conflict with libcurl.so.3

18:02:22 <inkel> Not yet

18:03:08 <[GNU]> maybe my redland is broken

18:03:22 <[GNU]> crschmidt: what redland are you using? from src or a .rpm?

18:03:43 <crschmidt> src

18:03:46 <pjz> crschmidt: btw, PEDAL seems to be the answer to my question about authentication description

18:04:06 <crschmidt> pjz: good to hear that you found it

18:04:17 <pjz> now to figure out how to use it... :)

18:04:49 <crschmidt> [GNU]: Redland config reports version as .9.18

18:05:05 <crschmidt> i have no doubt that it probably won't work with things other than that

18:05:47 <crschmidt> [GNU]: did you compile it with mysql? It mya not have found mysql when it was compiling, so you might not have the mysql bindings

18:05:52 <[GNU]> so i send my debian in an upgrade cycle

18:06:12 <crschmidt> Latest redland isn't in debian yet, i don't think, although others probably know better

18:06:14 <[GNU]> using binary debians from ac.uk

18:06:37 <[GNU]> Version: 0.9.18-3

18:06:38 <crschmidt> ah, okay

18:06:42 <crschmidt> hm

18:06:55 <crschmidt> i fixed the kissed query, by the way

18:07:00 <[GNU]> hmm, doens depend on mysql

18:07:03 <crschmidt> ordering of search terms makes a difference

18:07:03 <[GNU]> that sould worry me

18:07:15 <crschmidt> [GNU]: yeah, it will install mysql at compile-time if it finds it

18:09:15 <[GNU]> ok, just downgraded to 0.9.16 containing mysql support in the ac.uk debian binary packages

18:10:18 <[GNU]> now the redland mysql storage is working

18:10:30 * crschmidt nods, cool

18:10:56 <[GNU]> tables are created and data stored in them

18:11:02 <[GNU]> so... 0.9.16 is ok to

18:11:11 <[GNU]> should be the recommended version for debian users

18:11:55 <crschmidt> heh

18:13:08 <[GNU]> hmm... got a commandlist for me ? :)

18:13:58 <crschmidt> in svn now

18:14:14 <crschmidt> you can also just add them

18:14:18 <crschmidt> via ^newcommand

18:14:28 <[GNU]> i really love irc interfaces to svn commiting bots :)

18:17:20 <dustym> GNU, I had to compile with-mysql even though debs depend on mysql.

18:17:44 <[GNU]> hmm, mine 9.16 works ok

18:17:54 <crschmidt> the one that's in svn right now isn't actually up to date

18:18:41 <crschmidt> just committed

18:18:43 <[GNU]> and simply putting it in the ./ gives me plenty of 20:13:49) [GNU]: ^olb [GNU]

18:18:43 <[GNU]> (20:13:50) Datenboter: No such command.

18:18:43 <[GNU]> (20:14:02) [GNU]: ^sha [GNU]

18:18:43 <[GNU]> (20:14:03) Datenboter: No such command.

18:19:11 <dustym> cool. just letting you know that 0.9.18 will compile with mysql on sid just fine, in case you want to run the latest release w/mysql.

18:19:29 <[GNU]> dustym: just did a recompile?

18:19:40 <[GNU]> ahm... debian/rules binary ?

18:19:41 <crschmidt> [GNU]: Does ^commandlist do anything?

18:19:56 <[GNU]> (20:18:19) [GNU]: ^commandlist

18:19:56 <[GNU]> (20:18:19) Datenboter: Current commands: olb, like-pubs, like-books, img, sha, dob, pub-address, like-musicians, keywords, picOfA, contact, aim, like-same-music-as, like-musicalwork, based_near, icbm, xfn_met, kissed, mbox, phone,

18:20:21 <[GNU]> (20:18:43) [GNU]: ^sha [GNU]

18:20:21 <[GNU]> (20:18:44) Datenboter: No such command.

18:21:02 <crschmidt> i think "No such command" is code for "The query is broken"

18:21:07 <crschmidt> oh, hm

18:21:10 <crschmidt> you built .9.16?

18:21:14 <dustym> I didn't create any packages or anything.

18:21:35 <[GNU]> crschmidt: right, maybe upgrade to 0.9.18?! :)

18:21:54 <crschmidt> [GNU]: Rasqal isn't included in .9.16

18:21:57 <dustym> I just compiled --prefix=/usr

18:22:02 <crschmidt> you can't query unless you install rasqal

18:22:03 <crschmidt> ;)

18:22:45 <dustym> [GNU]: make sure you have libmysqlclient-dev installed.

18:24:05 <[GNU]> ja

18:25:54 <[GNU]> --with-mysql=no

18:25:59 <[GNU]> thats no good :)

18:27:35 <[GNU]> argh... no c++ compiler

18:39:25 <pjz> is a dc:description a valid thing to put on a Person ? there's no other 'about' thing I can find in FOAF

18:39:44 <danbri> bio:olb

18:39:50 <danbri> ...might be more what you're after

18:39:53 <danbri> .g bio:olb foaf

18:39:56 <phenny> bio:olb foaf: http://vocab.org/bio/0.1/

18:41:05 <crschmidt> ^olb Xavier

18:41:06 <julie> I'm a social geek who's into using technology to enhance and augment reality. I am also an information junkie who is rarely disconnected from the 'net, despite the physical world attempting to get in the way. ,

18:41:12 <danbri> coool

18:41:14 <crschmidt> for an example of what goes there

18:41:38 <crschmidt> i personally like this better:

18:41:43 <crschmidt> ^xfn_met redmonk

18:41:43 <julie> Nate Winesett , Damian Barrett , Robert Ochiallini , Jim Roepcke , Bill Brown , Wes Felter ,

18:41:45 <[GNU]> ^olb julie

18:41:46 <julie> Query select ?olb where (?p foaf:nick "julie") (?p bio:olb ?olb) using foaf for <http://xmlns.com/foaf/0.1/>, bio for <http://purl.org/vocab/bio/0.1/> returned no results

18:41:49 <crschmidt> heh

18:41:53 <crschmidt> julie has no foaf :)

18:41:55 <[GNU]> ahm... julie got no foaf?

18:42:03 <[GNU]> she's an agent

18:42:07 <[GNU]> needs to have a foaf file

18:42:18 <crschmidt> heh

18:42:25 <crschmidt> i have paying gigs to work on too :P

18:43:41 <[GNU]> :)

18:44:19 <crschmidt> Jess will soon start calling julie creusa

18:44:30 <crschmidt> give her a poisined robe, just like medea did

18:44:36 <[GNU]> .wn creusa

18:44:49 <[GNU]> .w creusa

18:44:52 <phenny> I couldn't find creusa in WordNet.

18:44:57 <[GNU]> hmm

18:45:22 <crschmidt> Creusa was neither a foreigner nor a sorceress, and so she seemed more attractive than Medea, whom Jason divorced in order to marry Creusa. Embittered, Medea sent Creusa a magnificent robe, soaked in a deadly potion. When Creusa donned the robe, it burst into fire and killed her in an instant.

18:45:29 <crschmidt> -- http://papa.essortment.com/jasonandthe_rjdy.htm

18:45:53 <pjz> what is it with poisoning robes? didn't someone try that with Hercules too?

18:45:58 <crschmidt> heh

18:45:59 <crschmidt> dunno

18:46:05 <crschmidt> but she named my laptop creusa

18:46:09 <crschmidt> and it's in the shop with a cracked lcd

18:46:13 <pjz> heh

18:46:25 <[GNU]> the powerbook is cracked?

18:46:38 <crschmidt> yep

18:47:27 <[GNU]> damned

18:47:41 <[GNU]> im just calculating to get a 15" with 1.5ghz

18:47:46 <[GNU]> but damned unchaep

18:47:48 <[GNU]> uncheap

18:49:35 <crschmidt> yeah, no kidding

18:49:47 <crschmidt> however, they sent it out for repairs as "warranty coverage recommended"

18:50:01 <crschmidt> if it's not, and the LCD is something i have to pay to repair... it's a $1200 fix.

18:54:16 <[GNU]> ja, sucks

18:54:42 <crschmidt> So, that machine will become a desktop :)

18:55:04 <crschmidt> But the website says "Repair started" as of wednesday, and they haven't called to let us know it's not warranty covered.

18:55:08 <crschmidt> so, hopefully all will be well.

18:55:54 <[GNU]> good luck :)

18:56:38 <crschmidt> yep :)

19:04:09 <pjz> anyone know why rdflib blows up on me on import due to a 'yield' statement?

19:04:32 <pjz> do I have to import __future__ or something?

19:05:24 <pjz> oh, hrm, not on import, but when I try to initialize a TripleStore()

19:07:22 <danbri> try in #rdfig, more likely right ppl notice

19:07:35 <danbri> or eikeon in #mindswap if he around, oh he's here too!

19:09:47 <pjz> ah, found the fix: syntax/serializers/XMLSerializer.py needs to 'from __future__ import generators'

19:16:08 * ndw nudges edd. You there?

19:16:43 <ndw> I guess not.

19:16:45 <ndw> brb.

19:23:45 <eikeon> pjz, It's either missing a from __future__ import generators statement...

19:24:01 <eikeon> or you're using a version of Python that's too old.

19:28:23 <pjz> it's missing that statement

19:28:50 <pjz> in rdflib/syntax/serializers/XMLSerializer.py

19:28:53 <pjz> if you care to fix it

19:29:01 <pjz> I'm using python 2.2.1, oldest you support

19:38:44 <eikeon> pjz, I'll double check... but think it's been reported and fixed... I've just yet to make the new release.

19:42:04 <pjz> ah, okay

19:42:08 <pjz> I just pulled it down today

19:53:46 <[GNU]> crschmidt: still arround??

19:54:04 <crschmidt> yep

19:54:08 <crschmidt> for another 10 minutes or so

19:54:36 <[GNU]> ah... ok... problem seems to be that if i fetch stuff with redland .9.16 it isnt recognized by .9.18

19:54:47 <[GNU]> which i have now, with mysql build as debian packages

19:55:02 <[GNU]> .sha works and .olb

19:55:38 <crschmidt> okay, so where's the problem? :)

19:55:39 <dajobe> what isn't recognised? The mysql schema has changed between 16 & 18 fwiw

19:55:51 <[GNU]> dajobe: ja, right

19:55:58 <[GNU]> crschmidt: ahm, nutting

19:56:08 <[GNU]> :) just wanted to say thank you :)

19:56:28 <crschmidt> ah, good :)

19:57:39 <crschmidt> ^query select query cause unknown error

19:57:39 <julie> Query select query cause unknown error returned no results

19:57:43 <crschmidt> hm.

19:58:06 <crschmidt> will fix when i get home, hopefully.

19:58:07 <crschmidt> later.

20:33:55 <pjz> is the bio spec written correctly?

20:35:43 <pjz> hrm, maybe I'm using a smart enough parsing engine (I'm just using rdflib)

20:42:49 <edd> norm!

20:42:55 <edd> ndw: just got back in

20:44:02 <danbri_dna> edd, hi

20:44:25 <danbri_dna>http://news.google.com/news?hl=en&ned=us&q=doap

20:44:27 <danbri_dna> doap is the new foaf :)

20:44:57 <danbri_dna> re the Adogato entry, DOAP's getting slagged for being XML

20:45:07 <danbri_dna> (not RDF :)

20:45:18 <ndw> hi edd!

20:46:12 <edd> danbri_dna: by you? :)

20:47:21 <danbri_dna> heheheh, no

20:47:23 <danbri_dna> i love xml

20:55:05 <Talliesin> Hit and run debate through the logs:

20:55:06 <Talliesin> A problem with leighd's model:

20:55:06 <Talliesin> libby -> memberOf -> mobloggers . mobloggers -> weblog -> moblog . doesn't mean that libby is posts to the moblog.

20:55:06 <Talliesin> Compare: I'm a member of the RSS1.0 WG (for whatever that might be worth these days). The RSS1.0 WG wrote the RSS1.0 spec. I however did not write any of the spec (I wasn't in the WG at the time, and in other case you can be a member of a group by not engaged in all it's activities).

20:55:12 <Talliesin> Hence with leighd's model we still need a connection between the person and the blog, whether direct or entailed by a predicate that says that all members of a group post.

20:55:15 <Talliesin> The connection between the person and the blog is what my property provides, so maybe there's still a case for it.

20:55:18 <Talliesin> loggerf: pointer

20:55:18 <Talliesin> See http://www.ilrt.bris.ac.uk/discovery/chatlogs/foaf/2004-09-03#T20-55-18

20:55:55 <Talliesin> phenny, ask leighd to take a look at http://www.ilrt.bris.ac.uk/discovery/chatlogs/foaf/2004-09-03#T20-55-18 and see if it counters the alternative model for shared properties he offered.

20:55:57 <phenny> Talliesin: I'll pass that on for you when leighd is around.

20:56:05 <Talliesin> Thanks phenny.

20:56:07 <phenny> You're welcome.

21:10:15 <pjz> Tall: hrm, so then maybe mobloggers -> bloggers ->libby ?

21:10:48 <pjz> oh, buh. nm.

21:11:21 <pjz> or maybe moblog -> author -> libby

21:12:03 <pjz> so you could explicitly see if moblog had >1 author

21:32:56 <edd> danbri, are you around in any of your incarnations

21:36:05 <danbri> hi

22:39:14 * dan-uk waves

22:58:46 * edd goes travelling

23:09:51 <teefal> hi morten... do you have link with photos?

23:10:40 <mortenf> hey

23:10:58 <mortenf> the ones that are already up are linked from the foaf-galway delicious tag

23:11:31 <mortenf> more to come...

23:11:35 <teefal> url?

23:11:45 <mortenf>http://del.icio.us/tag/foaf-galway

23:21:00 <teefal> did you get a pic of me with servlet exception?

23:21:24 <mortenf> well, i think only in separate ones

23:22:53 <mortenf> but see also .../5/raw/img_5684.jpg (2M) if you can't wait :)

23:24:27 <mortenf> hmm, seems the one i got was just a "service temporarily unavailable" :(

23:28:22 <dan-uk> mortenf: photoshop :)

23:28:31 <mortenf> right ;)

23:33:02 <teefal> ah... dialup speeds... will work later at galway "internet closet" :)

23:34:05 <dan-uk> you mean deri? ;)