Semantic Web Interest Group IRC Chat Logs for 2004-09-29

00:02:22 * crschmidt adds a photo to his key

00:02:37 <KjetilK> so, you caved in...? :-)

00:02:48 <crschmidt> Hm?

00:04:02 <KjetilK> weren't you here when we discussed whether it was any point to adding a photo...?

00:04:08 <crschmidt> Yeah, I thought there was :)

00:04:26 <KjetilK> hehe

00:05:11 <crschmidt> Hm.

00:05:25 <crschmidt> DanC_lap: It seems that when I changed my UID, I lost my signatures. I no longer ultimately trust myself.

00:05:59 <KjetilK> I think that isn't about signatures...

00:06:10 <KjetilK> but about the trustdb...?

00:06:24 <KjetilK> (but I'm not sure about these things...)

00:06:54 <crschmidt> Yeah, I'm sorry, misworded - what I meant was, after I changed my UID, the key (which I previously trusted) is now listed as not trusted.

00:07:08 <KjetilK> mmmm

00:07:35 <KjetilK> yeah, that is (I think) the behaviour I would expect... :-)

00:10:18 <crschmidt> Oh. Hm. I think I just broke something in mutt.

00:12:20 <KjetilK> ouch

00:13:15 <crschmidt> I get "Good signature from", but at the same time, get "PGP signature could NOT be verified"

00:15:12 <crschmidt> oh, i figured it out

00:16:05 <crschmidt> apparently, mutt uses the "pgp_good_sign" veriable to determine whether a signature is good. I thought it was text to use to print out if a signature was good, so I changed it, and all my signatures were "failing" cause the text didn't match.

00:19:22 <KjetilK> ok

00:23:11 <crschmidt> hm... difficult to back up secret keyrings, because you have to have control of any physical media they're stored on.

00:24:04 <KjetilK> yep

00:24:24 <KjetilK> and the media has to be relatively robust

00:25:00 <KjetilK> I used to have mine on a floppy locked down in my parent's house, but floppies die...

00:25:19 * crschmidt is still thinking of ways to identify himself without actually meeting people.

00:27:13 <crschmidt> I have a webcam. If I hold up piece of paper with key fingerprint, two forms of photo identification legible to the viewer - is that significantly different from actually meeting the person?

00:27:33 <crschmidt> I suppose there's the problem that video can be faked, wonder if that would be considered a significant problem.

00:27:58 <KjetilK> hm....

00:28:54 <KjetilK> yeah, it is easier to clip in a fake fingerprint beside your smiling face...

00:29:05 <KjetilK> meatspace is good for something...

00:29:48 <KjetilK> I would lsign based on that, but that wouldn't help your wot...

00:30:01 <crschmidt> yeah

00:31:06 <KjetilK> Boston can't be that bad, can it...? :-)

00:31:30 <crschmidt> heh

00:31:49 <crschmidt> it's more that i just don't know anyone, and would feel odd calling people and saying "Hey, can you sign my PGP key?"

00:31:56 <crschmidt> Plus, I don't have a car to get down there easily

00:32:00 * KjetilK does that... :-)

00:32:21 <KjetilK> those who sign up at biglumber are begging for it.... :-)

00:32:36 <KjetilK> I understand the car problem though

00:32:42 <KjetilK> I haven't got one either

00:32:45 <crschmidt> jess and i share a car

00:32:54 <KjetilK> ok

00:32:55 <crschmidt> which is fine, but for the fact that me taking it leaves her alone without one

00:33:03 <KjetilK> yeah

09:40:32 <chaalsBRS> Hola Ismael...

09:40:40 <chaalsBRS> todavia no he respuesto :-(

09:40:57 <chaalsBRS> Creí un vocabulario RDF en vez ...

09:41:04 <chaalsBRS> Me pongo al trabajo...

09:50:51 <Ismael> chaalsBRS no te preocupes

09:51:00 <Ismael> no hay prisa como para que te de un ataque de stress

09:51:02 <Ismael> :D

10:13:19 <chaalsBRS> He hecho una versión en el XML. Estoy describiendo una propuesta en RDF :-)

10:13:40 <Talliesin> Show off.

10:13:43 <chaalsBRS> Para que se puede saber como interpretarlo, visto que no hay esqueme del XML :-)

10:13:47 <chaalsBRS> :-P

10:14:16 <chaalsBRS> Hi Jon

10:14:48 <Ismael> jajja

10:14:50 <Ismael> eres un fiera

10:14:51 <Ismael> :D

10:14:58 <chaalsBRS> Anyone know of a class for "courses" (as in school or learning things)?

10:15:10 <Zenethian> that looks an awful lot about the very little Esperanto I know.

10:15:13 <Ismael> ay, si yo entiendiera como se usa realmente esto del rdf y el cwm

10:15:14 <Ismael> X-DD

10:17:32 * chaalsBRS finds CaptSolo's CV schema in Ontaria...

10:22:35 <Ismael> an schema for curricula?

10:22:43 <Ismael> please show the url

10:22:44 * chaalsBRS wonders about trying to convince captSolo about changes to the range of organizedBy (to include foaf:Person) and declaring courseDescription as a subProperty of dc:description

10:22:48 <chaalsBRS>http://kaste.lv/~captsolo/semweb/resume/cv.rdfs

10:22:49 <dc_rdfig> A: http://kaste.lv/~captsolo/semweb/resume/cv.rdfs from chaalsBRS

10:23:00 <chaalsBRS> !chump http://kaste.lv/~captsolo/semweb/resume/cv.rdfs

10:23:20 <chaalsBRS> oh. Bot not here.

10:23:39 <chaalsBRS> A:|CaptSolo's RDF Schema for Curricula Vitae

10:23:39 <dc_rdfig> Titled item A.

10:33:34 * chaalsBRS looks at Ivan's presentation ontology. Seems over complicated for what I am trying to do, and a little over-restrictive

10:33:41 <chaalsBRS> !chump http://kaste.lv/~captsolo/semweb/resume/cv.rdfs

10:33:45 <wh4experimental> no results found, sorry

10:34:11 <chaalsBRS>http://www.w3.org/2004/08/Presentations.owl

10:34:11 <dc_rdfig> B: http://www.w3.org/2004/08/Presentations.owl from chaalsBRS

10:34:16 <libby> !chump ~http://kaste.lv/

10:34:27 <wh4experimental> http://kaste.lv/~captsolo/info/blog_a.php/2003/11/13/p250

10:34:27 <wh4experimental> http://kaste.lv/~captsolo/info/blog_all.php/2004/01/06/p336

10:34:29 <wh4experimental> http://kaste.lv/~captsolo/info/blog_all.php/2004/04/14/p476

10:34:30 <wh4experimental> http://kaste.lv/~captsolo/info/blog_all.php/2004/05/06/p500

10:34:31 <wh4experimental> http://kaste.lv/~captsolo/semweb/index.html

10:34:57 <chaalsBRS> B:|Ivan Herman's Ontology for describing presentations.

10:34:57 <dc_rdfig> Titled item B.

10:35:32 <chaalsBRS> B:Made me realise why I love DanC's "rangeIntersects" property

10:35:33 <dc_rdfig> Added comment B1.

10:46:36 <chaalsBRS> B:Useful to kow that a foaf:Person is a subClass of the contact:Person that is the range of ivans:presenter

10:46:37 <dc_rdfig> Added comment B2.

10:47:20 * chaalsBRS thinks about how to include examples (skos:example parsetype="Literal" anyone?) in RDF vocabulary descriptions

10:57:41 <Talliesin> .g rangeIntersects

10:57:44 <phenny> rangeIntersects: http://dev.w3.org/cvsweb/2001/palmagent/icalVocab.rdf?rev=1.2

10:58:33 <Talliesin> file:/home/connolly/w3ccvs/WWW/2000/10/swap/util/sniffSchema# ?

10:59:56 <chaalsBRS> metadatos.cl have an ontology for stuff I want but it's in Java :-(

11:37:34 <Emmy> Hi all :)

11:37:59 <CaptSolo> hi libby

11:38:15 <CaptSolo> what about kaste.lv?

11:38:32 * CaptSolo should get a cool domain name instead of ~captsolo somewhere

11:38:34 <CaptSolo> i guess

11:38:53 <CaptSolo> ah...

11:39:00 * chaalsBRS should get a domain name.

11:39:13 <CaptSolo> chaals: i see where libby's query about kaste.lv came from

11:40:01 <CaptSolo> cv.rdfs is an ontology (classes & properties) rather than actual instance data that would be useable if loaded into the bot

11:40:47 <CaptSolo> a rather "green" ontology - needs many improvements, to mature up

11:41:27 <chaalsBRS> Sure. I used a bit of it in something I just did (not yet for publication - "commercial in confidence" ... :-)

11:41:54 <CaptSolo> "commercial in confidence" - sounds cryptic :)

11:41:56 <CaptSolo> hi danja :)

11:42:01 <chaalsBRS> I wanted something like the courseDescription property, but I went for dc:description instead, because it is better known.

11:42:05 <chaalsBRS> Hi danja

11:42:17 <danja> hi guys!

11:42:19 <CaptSolo> chaals: it could make sense to subclass

11:42:25 <chaalsBRS> ...so more intereoperability without looking up the term to see if there is something behind it.

11:42:41 <chaalsBRS> If you subclass then I will be really happy, and use the term (becausee it

11:42:53 <chaalsBRS> is more like what I want than dc:description is)

11:43:01 <CaptSolo> chaals: that is just what is needed for an ontology to mature - keep telling suggestions if they come to mind

11:43:39 <libby> hi all :)

11:43:58 <CaptSolo> chaals: i just do not know too much about what dc:description is for [intiutively i understand, but to make relations to it i must see for myself more]

11:43:59 <chaalsBRS> OT- "commercial in confidence" is the expression for why you don't publish details about how your system works, or for why governments don't publish information that shows they had a corrupt kickback deal going when they awarded an outsourcing contract...

11:44:06 <CaptSolo> hi libby :)

11:44:11 <Talliesin> hey danja, libby.

11:46:11 <danja> hi Talliesin, libby

11:47:39 <CaptSolo> libby: zool and schuyler are hi Talliesin!

11:47:52 * danja delights in Ontaria's knowledge of pets ;-)

11:48:08 <CaptSolo> chaals: i see - i shall remember that

11:48:09 <CaptSolo> ops

11:48:13 <CaptSolo> it was 2 lines:

11:48:19 <CaptSolo> libby: zool and schuyler are here again

11:48:23 <CaptSolo> hi Talliesin!

11:48:40 <eaon> hey guys

11:48:41 <eaon> brql was renamed to sparql?

11:48:47 <dajobe> in a sense

11:48:52 * dajobe lunchs

11:49:08 <dajobe> . http://lists.w3.org/Archives/Public/public-rdf-dawg/2004JulSep/0500.html

11:49:10 <chaalsBRS> Yeah, nice name huh? I guess it stands for "really useful way of finding semweb stuff" (a la OWL)

11:50:48 <danja> sparql is much better

11:51:38 <danja> brql sounds like someone about to vomit

11:51:47 <eaon> yep

11:52:49 <balbinus> :)

11:57:07 <jeen> sounds familiar...

11:58:53 <balbinus_> balbinus_ is now known as balbinus

12:07:12 <dajobe> I was looking for something pronouncable

12:07:29 <dajobe> so anything without a vowel was (mostly) out

12:16:37 <CaptSolo> danja: brql sounds as it's something out of the HitchHiker's guide to the galaxy

12:16:47 <CaptSolo> some alien's name or something :)

12:17:07 * CaptSolo met Valdis Krebs yesterday

12:17:20 <CaptSolo> was interesting to talk about social networks and their analysis

12:19:49 <JibberJim> did you try giving him a 20million foaf people?

12:21:08 <dajobe> dajobe has changed the topic to: Semantic Web / RDF chat n hack. blog http://rdfig.xmlhack.com/ (ADMIN: being upgraded, offline) logs http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/

12:21:31 <dajobe> dajobe has changed the topic to: Semantic Web / RDF chat n hack. blog http://rdfig.xmlhack.com/ (ADMIN: being upgraded, offline) logs http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/ | 2004-09-29 #foaf invasion 16:30UTC

12:25:00 <CaptSolo> Jim: nope, not yet at least

12:25:10 <CaptSolo> that was talk w/o computers

12:25:30 <CaptSolo> i told an short intro to a sem-web

12:25:45 <CaptSolo> ( if that can be told shortly at all )

12:27:24 <CaptSolo> Jim: 20 million foaf people - is that how many of us are out there?

12:27:31 <CaptSolo> i assume 95% are livejournal

12:29:08 <libby> I think LJ has about 2-4 million....

12:29:22 <libby> not sure of totals. 20million might be too high

12:29:45 <dajobe> tribe.net has foaf too I thought

12:30:43 <CaptSolo> ecademy has

12:31:01 <dajobe> " September 28, 2004 - 01:13 PM - FOAF import and export now on Tribe "

12:31:08 <JibberJim> ah no, LJ provides ~5.5million so probably around 6 million or so I reckon?

12:31:11 <CaptSolo> but if you add 'em up you should take into account that there may be duplicate persons

12:31:26 <CaptSolo> will make letss unique persons than the sum of FOAF persons

12:31:57 <crschmidt> tribe.net just st... yeah, what dajobe said

12:32:15 <JibberJim> how many do tribe have?

12:32:34 <dajobe>http://www.tribe.net/thread/81639067-3c2c-459d-9ba1-b6615f562d37?tribeid=94ba820b-162b-4540-9f87-ec807f343817&threads=true#$message.id

12:32:35 <dc_rdfig> C: http://www.tribe.net/thread/81639067-3c2c-459d-9ba1-b6615f562d37?tribeid=94ba820b-162b-4540-9f87-ec807f343817&threads=true#$message.id from dajobe

12:32:40 <libby> ah right, sorry

12:32:44 <dajobe> C:|September 28, 2004 - 01:13 PM - FOAF import and export now on Tribe, Paul Martino

12:32:45 <dc_rdfig> Titled item C.

12:50:19 <CaptSolo> Jim: so what is there to tell Valdis about the 20million of foaf people

12:50:57 <CaptSolo> i told some info, but i wonder where is the best place where a short intro the size of foaf universe is

12:51:31 <JibberJim> I imagine it's a larger dataset then he's worked with before...

12:53:01 <CaptSolo> i would think so

12:53:14 <CaptSolo> but there is a question of:

12:53:19 <CaptSolo> - data quality

12:54:08 <JibberJim> sure.

12:54:12 <CaptSolo> as i understood from him, in organization's research the best results are from questionaires that people fill in (of whom they consult to get some information and how frequently)

12:55:04 <CaptSolo> i.e., tracking emails (which also raises privacy issues) can tell you that these persons communicate, but does not say if it is about baseball, about situation reports or real knowledge transfer

12:55:56 <CaptSolo> same about FOAF - knowing that one person "knows" another would be sufficient for drawing nice social networks, but would not give you very much information that you could practically use

12:56:18 <CaptSolo> - privacy / confidentiality

12:56:33 <CaptSolo> in publicly available FOAF people would not put all their contacts

12:56:50 <CaptSolo> they would probably not mind giving them to you or me, but not to everybody

12:57:07 <CaptSolo> so the contents of FOAF often might be "censored"

12:57:31 <CaptSolo> Jim: but, still, I am looking for good pointers of FOAF info that i can show him

12:57:47 <CaptSolo> do you have ideas of some good urls?

13:53:49 <dajobe> dajobe has changed the topic to: Semantic Web / RDF chat n hack. blog http://rdfig.xmlhack.com/ logs http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/ | 2004-09-29 #foaf invasion 16:30UTC

14:50:57 <crschmidt> .t BST

14:51:00 <phenny> Wed, 29 Sep 2004 15:50:58 BST

15:03:29 <bparsia> bparsia is now known as bijan

15:06:31 <kpreid_> kpreid_ is now known as kpreid

15:23:14 <chaalsBRS> captsolo, what do you mean by "good" uris?

15:23:25 <sh1m> hi Charls.

15:24:29 <chaalsBRS> hi sh1m

15:24:33 <chaalsBRS> que tal?

15:26:14 <sh1m> Beunas. Tu?

15:30:18 <jeen> anybody know of a way to let a running process ignore hangup signals in Linux?

15:31:22 <jsled> `nohup`

15:31:36 <dajobe> you can't do it posthoc

15:31:37 <jsled> oh... "running"

15:31:42 <jeen> drat

15:31:46 <jeen> was afraid of that

15:31:55 <dajobe> to do with setting process group leaderships and detcaching from control terminals

15:31:58 <dajobe> detaching

15:32:17 <sh1m> You could make it very not nice... although I am not sure how much effect that would have more generally.

15:32:39 <jeen> was hoping it could be done, I found a man page for nohup somewhere that feature a -p switch that did just this... but GNU nohup does not support it

15:33:05 <dajobe> you could attach gdb

15:33:15 <dajobe> then invoke signals, syscalls by hand

15:33:16 <dajobe> ha ha

15:33:42 <dajobe> -- FOAF meeting here in +57mins --

15:34:21 <jeen> ah good, not necessary anymore: the process just threw a SocketException.

15:34:24 <jeen> grmbl.

15:35:11 <dajobe> suggest you use screen

16:18:29 <MacIntire> Meeting here in 10 min?

16:18:45 <mortenf> yup

16:24:00 * libby needs to chump agenda

16:25:47 <stpeter> greetings

16:27:00 <crschmidt> howdy stpeter

16:27:19 * eaon yawns

16:27:27 <stpeter> <-- Peter Saint-Andre ... http://www.jabber.org/people/stpeter.php

16:27:58 <stpeter> brb

16:27:58 <eaon> the invited one :)

16:28:24 <libby> BLURB: foaf meeting 1630 UTC 2004-09-29

16:28:24 * ear1grey squats into an active-lurk position

16:28:25 <dc_rdfig> D: foaf meeting 1630 UTC 2004-09-29 from libby

16:28:25 <stpeter> heh, but I'm not the anointed one ;-)

16:28:33 <libby> D:[http://rdfweb.org/pipermail/rdfweb-dev/2004-September/013737.html|agenda]

16:28:33 <dc_rdfig> Added comment D1.

16:29:09 <libby> D:attending [http://ilrt.org/people/libby|libby miller]

16:29:10 <dc_rdfig> Added comment D2.

16:29:10 <danbri> hi peter, glad you could make it

16:29:25 <danbri> libby's chairing the discussion today, thanks libby!

16:29:32 * stpeter cheers!

16:29:40 <libby> BLURB:Foaf Meeting previous actions

16:29:41 <dc_rdfig> E: Foaf Meeting previous actions from libby

16:29:49 <crschmidt> D: eating lunch [http://crschmidt.net|Christopher Schmidt]

16:29:49 <dc_rdfig> Added comment D3.

16:29:55 <danbri> D:attending [http://danbri.org/|dan brickley]

16:29:56 <dc_rdfig> Added comment D4.

16:29:56 <libby> E:|Foaf Meeting previous actions (E)

16:29:57 <dc_rdfig> Titled item E.

16:30:33 <libby> BLURB: Foaf meeting agenda item 1 - Domain of foaf:interest - Morten (F)

16:30:33 <dc_rdfig> F: Foaf meeting agenda item 1 - Domain of foaf:interest - Morten (F) from libby

16:30:40 * danbri glances over action list, wonders if crschmidt and bengee are here

16:30:53 * crschmidt is here, but did not complete naming action

16:30:56 <libby> F:see [http://rdfweb.org/pipermail/rdfweb-dev/2004-September/013708.html|morten's mail] and thread

16:30:57 <dc_rdfig> Added comment F1.

16:30:58 <mortenf> D:[http://www.wasab.dk/morten/|Morten Frederiksen] present

16:30:59 <dc_rdfig> Added comment D5.

16:31:04 <crschmidt> Was trying to pull it off last night, redland kept segfaulting on me

16:31:23 <libby> BLURB: foaf meetign agenda item 2 - Foaf views and profiles (G)

16:31:24 <dc_rdfig> G: foaf meetign agenda item 2 - Foaf views and profiles (G) from libby

16:31:29 <crschmidt> (A bug I still haven't tracked down enough to report, or upgraded to test if fixed)

16:31:33 <captainjim> D:attending [http://takepart.com|James Carlyle]

16:31:33 <dc_rdfig> Added comment D6.

16:31:54 <libby> G:see [http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-09-22.html#T17-32-27|agenda suggestion from last meet]

16:31:55 <dc_rdfig> Added comment G1.

16:32:01 <esigler> D: attending [http://esigler.2nw.net|Eric Sigler]

16:32:01 <dc_rdfig> Added comment D7.

16:32:23 <sh1mmer> D:[http://kid666.com/aboutme.xrdf|Tom Croucher] present

16:32:24 <dc_rdfig> Added comment D8.

16:32:30 <libby> BLURB: foaf meeting agenda item 3 - norm's mail (H)

16:32:31 <dc_rdfig> H: foaf meeting agenda item 3 - norm's mail (H) from libby

16:32:37 <balbinus> D:attending [http://foaf.balbinus.net/|Vincent Tabard]

16:32:37 <dc_rdfig> Added comment D9.

16:32:41 <libby> H:see [http://rdfweb.org/pipermail/rdfweb-dev/2004-September/013718.html|norm's mail]

16:32:42 <dc_rdfig> Added comment H1.

16:32:51 <libby> logger, pointer?

16:32:51 <libby> See http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-09-29#T16-32-51

16:33:09 <libby> D:[http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-09-29#T16-32-51|logs of this meeting]

16:33:09 <dc_rdfig> Added comment D10.

16:33:31 <libby> D:[http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-09-22.html#T16-28-52|logs of last meeting]

16:33:32 <dc_rdfig> Added comment D11.

16:33:54 <libby> D:[http://rdfig.xmlhack.com/2004/09/22/2004-09-22.html|agenda and actions from last meeting]

16:33:54 <dc_rdfig> Added comment D12.

16:34:02 <libby> .t UTC

16:34:04 <phenny> 2004-09-29T16:34:03Z

16:34:11 * libby back one sec

16:34:13 <mortenf> good job libby!

16:35:36 <libby> :)

16:35:39 <libby> more agenda items?

16:35:53 <JibberJim> D:[http://jibbering.com/|

16:35:54 <dc_rdfig> Added comment D13.

16:35:54 * libby sure we wanted to ask peter something specific....!

16:35:59 <danbri> stpeter, any other newcomers to #rdfig, just to note this is public and logged, with a weblog-ish summary at http://rdfig.xmlhack.com/

16:36:00 <eaon> D:[http://zine.niij.org/data/about-me|Michael Zeltner] present

16:36:02 <dc_rdfig> Added comment D14.

16:36:09 <JibberJim> D13:[http://jibbering.com/|Jim Ley] Present

16:36:09 <dc_rdfig> Replaced comment D13.

16:36:16 * stpeter nods to danbri

16:36:25 <danbri> libby, agenda+ Jabber update / discussion from stpeter?

16:36:35 <danbri> if you're still on for that, peter?

16:36:39 <stpeter> yeah, sure

16:36:41 <MacIntire> MacIntire is now known as bkdelong

16:36:50 <libby> yesh, let's do that

16:37:20 <stpeter> just let me know when you want me to dive in

16:37:21 <libby> BLURB:foaf agenda itme 4 - Jabber update/discussion from stpeter (I)

16:37:25 <danbri> quick run thru Actions maybe? how about continuing anything except things folk declare victory on now?

16:37:26 <stpeter> ik

16:37:27 <timbl> D:[http://www.w3.org/People/Berners-Lee/|Tim Berners-Lee] present

16:37:28 <dc_rdfig> I: foaf agenda itme 4 - Jabber update/discussion from stpeter (I) from libby

16:37:28 <dc_rdfig> Added comment D15.

16:37:29 <stpeter> ok, that is

16:37:56 <chaalsBRS> D:attending [chaals|http://www.w3.org/People/Charles/foaf.rdf]

16:37:57 <dc_rdfig> Added comment D16.

16:37:57 <libby> yep fine. we may want to change the order of the agenda items a bit

16:38:09 <danbri> completed-ACTION libby: send mail to rdfweb-dev re birthday stuff

16:38:24 <danbri> completed-ACTION danbri: invite Peter St Andre to a future IRC FOAF meeting

16:38:44 <mortenf> danbri, want to put an E: in front of those?

16:38:49 <danbri> E:completed-ACTION danbri propose to the list "that it should be a goal to ultimately get the FOAF spec in a shape (whatever shape that is) such that IETF, Dublin Core and W3C Working Groups are comfortable citing it ...

16:38:50 <dc_rdfig> Added comment E1.

16:38:53 <danbri> E:completed-ACTION danbri: invite Peter St Andre to a future IRC FOAF meeting

16:38:54 <dc_rdfig> Added comment E2.

16:38:56 <danbri> E:completed-ACTION libby: send mail to rdfweb-dev re birthday stuff

16:38:57 <dc_rdfig> Added comment E3.

16:39:06 <bkdelong> D:attending [bkdelong|http://foaf.brain-stream.org]

16:39:06 <dc_rdfig> Added comment D17.

16:39:14 <danbri> E:completed-ACTION: danbri let nicole know that Jim had problems with fontsize on new sit

16:39:15 <dc_rdfig> Added comment E4.

16:39:20 <libby> E:completed-ACTION bengee propose some clarifications re lifecycle/stability vocab to list, goal of having better machine-readable status for FOAF namespace

16:39:20 <dc_rdfig> Added comment E5.

16:39:37 <mcculley> D: attending [http://www.foafspace.com/|Gene McCulley]

16:39:38 <dc_rdfig> Added comment D18.

16:39:44 <danbri> crschmidt? did chris circulate any more stats? don't recall seeing...

16:39:51 <crschmidt> E: CONTINUED action on names - collecting more data (spidering) and also working to get it together into better stats

16:39:52 <dc_rdfig> Added comment E6.

16:39:59 <danbri> thanks chris

16:40:12 <danbri> E:Everything else continues...

16:40:13 <dc_rdfig> Added comment E7.

16:40:16 <mortenf> re bengee's action, his specific action may be completed, but the effects aren't?

16:40:18 <libby> E: CONTINUED action action danbri: get his FOAF talks from FOAFCamp and Galway online (and do the workshop report!)

16:40:18 <dc_rdfig> Added comment E8.

16:40:25 <libby> yep, neither are mine

16:40:41 * sh1mmer hides

16:40:44 * danbri wonders if the copy/paste continues can be done later, maybe move on?

16:40:45 <sh1mmer> I should do that too

16:40:51 <libby> yep, cool

16:40:57 <mortenf> ok, it'd just be sad to forget about them

16:41:16 <danbri> agree :)

16:41:37 <mortenf> i guess they should go on the issue list?

16:42:01 <libby> E:effects of bengees and libby's actions still in progress

16:42:01 <dc_rdfig> Added comment E9.

16:42:13 <mortenf> :)

16:42:14 * timbl wonders whether the topic should reflect the meeting

16:42:23 <libby> E:action Libby add these to the issue list

16:42:24 <dc_rdfig> Added comment E10.

16:42:41 <libby> Foaf meeting agenda item 1 - Domain of foaf:interest - Morten (F)

16:42:56 <danbri> danbri has changed the topic to: [#foaf invasion 16.30UTC 60-90mins] Semantic Web / RDF chat n hack. blog http://rdfig.xmlhack.com/ logs http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/ | 2004-09-29 #foaf invasion 16:30UTC

16:43:07 <mortenf> allright, basic idea outlined in message linked from the agenda item

16:43:37 <mortenf> i offer supporting evidence: the very place we are holding this meeting, the SW interest group...

16:43:46 <danbri> I saw a nice summary with pictures go past from Ian Davis, clarifies things to haev them drawn out...

16:43:55 <mortenf> i think it's fair to say that it's a group with an interest

16:44:16 <danbri> ah, ian's think was other issues, not so much the domain.

16:44:26 <mortenf> right, this agenda item shouldn't be taken as anything re ian's writeup and proposals

16:44:34 <captainjim> ian's issue was the range

16:44:36 <danbri> re-reading http://rdfweb.org/pipermail/rdfweb-dev/2004-September/013708.html

16:44:53 <danbri> IMHO it ought to be broadened to Agent

16:44:54 <mortenf> even if the change i propose goes through, it shouldn't make foaf:interest cast in stone

16:45:01 <danbri> I agree there too

16:45:38 <danbri> Characterising topics, interests etc is a huge issue. It may be regrettable that I used up the most generic property name on a construct that might seem to some to be a little quirky...

16:45:38 <captainjim> i agree foaf:interest domain should be Agent

16:45:39 <mortenf> i suggest we postpone discussion on ian's issue(s)

16:45:51 <danbri> ...that said, indirecting via characteristics of documents has some merit imho.

16:45:58 <mortenf> ... as mailing list discussion is still ongoing, and ian isn't here

16:46:01 <JibberJim> Agent's should definately be able to have interests.

16:46:03 <libby> seconded re ian's issues

16:46:12 * danja__ here while connection holds...

16:46:14 <JibberJim> foaf:Group is especially important

16:46:17 <danbri> mortenf, do you think we should alter rdfs:domain of foaf:interest to be foaf:Agent?

16:46:29 <mortenf> yep, that'd be nice.

16:46:47 <danbri> How about you propose it, we all go 'yeah!', see if anyone thinks otherwise, and action accordingly?

16:46:54 * danbri gets his 'yeah' button ready

16:46:55 <mortenf> :)

16:47:10 <mortenf> i propose changing the rdfs:domain of foaf:interest from foaf:Person to foaf:Agent.

16:47:14 <JibberJim> yeah

16:47:16 <eaon> yeah

16:47:17 <libby> yeah

16:47:18 <danja__> yeah

16:47:19 <captainjim> yeah

16:47:22 <danbri> +1 on changing rdfs:domain of foaf:interest to be foaf:Agent (after consulting rdfweb-dev)

16:47:27 <danbri> any objections?

16:47:28 * mortenf smiles

16:47:34 * danbri counts to 10

16:47:36 <danbri> 10

16:47:43 <balbinus> ...

16:47:53 * danbri counted fast; that was it ;)

16:48:01 <balbinus> :D

16:48:20 <chaalsBRS> yeah

16:48:22 <chaalsBRS> real fast

16:48:24 <danbri> E:ACTION: danbri update spec to broaden rdfs:domain of foaf:interest to be foaf:Agent (after consulting rdfweb-dev)

16:48:24 <mortenf> F:RESOLVED - rdfs:domain of foaf:interest to be changed to foaf:Agent

16:48:24 <dc_rdfig> Added comment F2.

16:48:25 <dc_rdfig> Added comment E11.

16:48:34 * libby proposes we bring item 4 (peter) to be next item

16:48:45 * mortenf seconds

16:48:46 <danbri> +1 re peter/jabber

16:49:00 * stpeter wakes up

16:49:15 <libby> foaf agenda itme 4 - Jabber update/discussion from stpeter (I)

16:49:18 <libby> :)

16:49:30 <stpeter> http://www.saint-andre.com/blog/2004-09.html#2004-09-17T09:25 is a bit of an overview

16:49:31 <dc_rdfig> J: http://www.saint-andre.com/blog/2004-09.html#2004-09-17T09:25 from stpeter

16:49:36 <danbri> Peter, I think a lot of folks are Jabber wellwishers, and sorta keep an eye on it, try out a client from time to time, vaguely know it's going through IETF as XMPP, but don't track the details....

16:49:44 <stpeter> ok

16:49:48 <stpeter> well, brief history

16:50:04 <stpeter> Jabber started out an an open-source IM project (actually a few related projects) in 1999

16:50:14 <libby> J:|overview of stpeter's comments re foaf and jabber

16:50:14 <dc_rdfig> Titled item J.

16:50:31 <stpeter> since then we've had more and more open-source projects, also many shareware and closed implementations as well

16:50:41 <libby> I:see [http://www.saint-andre.com/blog/2004-09.html#2004-09-17T09:25|stpeter's blog entry]

16:50:41 <danja__> D; attending [http://dannyayers.com|Danny Ayers]

16:50:41 <dc_rdfig> Added comment I1.

16:50:47 <jon-k> FWIW: i was at EDS ($22 B USD ISV in USA) and they use Jabber... I was shocked

16:50:48 <dc_rdfig> Label FWIW not found.

16:50:51 <stpeter> in 2001 we started the Jabber Software Foundation to manage the protocol development

16:51:02 <danja__> D: attending [http://dannyayers.com|Danny Ayers]

16:51:02 <dc_rdfig> Added comment D19.

16:51:14 * danja__ now waits for line to drop...

16:51:19 <stpeter> in late 2002 the core protocols were contributed to the IETF under the name Extensible Presence and Messaging Protocol (XMPP)

16:51:39 <stpeter> various additions were made etc during the IETF process (SASL/TLS/i18n stuff mainly)

16:51:49 <stpeter> we should have RFC numbers for those specs in a few weeks, I'd think

16:52:06 <stpeter> Jabber/XMPP is essentially a streaming XML technology (yes, it's wacky)

16:52:21 <stpeter> so your "session" is actually one big XML document that is built up over time

16:52:34 <stpeter> I'm sure timbl was not expecting someone to do that ....

16:53:10 * danbri nods, queues up a question

16:53:14 <stpeter> since the early days of the Jabber projects, we have used an XML representation of vCard, which never got beyond Internet-Draft in the IEF

16:53:18 <stpeter> IETf

16:53:20 <stpeter> er, whatever

16:53:46 <stpeter> documented here: http://www.jabber.org/jeps/jep-0054.html

16:53:49 <stpeter> that is a mess

16:53:53 <stpeter> and we need to move beyond it

16:53:54 * timbl wakes up was distracted

16:54:05 <stpeter> thus the interest in FOAF/RDF

16:54:37 <stpeter> we would like to describe not only persons but other kinds of entities on the network: bots, chatrooms, servers, etc.

16:55:17 <stpeter> at a minimum, we'd like to encapsulate the information that is in vCard as well as profile bits that are familiar from other IM systems, such as ICQ

16:55:23 <danbri> thanks for the renewed interest btw. One thing we talked about last week was taking steps to ensure that FOAF was citable from IETF, W3C, DC specs; whether or not that means getting it published thru such an org is another matter, but the will is there to be a 'respectable', citable spec...

16:55:43 <danbri> stpeter, how far along are you w/ requirements for describing 'agent' type things?

16:55:47 <stpeter> danbri: yes, I saw that -- it's a Good Thing [tm]

16:55:47 * chaalsBRS notes that there is interest in streaming XML in other places, like SVG

16:55:55 <timbl> (re streaming XML, David Gifford in CSAIL here was doing it some time ago, i think. Seems logical - we stram everything else. good browsers stream their html anyway)

16:56:04 <viclin> D: attending [http://www.schemaweb.info/|Victor Lindesay]

16:56:04 <dc_rdfig> Added comment D20.

16:56:10 <stpeter> I'm not very far along with requirements yet

16:56:22 <stpeter> I plan to write up some requirements in a more formal manner soon

16:56:42 <stpeter> basically the Jabber community has a lightweight standards process, similar to what the Python folks do

16:56:46 <danbri> ...eg being able to describe channels such as #rdfig is something I've always though FOAF good for, and in non-tech scene, being able to associate channels with places, people, topics ("this is a channel about Bristol " .."about George Bush", "about RDF") is a common wishlist.

16:56:58 * sandro__ is very confused about issues around streaming HTML. Is it better to stream or calculate the size first, so the user can see a % of how much is downloaded.....

16:57:10 <stpeter> the relevant docs are "JEPs", so I will work to write up a version 0.1 JEP on this in the next week or two

16:57:14 <danbri> I think it's pretty straightforward to do so in FOAF/RDF, given some conventions for identifying channels

16:57:20 <danbri> JEP, great, I look fwd to it

16:57:41 <stpeter> danbri: yes, being able to describe things like IRC channels or Jabber chatrooms is important to us

16:57:56 <danbri> I had a question, pasting in...

16:57:58 <stpeter> since we'd like to help people auto-discover places and people of interest if possible

16:58:27 <danbri> I'd *love* a way for chat channels for streets to be auto-findable, without needing a central website. Could have very interesting social effects...

16:58:28 * mortenf notes the partial overlap with foaf:AccountService

16:58:40 <danbri> Do jabber chatrooms have URIs?

16:58:49 <stpeter> mortenf: yes, or foaf:service?

16:59:02 <stpeter> I think foaf:Service could be interesting

16:59:05 <mortenf> foaf:service i haven't seen :)

16:59:21 <danbri> peter, did you seee http://xmlns.com/foaf/0.1/#term_jabberID

16:59:25 <stpeter> we don't normally refer to things as URIs, but any Jabber entity can have a URI

16:59:38 <stpeter> danbri: oh yes, that's in my personal FOAF file for sure ;-)

16:59:47 <mortenf> yep, there are certainly properties that apply well to account services (with logins) as well as channels

17:00:01 <danbri> it relates an 'Agent' (person, or other thingie) to a jabber ID. I wasn't sure if the associated text in the schema is adequate/precise enough. Do you think it needs any changes?

17:00:02 <stpeter> e.g., is this IRC channel a foaf:Group?

17:00:07 <stpeter> or something else?

17:00:26 <stpeter> danbri: will review again, but on first reading it seemed fine

17:00:27 * mortenf thinks it's something not yet named in foaf

17:00:48 <danbri> the individuals who hang out here are a Group, I think... but the technical/network entity that freenode IRC provides us... yes, is not modelled yet.

17:00:59 <danbri> anybody willing to have a go at that modelling task?

17:01:18 <mortenf> sounds like the coming JEP will have some input on that?

17:01:31 <stpeter> http://rdfweb.org/topic/AccountServiceIssue and http://rdfweb.org/topic/QuoteIssue may be relevant here as well

17:01:32 <dc_rdfig> K: http://rdfweb.org/topic/AccountServiceIssue from stpeter

17:01:41 <mortenf> indeed

17:01:50 <mortenf> (the first)

17:01:55 <stpeter> well, if Jabberites have recommendations for things to add to FOAF, we will do that through your process

17:02:11 <mortenf> the second interesting as well, need spec writeup etc.

17:02:28 <stpeter> the JEP will specify how Jabberites re-use what you've done, not define new terms etc.

17:02:29 <danbri> OK, bewarned that the process is kinda evolving and informal

17:02:43 <danbri> having more IRC catchups is at least creating a bit of a heartbeat re issue management

17:02:44 <stpeter> danbri: can't be more informal than the early days of Jabber ;-)

17:02:49 <mortenf> i was thinking that, if jabber people come up with a model for channels etc., foaf could piggyback off that?

17:02:53 <danbri> I'm interested to learn more aobut the Jabber experience on that front...

17:03:07 <danbri> my big pre-typed question:

17:03:08 <danbri> [[

17:03:09 <danbri> Q: The rendevous/zeroconf mode of Apple's iChat client uses Jabber

17:03:09 <danbri> directly in a machine-to-machine way, without intervening servers. This

17:03:09 <danbri> is close to some experiments people have been doing with FOAF (Damian

17:03:09 <danbri> Steer's FOAFFinger over zeroconf; Edd Dumbill's BlueFOAF app over

17:03:09 <dc_rdfig> Label Q not found.

17:03:10 <danbri> Bluetooth discovery). Being able to scan a local environment and pick up

17:03:11 <stpeter> danbri: we can chat offline about that

17:03:16 <danbri> characteristics (weblog URIs, photos etc) of people in 'public' mode

17:03:18 <danbri> locally seems a common theme. Anything happening standards-track or

17:03:20 <danbri> toolwise in Jabber community we could follow up on, in this area?

17:03:22 <danbri> ]]

17:03:26 * danbri nods, that'd be good

17:03:46 <stpeter> BTW, don't want to take more than my 5-10 minutes here, feel free to move on when desired

17:04:08 <stpeter> yes, there is some p2p Jabber protocol going on in iChat Rendezvous mode

17:04:14 <danbri> libby?

17:04:38 <stpeter> what is the best next step for me -- to write up our requirements a bit more formally?

17:04:58 * libby thinks so

17:05:01 <danbri> Dirk-Willem van Gulik showed me the ichat/jabber thing, am interested if there's a JEP or IETF writeup of the techniques...

17:05:04 <mortenf> that's a good step indeed

17:05:18 <danbri> yes, if you can write them up, we can shop them about and get reviews/comment, maybe implementation...

17:05:20 <stpeter> mainly I would like to participate in your process in an organic way rather than present a list of demands or anything like that

17:05:25 <libby> stpeter, will you repoirt back to the list when you have something?

17:05:33 <stpeter> libby: yepper

17:05:44 <danbri> sure, appreciated. just hang out in irc, email etc and we'll take care to catchup periodically

17:05:54 <danbri> Someday #rdfig and #foaf will be Jabber channels, I'm sure

17:05:55 <libby> thanks :)

17:05:58 <stpeter> I won't be able to attend the next two IRC chats

17:06:10 <stpeter> but I'll be on the list

17:06:12 <libby> ok, thanks for letting us know

17:06:19 <danbri> (I might have difficulty too, in Oct)

17:06:35 <danbri> thanks for the updates :)

17:06:42 <libby> if when you have something more formal, you'd like to make it an agenda item on a chat you make, just suggest something

17:06:43 <stpeter> so action item for me is to write up Jabber requirements and report to RDFWEB

17:06:47 <libby> :)

17:07:02 <libby> I:action stpeter: write up Jabber requirements and report to RDFWEB

17:07:03 <dc_rdfig> Added comment I2.

17:07:07 <stpeter> thanks for your interest :-)

17:07:20 <libby> ok, next up

17:07:24 <libby> --foaf meetign agenda item 2 - Foaf views and profiles (G)

17:07:59 <libby> captsolo, teefal intersted in this; pjz also I think:

17:08:09 <libby> http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-09-22.html#T17-32-27

17:08:23 <danbri> is that profiles in the sense of schemarama, application profiles, app-specific profiles of which properties etc would appear?

17:08:29 <danbri> guess so

17:08:41 <libby> hm, only pjz is here out of that list

17:08:49 * libby keen too of course

17:09:21 * mortenf thinks this is (also) about acl

17:09:40 <libby> yeah? can you expand a bit?

17:09:53 <danbri> about saying, 'you can see my day-of-birth but not my year-of-birth'?

17:10:16 <mortenf> <teefal>we've been talking about "views" (same as profiles) ... subsets of foaf properties that have ACLs applied to them

17:10:37 <danbri> G:SeeishAlsoish, [http://www.w3.org/2002/01/pedal/thesis.html|Personal Data Protection in the Semantic Web], Ryan Lee's thesis.

17:10:38 <dc_rdfig> Added comment G2.

17:10:41 * mortenf wasn't there...

17:11:01 <libby> hm, ok

17:11:10 <libby> I'm not sure what he means really

17:11:16 <stpeter> Jabber folk will probably be interested in such things as well -- defining "view"

17:11:29 <libby> postpone till teefal is back?

17:11:38 <danbri> there was some work Leigh did, using XML schematron rules to indicate profiles for different purposes

17:11:38 <mortenf> G:kind-of-seeAlso: [http://www.w3.org/2001/sw/Europe/events/foaf-galway/papers/pp/ghosts_in_the_semantic_web/|Ghosts in the Semantic Web Machine?]

17:11:39 <dc_rdfig> Added comment G3.

17:11:55 <danbri> some apps require certain fields, even though those fields aren't mandatory across all of foaf

17:12:00 <timbl> Hmmmm..... sets of properties are also relevant in form-filling .... sets of properties which must be defined in a document for the entry to be "valid" in some sense.

17:12:03 <libby> yes indeed

17:12:14 <libby> re both

17:12:15 <danbri> seealso libby/damian galway paper + damian's swad-e xform experiment

17:12:22 <danbri> is latter online anywhere yet?

17:12:25 <libby> yeah, I don;t think the latter is online yet

17:12:25 <ear1grey> D:attending [http://boakes.org/|Rich Boakes]

17:12:26 <dc_rdfig> Added comment D21.

17:12:37 <danbri> timbl, yes, exactly

17:12:53 <libby> shellac's been working on creating xforms from owl

17:12:58 <danbri> one lesson i got from being RDFS editor is that people really really expect RDF schema to do that work for them, and it simply doesn't.

17:13:18 <danbri> Mixing in OWL, per damian's experiment, was an interesting twist but I think the jury is still out there...

17:13:24 <libby> from experience, application profile type stuff is dead useful

17:13:47 <libby> if you're getting data from multiple sources, to show the sorts of things you'll accept/display

17:14:20 <danbri> anything we can get done today? other than agree This Is Important And Interesting? (which it is?)

17:14:28 <libby> G:related [http://www.w3.org/2001/sw/Europe/events/foaf-galway/papers/pp/validating_rdf/|Validating RDF with TreeHugger and Schematron]

17:14:29 <dc_rdfig> Added comment G4.

17:14:30 <danbri> maybe chase damian for xform demos...

17:14:36 <ear1grey> i'm wondering if forms/profiles could solve my "when is a scheme complete" question - since it could then be defined in terms of required presence or absence of properties

17:14:53 * chaalsBRS thought it was more useful to have a away of dscribing the things you have a nice way of handling in general, without it being an application profile as such...

17:15:02 <timbl> Well, folks will have to realize that the requirements of a given form just that and nothing about the ontology in general -- but given that, a solution in this area would probably go down quite well.

17:15:04 <mortenf> i guess we could ask the people behind the agenda item to start a thread on the list?

17:15:13 <danja__> how about one or two use cases, to make a sample profile or two?

17:15:21 <chaalsBRS> e.g. tool foo creates / presents / "understands" list X or properties and classes.

17:15:36 <chaalsBRS> Foafnaut being an example that I often have in mind

17:15:48 <libby> I think both important and interesting and at least 2 threads there

17:16:13 <timbl> (Profiles are also maybe interesting in queries (DAWG?) like "tell me all the financial stuff about xxx")

17:16:21 <chaalsBRS> Or foaf-a-matic creates the following kinds of info...

17:16:23 <bkdelong> I'm not sure I understand what is being discussed...does this have to do with say having a Master FOAF that is encrypted and have little FOAFlets of information accessible only to people whose public key it's encrypted with? ala ACLs?

17:16:37 <libby> no, though I think people want to do that

17:16:43 <chaalsBRS> bk, I think that is one of the threads here.

17:16:52 <danbri> yep, libby and i played around with using RDF queries to characterise RDF profiles a while back...

17:16:55 <danbri> .g schemarama libby

17:16:58 <phenny> schemarama libby: http://ilrt.org/discovery/2001/02/schemarama/

17:17:14 <danbri> ...new DAWG language more expressive than Squish, esp w/ optionals, might be worth revisiting then.

17:17:18 <libby> this is somewhat different bkdelong: an application can usually only display certain types of thing, therefore it's useful for others to knwo what these are

17:17:30 <libby> yeah, that could work. really needs optionals

17:17:30 <bkdelong> gotcha.

17:17:41 <libby> otherwise you end up with lots of if/elses

17:18:15 <libby> I'm not sure you could do it without some extra code though, even with optionals

17:18:21 * libby proposes moving on

17:18:24 <danbri> its a way of probing for some sort of 'descriptive adequacy' relative to some task at hand, rather than (per RDFS/OWL usually) just checking for inconsistent data

17:18:28 <danbri> +1 re moving on

17:19:00 <libby> G:action libby: start one thread on this and suggest to teelfal/captsolo/pjz start another

17:19:01 <dc_rdfig> Added comment G5.

17:19:07 <danbri> no norm?

17:19:29 <libby> don;t see him

17:19:39 <libby> he didn;t acually propose this as an agenda item

17:19:42 <libby> --foaf meeting agenda item 3 - norm's mail (H)

17:19:52 <libby> I just thought miht be useful to talk about it

17:19:53 <danbri> re http://rdfweb.org/pipermail/rdfweb-dev/2004-September/013718.html I have lately been wondering whether having a big fat spec at the namespace URI is the best thing

17:20:03 <libby> but we could do soemthign else...

17:20:06 <danbri> ...as we add more general bits into the doc, and have translations etc too

17:20:21 <mortenf> me would, with the increasing amount of term defs et al, like for a paged doc somewhere else...

17:20:42 <libby> use RDDL?

17:20:45 <danbri> I still want something human-friendly at the namespace URI

17:20:45 * crschmidt has often felt the schema doc is too big - takes *long* time to load/render

17:20:50 <libby> or just slit the document up?

17:20:57 <libby> split

17:20:58 <danbri> RDDL, is that human friendly? hmm

17:21:06 <libby> yeah, it's not bad

17:21:14 <danbri> anyone else finding http://xmlns.com/foaf/0.1/ takes ages to download?

17:21:19 <mortenf> yep

17:21:23 <danbri> this could be a hosting issue too :(

17:21:27 <libby> :(

17:21:42 <balbinus> so do i

17:22:03 <danbri> edd once suggested we could look into going into their hosting deal, maybe we should talk to edd, dajobe et al, libby?

17:22:04 <mortenf> how about the schema at the ns uri, with a rdf:parseType="Literal" chunk at the top that links elsewhere?

17:22:34 <crschmidt> Content-Length: 176573 is big - even at 20KB/s is 8-10 seconds to download

17:22:37 <libby> might make sense danbri. or look into better pipes here

17:22:51 <danbri> Ideally TAG and SW Best Practices WG would've figured out a suits-everone answer...

17:22:51 <timbl> Please put something which parses as RDF at the namespace name

17:23:03 * mortenf agrees

17:23:06 <danbri> does RDF inside XHTML meet that req., timbl?

17:23:12 <danbri> we have that now

17:23:20 <timbl> The TAG notoced that a suits-everyone answer is not realistic.

17:23:21 <danbri> also we have content negotiation

17:23:24 <danbri> yeah

17:23:36 <timbl> You could use content negotiation.

17:23:40 <danja__> wasn't norm's point that he couldn't easily get the rdf/xml?

17:23:44 <libby> yeah

17:23:49 <danbri> I don't find RDDL very interesting, if you're in the RDF world. For the XML scene, where there are a load of conflicting schema languages, it makes more sense.

17:24:02 <danbri> norm's using XSLT which doesn't support conneg

17:24:10 <timbl> I think RDF-in-HTML woudl meet that on the forward assumption that it becomes adopted, which is reasonable for FOAF to push. Or GRDDL.

17:24:12 <danbri> maybe he could still get to the embedded-in-xhtml version

17:24:30 <danja__> that'd be interesting...

17:24:33 <danbri> Yep, when it's in more parsers, we could switch to the new RDF-in-HTML syntax

17:24:51 <JibberJim> I want a text/html if text/html is sent, I think the foaf namespace becomes too google popular being used so much, and non-tech people want to get it, tech people should be able to organise content-neg if they want something else

17:24:55 <danja__> better for text-only browsers too

17:25:06 <danbri> I reckon for now, some continaution of current practice, ie. XHTML with RDF inside at the ns URI, but look to have 'The FOAF Spec' be a separate document, and minimal HTML doc only.

17:25:34 <mortenf> i propose rdf with xhtml inside...

17:25:36 <danja__> yep, that sounds good - get slitting ;-)

17:25:38 <danbri> jim, you don't want RDF-inside? you wouldn't count that as text/html ?

17:26:08 <JibberJim> RDF inside is fine - and text/html I'd count as just about anythign

17:26:11 <ear1grey> on the size issue: has it been considered to have a schema file specify mirrors rather than content?

17:26:26 <danbri> ok, phew, thought you were gonna launch some deep pedantry on me jim :)

17:26:27 <KjetilK> D: attending [http://www.kjetil.kjernsmo.net/|Kjetil Kjernsmo]

17:26:27 <dc_rdfig> Added comment D22.

17:26:46 <stpeter> sorry to chat and run, but I need to clear out -- will post to the list regarding my action item

17:26:50 <JibberJim> smaller file would also be good, my main concern is that people find it wanting to know about the spec, due to its exposure on the web as a URI

17:26:53 <chaalsBRS> or breaking up the schema so it's just a list of terms, and each term is defined by a document at the URI that is its name?

17:26:54 <danbri> we sorta could do mirrors with seeAlso-ish pointers, but no, don't think people looked much at that yet.

17:26:56 <mortenf> rapper: Error - URI http://xmlns.com/foaf/0.1/:3 - Using property attribute 'lang' without a namespace is forbidden.

17:27:25 <danja__> rapperslap

17:27:34 <danbri> I think there are significant communities, some wanting RDF at the ns, some wanting HTML, so we'll have to keep 'em both happyish somehow.

17:28:03 * libby gotta run too in a sec, sorry

17:28:06 <timbl> I think that if you content negotiate a resource, then you really are offering two encodings of the *same information* -- so the HTML better be generated from the RDF for example. Suggest that is the ontology, which has a pointer to the spec.

17:28:08 <danbri> Also, XSLT 2 is still in last call I believe, could always remind them that content negotiation is an important part of the web, and plead for support...

17:28:12 <danja__> I think it was only the conneg that was messing norm up, so XHTML should work for him

17:28:16 <libby> shall we action ourselves to look into hosting options danbri?

17:28:19 <danbri> timbl, yeah agree

17:28:25 <danbri> yes pls libby

17:29:04 <libby> H:action danbri/libby look into other hosting options for xmlns.com

17:29:05 <dc_rdfig> Added comment H2.

17:29:25 * mortenf notes that practically every other rdf ns serves rdf/xml at the ns uri, without conneg - also, foaf has a homepage at foaf-project.org

17:29:49 <danbri> H:danbri investigate splitting bulky HTML text out into a separate 'spec' document, leaving slimmer XHTML and RDF representations of the ns doc.

17:29:50 <dc_rdfig> Added comment H3.

17:29:55 <KjetilK> will it always be possible to meaningfully markup the same information in e.g. both HTML and RDF?

17:29:58 <danbri> H3:ACTION danbri investigate splitting bulky HTML text out into a separate 'spec' document, leaving slimmer XHTML and RDF representations of the ns doc.

17:29:59 <dc_rdfig> Replaced comment H3.

17:30:37 <danbri> The main difference is that we have big chunks of per-term XHTML documentation, which aren't currently exposed in the RDF representation

17:30:43 <danbri> though they could be...

17:30:51 <KjetilK> yup

17:30:56 <danbri> ...especially if schemaweb, ontaria etc could consume such additioanl notes

17:31:09 <danbri> (same goes for alternate language translations)

17:31:22 <danbri> <mortenf> rapper: Error - URI http://xmlns.com/foaf/0.1/:3 - Using property attribute 'lang' without a namespace is forbidden.

17:31:26 <danbri> hmm

17:31:44 <danbri> H:danbri investigate/fix <mortenf> rapper: Error - URI http://xmlns.com/foaf/0.1/:3 - Using property attribute 'lang' without a namespace is forbidden.

17:31:44 <dc_rdfig> Added comment H4.

17:31:54 <danbri> We better wind up before any more actions get dished out...

17:32:04 <mortenf> that's without the --scan option, which finds the 467 statements fine

17:32:08 * danbri not yet sure can make next week

17:32:21 <danja__> anyone considered <link rel="..."> to an rdf/xml version?

17:32:34 <danbri> hmm yeah could easily do that, if we don't already

17:32:35 <mortenf> also, when rapper fetches directly, it sends an accept header

17:32:42 <libby> next week would be 6th oct

17:32:48 <danbri> re "mortenf notes that practically every other rdf ns serves rdf/xml at the ns"

17:32:57 <danbri> ...yeah, I hoped we could do better ;)

17:33:10 <danbri> it can be quite unfriendly to find just a bunch of rdf/xml when someone points you at a vocabulary

17:33:16 * libby can run next week if there's demand

17:33:41 <mortenf> true, but the namespace is just that, and if it linked to the "real spec"

17:34:00 <mortenf> ... via xhtml in rdf

17:34:01 <danbri> yep, I think having it look ok-ish in a browser is a reasonable expectation tho

17:34:07 * danbri nods

17:34:21 <danbri> shall we adjourn and continue this rathole topic at leisure?

17:34:23 <mortenf> np, especially with the syntax hacks (that ndw dislikes)

17:34:27 <mortenf> yep

17:34:41 <danbri> next meeting: unconfirmed, maybe next week w/ libby running the show again

17:34:44 <timbl> unfriendly? Put a style sheet on it.

17:34:49 <danbri> CSS or XSLT?

17:34:51 <danbri> CSS can't do links

17:34:56 <danbri> XSLT is turing complete, trust issues

17:34:58 <timbl> And stick stg nearthe top pointing firmly at the home page.

17:35:29 * mortenf thinks of doing a proof-of-concept for his suggestion

17:35:39 <danbri> html inside rdf?

17:35:40 <mortenf> go ahead and action me on that :)

17:35:42 <danja__> action mortenf

17:35:43 <mortenf> yep

17:35:46 <mortenf> heh

17:36:23 <danbri> H:action mortenf draft a demo namespace showing HTML fragment inside RDF/XML schema markup, so it can link stylishly to external documents when viewed in browser.

17:36:23 <dc_rdfig> Added comment H5.

17:36:41 <danbri> OK let...

17:36:44 <danbri> ADJOURN

17:36:51 <danbri> aka fix the /topic

17:36:52 <mortenf> seconded :)

17:36:59 <mortenf> you're the op...

17:37:10 <danbri> danbri has changed the topic to: Semantic Web / RDF chat n hack. blog http://rdfig.xmlhack.com/ logs http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/

17:37:25 <danja__> what if...you still used conneg, but got rdf in html for text/html and html in rdf/xml for application/rdf+xml?

17:37:43 <crschmidt> mortenf: channel is -t, anyone can change a topic :)

17:37:47 <ear1grey> i've had to hack around the ns issue this and think a server should by default return the rdf/xml at a particular and have the browser negotiate for html - rather than the other way round

17:37:54 * danja__ adjourns to dogwalk

17:37:56 <mortenf> danja, that wouldn't help ndw with xslt

17:41:18 <mortenf> K:

17:41:18 <dc_rdfig>http://rdfweb.org/topic/AccountServiceIssue

17:41:47 <mortenf> K:|AccountServiceIssue

17:41:47 <dc_rdfig> Titled item K.

19:12:21 <balbinus> do you think it's better to use different base classes for different types of things, or use rdf:type? (in both cases, all classes are sub-classes of an unique superclass)

19:17:52 <chaalsBRS> balbinus, got some more context / example?

19:18:56 <Talliesin> balbinus, expand please.

19:19:11 <balbinus> :)

19:19:23 <balbinus> well, FOAF is an example (not a good one, but one anyway)

19:19:48 <balbinus> <foaf:Person>, <foaf:Organization>

19:20:30 <balbinus> or <foaf:Agent rdf:type="http://xmlns.com/foaf/0.1/Agent"> (i think it's correct

19:20:37 <chaalsBRS> Hmmm

19:21:33 <balbinus> )

19:21:34 <Talliesin> well <foaf:Person> is another way of saying:

19:21:41 <chaalsBRS> <foaf:Person> blablabla </foaf:Person> and <r:Description><r:type r:resource="http://xmlns.com/foaf/0.1/Person" /> blablabla </r:Description> ?

19:21:46 <chaalsBRS> They are the same thing

19:22:00 <Talliesin> <rdf:Description><rdf:type rdf:resource="http://xmlns.com/foaf/0.1/Person"/>...

19:22:03 <chaalsBRS> If you put them in the validator an ask it to show the graph, you get the same picutre

19:22:07 <balbinus> i know :)

19:22:15 <balbinus> but which one should be prefered?

19:22:30 <chaalsBRS> I don't see any good reason to prefer one or the other.

19:22:35 <balbinus> with something even closer than foaf:Agent and foaf:Organization

19:22:43 <balbinus> XSLT "parsing"

19:22:55 <Talliesin> You might possibly want to do something like:

19:23:12 <chaalsBRS> (your example: <foaf:Agent rdf:type="http://xmlns.com/foaf/0.1/Agent"> is tautological - <foaf:Agent already says that something is of that type)

19:23:35 <chaalsBRS> You mean generating, or reading?

19:23:49 <chaalsBRS> For generating it doesn't matter. For reading you need to handle both cases.

19:23:52 <Talliesin> <foaf:Person><rdf:type rdf:resource="http://example.net/ThingsFavouredByTheEmporer"/></>

19:24:04 <chaalsBRS> (unless you only read things that you have created...)

19:24:18 <Talliesin> In the case were an instance is an instance of two orthogonal classes.

19:24:18 <balbinus> chaalsBRS: i meant rdf:type="http://xmlns.com/foaf/0.1/Person", sorry :)

19:25:17 <Talliesin> Saying something is both a foaf:Agent and a foaf:Person is saying something, but that something will be understood by anything that does entailment, or by anything that is FOAF-specialised.

19:25:31 <balbinus> yes

19:25:38 <Talliesin> since all <foaf:Person>s are <foaf:Agent>s

19:26:06 <balbinus> but i don't think everyone will think of handling all possible cases in a not RDF-oriented language as XSLT..

19:26:31 <chaalsBRS> True. That's why most people shouldn't use XSLT to parse RDF...

19:26:35 <crschmidt> XSLT shouldn't ... right.

19:26:42 <balbinus> i know :)

19:27:05 <Talliesin> Unless you have a strong use-case where a consumer will both a. Not be able to make that entailment, and b. Care! you should just use the the most specific class

19:27:19 <Talliesin> One case where I do see some value would be in the case of:

19:27:28 <balbinus> but, there's no RDF equivalent to XSLT

19:28:05 <chaalsBRS> Well, there are a bunch of them, unless the equivalence you are looking for is a standardised declarative way of doing things with general implementation across toolsets...

19:28:48 <chaalsBRS> Hera actually does something ike what you are suggesting, to cope with tools that don't understand rdfs:sub*

19:28:59 <Talliesin> <ex:Employee><rdf:type rdf:resource="http://xmlns.com/foaf/0.1/Person"/></ex:Employee> if it was forseen that the data would be seen my software that doesn't do entailment and which understands the common FOAF vocab but not the rarer ex vocab.

19:29:43 <chaalsBRS> We say that a result has r:type &earl;fail and r:type &sidar;partial - having declared that sidar:partial s:subClass earl:fail

19:29:57 * Talliesin has termed this "Poor-man's entailment" in a desperate attempt to leave the technical lexicon with something of his coining ;)

19:30:12 <balbinus> :)

19:30:23 <chaalsBRS> just the sort of thing Talliesin is talking about. It's about giving people excuses for not implementing proper systems, and is a Bad Idea (TM)

19:30:30 <chaalsBRS> (IMHO :-)

19:30:33 <balbinus> ok :)

19:30:49 <Talliesin> Another type of Poor-man's entailment is to query twice when one query + entailment should be enough.

19:30:53 <balbinus> thx Talliesin, chaalsBRS :)

19:31:34 <Talliesin> Yes. It can lead to very bad things happening. It's a technique to put on a high shelf with a childproof lid on.

19:31:51 <balbinus> :D

19:32:31 <crschmidt> mmm, childproof lids.

19:32:37 <Talliesin> Remember though, that using rdf:type to say that something is an instance of two orthogonal classes is perfectly valid.

19:32:54 <balbinus> yep

19:51:45 <chaalsBRS> jibberjim?

19:52:01 <chaalsBRS> is your worldmap mercator projection?

19:52:09 <chaalsBRS> (the svg one)

19:56:42 <JibberJim> er, dunno.

20:03:26 <chaalsBRS> Fair enough :-)

21:48:41 * Chopinhauer is away: Good night

23:39:13 <ows> night