Semantic Web Interest Group IRC Chat Logs for 2004-08-26

00:32:42 <Talliesin_> Talliesin_ is now known as Talliesin

00:40:12 <Talliesin_> Talliesin_ is now known as Talliesin

01:15:54 <MarkB> hmmm... REST 0 isn't REST

01:15:59 <MarkB> oopsie, wrong channel

01:38:01 <jsled> [http://asynchronous.org/blog/archives/2004/08/25/levels_of_rest.html if anyone cares about the referrant of that outburst, [and hopefully a bit more link-love for me ;)]]

01:38:12 <ndw> Hey MarkB :-)

02:07:20 * MarkB waves to ndw, runs back to #rest

02:24:17 * sanctius Bonjour tout le monde / Good Morning all

03:21:29 <Ant_Prime> PARTY CHANNEL!

03:51:11 * logger is already logging

07:24:01 <Florent> hi all

08:28:07 <balbinus_> balbinus_ is now known as balbinus

09:13:41 <balbinus> hi

09:14:36 <balbinus> how can i indicate that there are more than one language in a rdfs:seeAlso? (i'm currently using xml:lang to indicate the "language of link")

09:17:27 <dajobe> xml:lang is about the language of literals

09:17:38 <dajobe> not about the language of a resource such as a web page

09:17:43 <dajobe> for that you could use the DC term dc:language

09:18:04 <balbinus> ok... thanks :)

09:18:22 <dajobe> this is one way to do it <rdfs:seeAlso dc:language="fr" rdf:resource="http://otherplace/" />

09:18:29 <balbinus> yep

09:18:32 <dajobe> which says that the otherplace has dc:language=fr

09:18:40 <balbinus> another one is using a foaf:Document...

09:18:56 <dajobe> that's a class, rdfs:seeAlso and dc:language are properties

09:19:03 <balbinus> yes

09:19:12 <dajobe> so foaf:Document can be added additionally, but doesn't say anything morea bout language

09:19:41 <balbinus> but you can do <rdfs:seeAlso><foaf:Document><dc:language>fr</dc:language></foaf:Document></rdfs:seeAlso>

09:20:12 <dajobe> yes

09:20:29 <dajobe> but I expect you'd want a URI for the foaf:Document rdf:about="..."

09:20:43 <balbinus> yes :) forgot it...

09:20:48 <dajobe> looks good

09:21:14 <balbinus> btw, what class can i use to define more precisely a dc:subject?

09:21:45 <dajobe> it's quite open. there are lots of classification hierarchies

09:21:58 <dajobe> or you can stick to keywords/literals

09:22:04 <balbinus> yes

09:22:22 <balbinus> any class for a keyword or category?

09:22:43 <dajobe> there are classes in wordnet to represent words

09:22:51 <dajobe> but literal strings are not classes

09:22:57 <dajobe> unless you get pedantic

09:24:00 <balbinus> ok... thanks :)

09:42:11 <Code_Poet> Code_Poet is now known as Talliesin

11:14:22 <bblfish> bblfish is now known as bblfish_lunch

13:55:19 <GregElin> monring

13:55:36 <balbinus> mroinng

13:55:56 <ndw> morinng

13:56:20 <Talliesin> mornnig

13:58:51 * balbinus counts 25 possibilities (w/o moving "m" and "g")... is he right?

14:02:10 <ndw> Uhm. I get 60, assuming the two n's are indistinguishable

14:03:47 <balbinus> i get 120 :)

14:04:00 <crschmidt> hm. i got 96, but I assume that I did something wrong. (I was thinking 4*4*3*2*1)

14:04:29 <ndw> I got 120 then tried to finesse the fact that the two n's aren't distinct

14:04:34 <ndw> 96 sounds better to me

14:04:41 <balbinus> ok

14:04:50 <balbinus> forgot the two "n"'s

14:04:57 * balbinus wonders how he got 25???

14:05:16 <ndw> 5*5 would be my guess. :-)

14:07:30 <balbinus> yep

14:07:35 <balbinus> probably

14:07:44 * balbinus should do less things at once

14:11:04 <ndw> Hmmm. I get 60

14:11:39 <ndw> minnorg

14:11:39 <ndw> minnrog

14:11:39 <ndw> minonrg

14:11:39 <ndw> minorng

14:11:39 <ndw> minrnog

14:11:40 <ndw> minrong

14:11:44 <ndw> mionnrg

14:11:46 <ndw> mionrng

14:11:48 <ndw> miornng

14:11:50 <ndw> mirnnog

14:11:56 <ndw> mirnong

14:11:58 <ndw> mironng

14:12:00 <ndw> mninorg

14:12:02 <ndw> mninrog

14:12:04 <ndw> mnionrg

14:12:06 <ndw> mniorng

14:12:08 <ndw> mnirnog

14:12:10 <ndw> mnirong

14:12:14 <ndw> mnniorg

14:12:16 <ndw> mnnirog

14:12:18 <ndw> mnnoirg

14:12:18 <Talliesin> Stop. Stop right now.

14:12:20 <ndw> mnnorig

14:12:22 <ndw> mnnriog

14:12:24 <ndw> mnnroig

14:12:26 <ndw> mnoinrg

14:12:28 <ndw> mnoirng

14:12:30 <crschmidt> I'm assuming it was probably accidental.

14:12:30 <ndw> mnonirg

14:12:32 <ndw> mnonrig

14:12:34 <ndw> mnoring

14:12:36 <ndw> mnornig

14:12:36 <balbinus> well...

14:12:38 <ndw> mnrinog

14:12:41 <ndw> mnriong

14:12:44 <ndw> mnrniog

14:12:45 <balbinus> interesting logs :)

14:12:46 <ndw> mnrnoig

14:12:48 <ndw> mnroing

14:12:48 * crschmidt ignores ndw for the next minute or two.

14:12:50 <ndw> mnronig

14:12:54 <ndw> moinnrg

14:12:56 <ndw> moinrng

14:12:58 <ndw> moirnng

14:13:00 <ndw> moninrg

14:13:01 <Talliesin> Argh, darn this freenode habbit of not having operators ready with the kick

14:13:02 <ndw> monirng

14:13:04 <ndw> monnirg

14:13:06 <ndw> monnrig

14:13:08 <ndw> monring

14:13:10 <ndw> monrnig

14:13:14 <ndw> morinng

14:13:16 <ndw> morning

14:13:18 <ndw> mornnig

14:13:19 <sbp`> ooh, a moninrg!

14:13:20 <ndw> mrinnog

14:13:21 <dajobe> operators are not so fun

14:13:22 <ndw> mrinong

14:13:24 <ndw> mrionng

14:13:26 <ndw> mrninog

14:13:28 <ndw> mrniong

14:13:30 <ndw> mrnniog

14:13:32 <ndw> mrnnoig

14:13:34 <ndw> mrnoing

14:13:35 * GregElin looks for an IRC client with a spell checker for future use

14:13:36 <ndw> mrnonig

14:13:38 <ndw> mroinng

14:13:39 <sbp`> the +q would probably be more relevant here

14:13:40 <ndw> mroning

14:13:44 <ndw> mronnig

14:13:46 <ndw> Though I suppose 59 is a better answer, surely "morning" doesn't count :-)

14:13:48 <ndw> Oh, c'mon, that was five minutes of pure fun.

14:13:50 <ndw> Heh.

14:13:51 <crschmidt> #rdfig: Contact: AaronSw, last seen: 9 weeks 6 days (15h 39m 43s) ago

14:13:52 <dajobe> :)

14:13:56 <crschmidt> heh

14:14:10 * ndw goes back to building DocBook V4.4b2. Pphhhphttpbbt!

14:14:22 <balbinus> GregElin: lol :)

14:19:23 <balbinus>http://www.bonjourlesmouettes.org/morning.txt

14:19:24 <dc_rdfig> A: http://www.bonjourlesmouettes.org/morning.txt from balbinus

14:19:53 <balbinus> A:|All the answers to the "monring" contest

14:19:53 <dc_rdfig> Titled item A.

14:20:01 <balbinus> A:Launched by GregElin

14:20:02 <dc_rdfig> Added comment A1.

14:20:14 <balbinus> A:Thanks to ndw for the answers

14:20:14 <dc_rdfig> Added comment A2.

14:20:24 * ndw chuckles

14:20:29 <ndw> I guess I don't have to blog it now :-)

14:20:48 <balbinus> A::)

14:20:48 <dc_rdfig> Added comment A3.

14:20:54 <balbinus> A3:""

14:20:54 <dc_rdfig> Deleted comment A3.

14:20:57 <balbinus> aoh

14:20:59 <GregElin> Thank you, ndw. Ever thought of running a domain naming service?

14:21:08 <balbinus> lol

14:22:13 <balbinus> A:Also of interest, [chat logs|http://www.ilrt.bris.ac.uk/discover/chatlogs/rdfig/2004-08-26.html#T14-11-04] during the contest

14:22:14 <dc_rdfig> Added comment A3.

14:23:19 <balbinus> A3:Also of interest, [chat logs|http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-08-26.html#T14-11-04] during the contest

14:23:19 <dc_rdfig> Replaced comment A3.

14:25:23 <ndw> Wow. I got all 60 answers joined together. I didn't get the comments interspersed.

14:25:32 <ndw> I guess my lag is 0 to me. heh.

14:27:16 * balbinus deleted all comments by hand from the logs to generate the file... to lazy to code sth :)

14:27:34 <crschmidt> Clients will often space out the sending of many lines.

14:27:47 <crschmidt> if it all got sent at once, you'd have been kicked off the network with Excess Flood

14:28:00 <balbinus> yep

14:28:01 * ndw apologizes for behavior that was probably a little rude.

14:28:14 * balbinus is interested in ndw's algorithm :)

14:28:25 <ndw> A crude perl script.

14:29:55 * ndw blogs it after all

14:30:09 * balbinus too :)

14:34:55 <dajobe>http://www.w3.org/TR/2004/CR-xop10-20040826/

14:34:56 <dc_rdfig> B: http://www.w3.org/TR/2004/CR-xop10-20040826/ from dajobe

14:35:04 <dajobe> B:|XML-binary Optimized Packaging - w3c candidate rec

14:35:05 <dc_rdfig> Titled item B.

14:35:14 <dajobe> B:"XOP ... a means of more efficiently serializing XML Infosets that have certain types of content."

14:35:15 <dc_rdfig> Added comment B1.

14:35:18 <dajobe> B:doesn't say what types

14:35:19 <dc_rdfig> Added comment B2.

14:35:23 <dajobe> B:anyone know what this is?

14:35:23 <dc_rdfig> Added comment B3.

14:35:34 <dajobe> B3:anyone know what this is for?

14:35:35 <dc_rdfig> Replaced comment B3.

14:37:09 <dajobe> B:or should I care?

14:37:10 <dc_rdfig> Added comment B4.

14:37:55 <dajobe> B:seems to be a way of doing XML incl binary bits into mime/multipart XML + references to attached binary bits

14:37:56 <dc_rdfig> Added comment B5.

14:39:51 <dajobe> B:as used by [http://www.w3.org/TR/2004/CR-soap12-mtom-20040826/|SOAP message optimization]

14:39:52 <dc_rdfig> Added comment B6.

14:39:58 <dajobe> B:I'm tending towards "I don't care" now :)

14:39:59 <dc_rdfig> Added comment B7.

14:40:08 <ndw> It's a binary packaging thing

14:40:13 <dajobe> yeah

14:40:21 <ndw> Designed to get binary blobs into the infoset so that SOAP can process them

14:40:25 <dajobe> I've a backburner musing about binary "rdf infosets"

14:40:38 <dajobe> since a usecase came up, someone doing rdf on phones

14:40:43 <dajobe> ndw: yeah

14:40:51 <dajobe> ... phones with no xml parser

14:40:56 <ndw> "We have an abstraction. Everything will fit in our abstraction! Or else!"

14:41:07 * dajobe applies the xml hammer

14:41:27 <Talliesin> Why would a phone with no xml parser have a binary RDF parser?

14:41:41 <dajobe> it woudl have a non-xml thing

14:41:56 <dajobe> since an xml parser would be too slow and/or too large to add or download

14:42:00 <dajobe> if it wasn't built in

14:42:25 <dajobe> but optimisation of B: is for something else

14:44:08 <Talliesin> Why not n-triples then?

14:44:21 <dajobe> that's what they used

14:45:35 <Talliesin> Yay! I hit to eventual solution without really thinking about it. It's like finding Waldo that is.

14:46:15 <dajobe> I guess I better not say any more, or ASN.1 will be mentioned.

14:46:16 <dajobe> oops

14:46:42 * Talliesin puts his fingers in his ears, closes his eyes and counts to a thousand.

14:48:21 <jeen> mattb, you here?

15:05:18 * ndw points balbinus to http://norman.walsh.name/2004/08/26/goodMorning and then promises to *let it go* :-)

15:06:28 <jeen> omg

15:06:56 <libby> eh

15:06:58 <libby> heh

15:06:59 * balbinus points ndw to http://bonjourlesmouettes.org/index.php/2004/08/26/98-monring-contest and do not promise anything :)

15:08:07 * ndw wonders, is une mouette a seagull?

15:08:22 * balbinus answers "yes"

15:09:18 <crschmidt> ndw: what package do you use for blogging?

15:10:19 <crschmidt> or is it something you wrote yourself?

15:10:24 <ndw> Ha. My own bailing wire and duct tape and RDF and XML and DocBook and ... stuff: http://norman.walsh.name/2003/05/14/how

15:10:30 * crschmidt nods

15:10:51 * ndw points crschmidt to http://norman.walsh.name/Makefile and invites him to "have at it!"

15:10:55 <crschmidt> i'd never seen anything else offer a link to a .pdf, so i guessed as much

15:11:23 <balbinus> wow

15:23:02 <crschmidt> ndw: i'm quite happy with wordpress, thanks :)

15:24:37 <ndw> :-)

15:26:11 * balbinus thinks of using XML for DOAP-fr...

15:28:49 <balbinus> but i'm a little bit afraid of the time i'll need to do it :)

16:05:57 <ndw> ndw is now known as ndw^phone

16:10:17 <inkel> inkel is now known as inkel|lunch

16:17:39 <mattb> jeen: you were looking for me earlier?

16:18:00 <jeen> hi matt, yeah that's right

16:18:09 <jeen> that scutter of yours, are you still actively developing that?

16:18:20 <mattb> not very actively, no

16:18:24 <mattb> the java or the python one, by the way?

16:18:28 <jeen> java :)

16:18:35 <mattb> you should ask aharth about that

16:18:45 <mattb> he picked it up and made it into something better

16:19:08 <jeen> reason I ask: a colleague of mine has adapted your scutter to work with Sesame instead of Jena, and we're considering distributing it as part of Sesame.

16:19:42 <mattb> oh cool

16:19:58 <mattb> do you need a statement of licensing for the original codebase?

16:20:13 <bblfish_lunch> bblfish_lunch is now known as bblfish

16:20:22 <jeen> well nothing really formal, really, just yet anyway, just wanted to know if you were ok with it.

16:20:48 <jeen> in principle

16:21:16 <jeen> is it GPL?

16:21:22 <mattb> in principle

16:21:28 <jeen> heh

16:21:34 <mattb> i need to make a blanket statement of licensing on my site

16:21:44 <bparsia> bparsia is now known as bijan

16:21:52 <mattb> everything up there is there for people to pick up and play with and make better things from

16:22:04 <jeen> that's great

16:22:04 <mattb> i claim no ownership

16:22:11 <mattb> credit is always nice :)

16:22:20 <jeen> of course, that goes without saying :)

16:22:37 <mattb> let me know if/when you need anything more concrete than these irc logs

16:22:49 <jeen> I have to leave now, but it's good to know this. I'll be in touch about it ok? thanks

16:23:24 <jeen> seeya

17:15:21 <mchncl> iä iä

17:43:46 <ndw^phone> ndw^phone is now known as ndw

17:44:02 * ndw nudges DanC

17:44:30 <ndw> You there, DanC?

17:47:19 <DanC> about IETF/MIME foo, you say?

17:47:57 <ndw> I do.

18:40:42 <IRCMonkey> test

18:41:01 <GregElin> IRC IS WORKING.

18:42:19 <inkel|lunch> inkel|lunch is now known as inkel

19:40:34 <DanC> epydoc... yum!

20:01:08 <ndw> epydoc?

20:02:16 <ndw> .g epydoc

20:02:19 <phenny> epydoc: http://epydoc.sourceforge.net/

20:20:42 <rawles> shot in the dark: does anyone use swi-prolog for parsing and loading rdf files?

20:21:13 <crschmidt> .g prolog rdf

20:21:16 <phenny> prolog rdf: http://www.xml.com/pub/a/2001/07/25/prologrdf.html

20:21:39 <crschmidt> So, it sounds like bijan might be

20:21:46 <crschmidt> !namefornick bijan

20:21:48 <wh4experimental> bijan Bijan Parsia

20:21:49 <wh4experimental> hBijan Bijan Parsia

20:21:57 <crschmidt> (good, they are the same person)

20:22:51 <rawles> oh this is a nice article

20:25:16 <rawles> i had a specific question but i'll read the article first

20:25:47 <crschmidt> bijan stops by here at least occasionally, i think

20:27:15 <rawles> that would be cool

21:09:03 <sandroMIT> rawles, I used swi-prolog for RDF, lots and lots.

21:09:26 <sandroMIT> (as you can tell by my messages in the swi-prolog mailing list archives and bug reports.)

21:11:34 <rawles> fantastic

21:11:47 <rawles> do you have 5 minutes to help me, i'd be pulling my hair out if i wasnt bald.

21:14:47 <CLoCkWeRX[a]> CLoCkWeRX[a] is now known as CloCkWeRX

21:14:57 <rawles> sandro are you there?

21:48:12 <lyle> lyle is now known as lyle|away

22:04:33 <sandroMIT> sorry, rawles, in another window.... :-(

22:04:57 <sh1mmer> hey sandroMIT :)

22:04:59 <sh1mmer> ltns

22:05:03 <sh1mmer> got a min?

22:05:49 <sandroMIT> mmmm, maybe.

22:06:42 <sh1mmer> I am going to propose doing my diseration on composite IFPs or "unamibigous property constellations"

22:06:44 <sh1mmer> any thoughts?

22:07:14 <JibberJim> hmm, not much SVG in it...

22:07:20 <mortenf> heh

22:07:25 <sandroMIT> sounds like a useful thing to have, but I know nothing about the kinds of theory work that would really involve.....

22:07:50 <sandroMIT> (I'm an implement-it-and-see-if-it-works guy, which doesn't get PhD's.)

22:07:53 <sh1mmer> ok, it's just Sean recons you could the term

22:08:03 <sh1mmer> this isn't a PhD. It's for my undergrad

22:08:21 <sandroMIT> Ah. Then an implementation would probably be good enough.

22:08:42 * sh1mmer got told of a bit for trying to pick topics none of the examiners would be able to mark (or possibly understand :] )

22:08:52 <sandroMIT> Of course it's not called UnambiguousProperty any more; now it's owl:InverseFunctionalProperty.

22:09:06 <sh1mmer> indeed

22:09:06 <sandroMIT> Have you though about how you'd express what you want in RDF?

22:09:14 <sh1mmer> sure

22:09:18 <sh1mmer> Sean did actually

22:09:24 * sh1mmer finds

22:09:44 <crschmidt> heh

22:10:22 <sh1mmer> (:bankAccountNumber :name) rdf:type ont:UnambiguousPropConst .

22:10:30 <sh1mmer> pretty simple.

22:10:39 <sh1mmer> I have a few use cases.

22:10:48 * mortenf has a few as well

22:10:55 <sh1mmer> Smushing, negociation, and optimisation of data stores.

22:11:09 <sh1mmer> mortenf oh?

22:11:18 <mortenf> hmm, doesn't sound like *use* cases :)

22:11:30 <mortenf> yeah...

22:11:36 <sh1mmer> it doesn't?

22:11:46 * sh1mmer lives for brevity

22:12:04 <mortenf> well, which goals are you trying to meet?

22:12:29 <mortenf> i mean, it sounds about right, but are they *use* cases?

22:13:05 <sh1mmer> sure.

22:14:23 <mortenf> anyway, my use cases: identifying online accounts (see foaf:holdsAccount) and waypoints (local device "namespace" for id's)

22:14:26 <sandroMIT> How is it different than saying some property is (inverse)functional, and the property's value is always a list of the values of certain other properties? Same meaning, right?

22:14:27 <sh1mmer> If you know the possible CIFPs in a system then it makes smushing a lot easier and definitive, it makes negociation (ie smushing in negociation) easier

22:14:55 <sh1mmer> optimising data stores is about self-organising by the CIFPs to make retrieval of unique stuff faster and more efficient.

22:15:13 <mortenf> sounds interesting

22:15:35 <sh1mmer> sandroMIT I guess so.

22:16:30 <sh1mmer> sandroMIT I like the idea of the group having a URI too, that useful.

22:24:58 <sh1mmer> well that conversation died.

22:25:26 <sh1mmer> mortenf what do you mean by waypoints?

22:26:00 <mortenf> gps

22:26:31 <sh1mmer> I don't follow

22:26:39 <sh1mmer> Can you expand on what you mean a bit

22:26:56 <mortenf> in my gps (geko 201) i can mark waypoints, e.g. where i am now, with an id and a symbol

22:27:19 <mortenf> however, the id has to be unique (for the device) and can only be 6 chars

22:27:50 <mortenf> i'd like to be able to keep a (somewhat) separate storage with further descriptions of each waypoint, and also share them

22:28:03 <sh1mmer> ok, and how would CIFPs (as I like to call them unless someone has a better name) help?

22:28:10 <mortenf> for that, i need to identify them, which would be by device + id

22:28:13 <sandroMIT> use case: (id for my camera, the day the picture was taken, and the filename the camera assigns the picture) -- unambiguous constellation.

22:28:25 <mortenf> right

22:28:26 <sh1mmer> ah

22:28:27 <sh1mmer> :)

22:28:29 <sh1mmer> gotcha

22:29:00 <mortenf> (of course, i could construct a uri for each waypoint with the id, but...)

22:29:56 <sh1mmer> yeah thats what I was going to suggest at first, but it's interesting

22:30:36 <mortenf> the foaf:holdAccount would be more interesting in the short term, since danbri has all but accepted making it an IFP

22:30:47 <mortenf> erh, foaf:holdsAccount

22:31:48 <sh1mmer> right, I use that for my SF account

23:01:27 <sandroMIT> but if that's an IFP it's not interesting for this, right?

23:01:51 <sh1mmer> heh, no.

23:02:09 <sandroMIT> because several people can share an account?

23:02:47 <sh1mmer> well same goes for an email, its an articificial constraint.

23:03:50 <mortenf> it's interesting, because how do you identify an account?

23:04:05 <mortenf> answer: by accountServiceHomepage + accountName