Semantic Web Interest Group IRC Chat Logs for 2002-04-16

00:04:06 <mnot> sandro: consensus doesn't work that way (re: authors on an RFC). Then again, consensus would have it that a solution acceptable to all would surface; that doesn't seem to be happening yet.

00:33:04 <danbri> http://www.schafercorp-ballston.com/spawar_meeting2002/

00:33:04 <dc_rdfig> A: http://www.schafercorp-ballston.com/spawar_meeting2002/ from danbri

00:33:49 <danbri> A:|DARPA Semantic Web for Military Users, Joint Meeting

00:33:49 <dc_rdfig> titled item A

00:35:09 <danbri> A:"There is a collective vision of a future warfare environment that describes how agent-based systems and the semantic web may provide the war fighter with actionable information and a means to execute responses in dramatically improved terms. "

00:35:09 <dc_rdfig> added comment A1

00:35:48 <AaronSw> Hm, my inner pacifist is reconsidering my work on the Semantic Web. ;-)

00:35:55 <danbri> A:I wonder if they mean the _American_ war fighter? It being a world-wide Web, and all... Or just fighters generally...?

00:35:55 <dc_rdfig> added comment A2

00:37:09 <danbri> A:"The DAML program, with new program management and a new DARPA Office providing context, is recasting its future in nearer term operationally significant context."

00:37:10 <dc_rdfig> added comment A3

00:37:48 <danbri> hi jim

00:38:36 <jhendler1> hi dan

00:39:05 <bijan> Hmmm. My new motto:

00:39:16 <bijan> Building the semantic web, one argument at a time.

00:39:23 <bijan> Works on several different levels :)

00:41:21 <danbri> A:See also [http://orlando.drc.com/daml/Ontology/Fugitive/3.1/Fugitive-ont.daml|fugitive ontology]

00:41:21 <dc_rdfig> added comment A4

00:41:35 <sandro> good slogan, Bijan. :-)

00:41:52 <bijan> In general, or for me *personally*, sandro?

00:42:00 <bijan> :)

00:44:23 <jhendler1> danbri - I simply want to assert, with respect to "A:" that I'm happy that

00:44:25 <jhendler1> "The DAML program, with new program management and a new DARPA Office providing context, is recasting its future in nearer term operationally significant context."

00:44:42 <jhendler1> makes it clear that it is NEW program management...

00:48:35 * danbri happy about that too

00:48:51 <AaronSw> why?

00:49:14 <AaronSw> don't want to tarnish your reputation as the pacifist military employee?

00:50:31 <jhendler1> Aaron - pacificst military employee - I like that, should have had it on my business cards

00:50:54 <AaronSw> I've decided that my title is either "Meddling Kid" or "Outer Child".

00:51:11 * danbri wonders if some DARPA RDF database knows that I was once a member of CND...

00:51:11 <bijan> Meddling kid

00:51:14 <bijan> No question.

00:51:21 <jhendler1> CND?

00:51:24 <AaronSw> Yeah, that's what I figured, bijan.

00:51:25 <bijan> All agree.

00:51:35 <jhendler1> I like Outer Child.

00:51:38 <bijan> Indeed, I have that tattooed on my kneecap.

00:51:39 <danbri> Campaign for Nuclear Disarmament

00:51:46 <bijan> "Aaron is a Meddling Kid"

00:51:54 <bijan> Though I actually have "meddlesome"

00:51:57 <danbri> Didn't keep up my subscription after that cold war thing petered out...

00:52:04 <bijan> As in, "Will no one rid me of this meddlesome kid?"

00:52:04 <bijan> :)

00:52:13 <jhendler1> Bijan has "Aaron is a meddling kid" tatooed on his kneecap??

00:52:18 <AaronSw> we can break some kneecaps

00:52:49 <jhendler1> danbri - they didn't seem to mind my Vietnam activities

00:53:00 <AaronSw> danbri a :EnemyOfTheState .

00:53:46 * danbri wonders what your vietnam activities were

00:53:53 <bijan> jim, don't point them to monkeyfist. :)

00:54:03 <jhendler1> of course, I was actually refering more to the emphasis of the DAML program -- mine was on the semantic web for everyone.

00:54:20 <bijan> (Having just received another round of hate mail)

00:54:38 <AaronSw> MonkeyFist: Our Monkey, DARPA's Fist

00:54:42 * jhendler1 it's a good thing the web wasn't around in my vietnam days, I don't have to admit it...

00:54:46 <danbri> for monkeyfist? from primatologists?

00:54:51 <bijan> We Looney Leftist Fifth Columnists must be careful.

00:55:05 <bijan> No, lately it's been homophobes and rather scary racists.

00:55:27 <bijan> Though, espeically with the latter, rather impersonal.

00:55:48 <bijan> When they start dropping hints about Iraqi sympathies and greasy arabs, I start getting nervous.

00:56:05 <bijan> Especially as only I get those particular aspertions, even when I co author :)

00:56:20 <bijan> I been called some interesting things in NC.

00:56:46 <bijan> Someone asked me if I was an Aaaay-raaaab.

00:57:00 <bijan> Another call me a "Damned Ct Yankee"

00:57:03 <AaronSw> Well, the lack of bijan pictures...

00:57:13 <bijan> Which had me rolling in the aisle.

00:57:24 <jhendler1> oh good, we haven't had any fun political debates in my lab for a while.

00:57:39 <bijan> Soon after that I was in philly and someone called me a Damn Cracker.

00:57:49 <jhendler1> used to have two old former hippies and two Rush Limbaugh fans

00:57:55 <bijan> Heh. I'm all for forced redistribution of wealth!

00:58:02 <bijan> *Your* wealth, in particular.

00:58:03 <bijan> :)

00:58:16 <bijan> Though I'd take aaron's too, that damn meddling kid!

00:58:25 <jhendler1> made for fun discussion -- even more so when I hired a Chilean student who was pro-dictatorship...

00:58:32 <AaronSw> I don't have any, I gave it all to the poor.

00:58:33 <bijan> Eek!

00:58:47 <bijan> Well, go get it back.

00:58:54 <bijan> Forcably redistribute it again!

00:59:02 <jhendler1> ahh, Bijan's a republican - he wants to take money from the poor!

00:59:21 <bijan> Hmm. THat's the first time i've *ever* been accused of being a republican! :)

00:59:22 <AaronSw> phew, i was worried there

00:59:32 <AaronSw> i don't need to pay no republican

00:59:36 <bijan> But I want to give it to the *other* poor.

01:00:00 <bijan> If you give it to the poor they have to pay taxes which get back to the rich via defense contracts, see.

01:00:00 <AaronSw> Who?

01:00:09 <bijan> So it's better to get it away from them so it doesn't trickle up!

01:00:38 <AaronSw> bijan's trickle up economics

01:00:41 <bijan> (I'm also fond of generating silly theories, in case that wasn't evident :))

01:00:59 <jhendler1> makes more sense than much of what we see in #rdfig

01:01:08 <bijan> Heheheh.

01:01:11 <bijan> Ouch!

01:01:30 <bijan> Jim's in a puckish mood -- must have gotten his laptop fixed.

01:03:05 <jhendler1> jhendler1 has changed the topic to: Dark Triples and the "Satanic Web" (weblog: http://rdfig.xmlhack.com/)

01:03:51 <jhendler1> got my harddisk transferred into a faster box - so I'm in good shape -- of course, the student expecting the laptop got screwed...

01:04:14 <bijan> Heh.

01:04:34 <AaronSw> No, it just got turned into an Educational Opportunity for him to learn about laptops

01:05:42 <bijan> Hmm. jim's comment about me being repug is killing my rep.

01:05:53 <bijan> All across the net my hate sites are being revamped.

01:06:06 <bijan> Eek! Is that David Horowitz on the phone with a writing gig?!?!?!?

01:06:13 <AaronSw> lol!

01:06:35 <jhendler1> bijan - just want you to have a shot at the new NY Sun.

01:06:52 <bijan> I guess I better start my "how I betrayed all my political principles, causes, and friends" memoir...

01:07:05 <jhendler1> see, you ARE a Republican :->

01:07:08 <AaronSw> Blinded By The Dark

01:07:12 <AaronSw> err, left

01:07:13 <bijan> Heheh.

01:07:17 <bijan> Triples!

01:07:36 <bijan> "Mildly incapacitied by the left"

01:07:49 <bijan> "Mildly impoverished by the left"

01:07:57 <bijan> "Somewhat ignored by everyone"

01:08:14 <jhendler1> ahh, we finally get back to the seamntic web!!

01:08:17 <bijan> Hmm. To complete this, I guess I should join up with the web services bandwagon!

01:08:24 <bijan> Viva SOAP! Viva WDSL!

01:08:40 <bijan> LIBERTE, EGALITE, SOAPALITE!

01:08:46 <jhendler1> time for bijan to revise his GET-7 approach...

01:09:14 <AaronSw> lol @ L, E, S

01:09:34 <AaronSw> Gotta run

01:09:37 <bijan> There's safty in numbers.

01:09:51 <bijan> So the more people who misuse GET, the safer it is.

01:10:22 <jhendler1> g'night Aaron

01:33:43 <AaronSw> back.

03:58:17 <AaronSw> hm. HP employees working on free software being asked to justify their existence: http://www.advogato.org/person/mbp/diary.html?start=197

10:02:18 <JibberJim> Libby, have you stopped looking for the RDF rdfweb files now?

10:03:36 <libby> hey jim, how dyou mean? is it not harvesting?

10:06:26 <JibberJim> I've not seen the avalanche of 404's I used to get (I monitor 404's...)

10:06:32 <libby> heh

10:06:51 <libby> I'll check - I've been away, so probably everything's fdallen over ;)

10:07:02 <JibberJim> Ah, the normal problem of going away...

10:08:08 <libby> yeah

10:38:17 <danbri> http://tmitwww.tm.tue.nl/staff/gwagner/RuleML-BR-SW.html

10:38:18 <dc_rdfig> B: http://tmitwww.tm.tue.nl/staff/gwagner/RuleML-BR-SW.html from danbri

10:38:56 <danbri> B:|workshop on business rule markup languages, 14 June 2002, Sardinia (Italy), ISWC2002

10:38:56 <dc_rdfig> titled item B

12:53:41 <JibberJim> danbri?

12:54:12 <JibberJim> I see you're linking to the screen-saver, if I make changes to it, do you want me to keep the old version as is, or can I update it?

13:44:25 <AaronSw> cool, Jos did run the test cases

13:49:19 <AaronSw> cool, Jos did run the OWL intersection test cases

13:57:39 <DanC> quite.

13:57:57 <DanC> so... er... that example doesn't motivate squat. next?

13:58:34 <AaronSw> this is somewhat more helpful: http://lists.w3.org/Archives/Public/www-archive/2002Apr/0003.html

14:00:31 <DanC> yes, that's quite a clear write-up

14:00:38 * DanC has a telcon in... er.. now

15:30:57 <libby>http://ilrt.org/~ecdb/crp/reports.xml

15:30:57 <dc_rdfig> C: http://ilrt.org/~ecdb/crp/reports.xml from libby

15:31:06 <libby> libby is now known as danb_ilrt

15:31:25 <danb_ilrt> C:|SW application scenarios: Company Accounts

15:31:25 <dc_rdfig> titled item C

15:31:53 <danb_ilrt> C:Idea is to take some demo data from biz/ed site and experiment with it in Cwm, MathML, SVG, RDF-logic...

15:31:53 <dc_rdfig> added comment C1

15:32:15 <danb_ilrt> C:Anyone fancy trying the Cwm Math library with this dataset?

15:32:15 <dc_rdfig> added comment C2

15:32:18 <danb_ilrt> danb_ilrt is now known as libby

15:32:35 <AaronSw> are nuks related to nukes?

15:48:17 * Seth thinks the darkness of triples is relative

16:27:08 <danbri>http://www.w3.org/2002/04/p3p-pressrelease

16:27:09 <dc_rdfig> D: http://www.w3.org/2002/04/p3p-pressrelease from danbri

16:27:17 <danbri> D:|"World Wide Web Consortium Issues P3P 1.0 as a W3C Recommendation"

16:27:17 <dc_rdfig> titled item D

16:28:30 <AaronSw> wait a second... how'd they get out of CR, i thought they needed an rdf schema!

16:28:47 <JibberJim> Still nothing on TRUST wrt to p3p though...

16:28:48 <danbri> D:Nearby, [http://www.w3.org/TR/p3p-rdfschema/|An RDF Schema for P3P], by Brian McBride, Rigo Wenning, Lorrie Cranor.

16:28:48 <dc_rdfig> added comment D1

16:29:10 <AaronSw> I'd figured that the rdf schema requirement would keep them stuck forever!

16:34:39 <AaronSw> Hm. Why is my calendar telling me there's an RDFCore F2F tomorrow in Bristol?

17:14:44 <JibberJim> danbri_?

17:16:18 <JibberJim> - Another alternative for dsiplaying the FOAF region data http://jibbering.com/rdf/foafwhoss3.1

17:28:05 <Seth>http://robustai.net/mentography/annotation.gif

17:28:05 <dc_rdfig> E: http://robustai.net/mentography/annotation.gif from Seth

17:28:31 <Seth> E: a mentograph of annotation

17:28:31 <dc_rdfig> added comment E1

17:28:53 <Seth> E1:''

17:28:53 <dc_rdfig> replaced comment E1

17:29:19 <Seth> E:|A mentograph of annotation

17:29:20 <dc_rdfig> titled item E

17:30:13 <Seth> E: in response to [http://lists.w3.org/Archives/Public/www-rdf-interest/2002Apr/0180.html|Sean's post]

17:30:13 <dc_rdfig> added comment E2

17:33:11 <Seth> E1: introduces green nodes who's identity is a statement id

17:33:11 <dc_rdfig> replaced comment E1

17:37:37 <Seth> E: this technique allows one graph to have an attitude twards another graph

17:37:37 <dc_rdfig> added comment E3

17:52:52 * chaals waves to niq

17:53:04 <niq> hi

18:09:36 <MysticOne> [GlobalNotice] Hello everyone! I hope everyone's day is going well. I just wanted to remind everyone that we're still accepting artwork for our logo and t-shirt campaign. Questions or submissions may be sent to mysticone@openprojects.net, or you may contact me here with the nick MysticOne. Any help you can give us is much appreciate, and as always we thank you for using the Open Projects Network.

18:09:57 <danbri_> C:Oops, the main URL I meant to chump was the [http://ilrt.org/~ecdb/crp/semweb-scenario.html|html writeup]. I generated some schema fragments...

18:09:57 <dc_rdfig> added comment C3

18:10:54 <AaronSw> how many ilrt accounts do you have?

18:11:03 <AaronSw> did you encroll in economics too?

18:11:29 <danbri_> ILRT accounts: 7 years worth...

18:11:35 * chaals suspects aaron of snooping around the semantic web

18:12:15 <danbri_> ILRT was merged from the Educational Technology Service (ETS) and the Centre for Computing in the Social Science. Hence foaf:nick of ec* (ecdb, eclm etc).

18:12:34 <AaronSw> educational computing?

18:12:52 <danbri_> that kind of thing, yeah

18:13:22 <danbri_> Aaaron, do you know the Cwm maths stuff well enough to have a look at the company account thing? It's a black art to me...

18:13:50 <AaronSw> i can probably manage it... what do you want to do?

18:15:30 <danbri_> Haven't got that clearly writen up yet, but basically see the XML dump at http://ilrt.org/~ecdb/crp/semweb-scenario.html and find a property (eg. ATR = (Current Assets - Stocks) /Current Liabilities ) and translate the rules into Cwm/N3 form.

18:18:57 <MysticOne> [GlobalNotice] Just to clarify a few things. We're looking for a logo that emphasis community involvement and work, not just software development. Also, high resolution/dpi is desired if you're submitting a final proof, but concept sketches are fine too. Just e-mail submissions to mysticone@openprojects.net. Thank you!

18:24:34 <AaronSw> _:spe = [math:divide (?sr ?noe)

18:24:36 <AaronSw> ].

18:26:47 <AaronSw> { ?corp :sr ?sr . ?corp :noe ?noe . } => { ?corp :spe _:spe . _:spe math:quotientOf (?sr ?noe)} .

18:28:23 <AaronSw> hm, that'd work except the data has no noe, but does have an spe

18:29:06 <AaronSw> so noe = sr / spe

18:30:05 <AaronSw> { ?corp :sr ?sr . ?corp :spe ?spe . } => { ?corp :noe _:noe . _:noe math:quotientOf (?sr ?spe)} .

18:34:28 <AaronSw> hm, doesn't seem to work. you'll have to ask sbp

18:48:15 <danbri_> Thanks for trying. I haven't got my head around the various fields yet...

19:01:38 <danbri_>http://www.w3.org/2002/Talks/0411-ws-f2f-sweb/

19:01:39 <dc_rdfig> F: http://www.w3.org/2002/Talks/0411-ws-f2f-sweb/ from danbri_

19:02:02 <danbri_> F:|Web Service Face to Face - Aligning With the Semantic Web, Eric Prud'hommeaux & David Booth (April 8-12, 2002)

19:02:02 <dc_rdfig> titled item F

19:02:16 <danbri_> F:From recent W3C Web Services WG Face-to-Face Meetings

19:02:16 <dc_rdfig> added comment F1

19:13:09 <AaronSw> hm: http://swada.w3.org/

19:15:33 <AaronSw>http://www.mindswap.org/~rreck/pp_overview.shtml

19:15:33 <dc_rdfig> G: http://www.mindswap.org/~rreck/pp_overview.shtml from AaronSw

19:15:54 <AaronSw> G:|Status of Parka Database Software for RDF

19:15:54 <dc_rdfig> titled item G

19:16:25 <rreck> hey where did THAT come from ?

19:16:31 <AaronSw> *cough*

19:17:22 <AaronSw> ooh, Apache 2.0. really should have used http://www.mindswap.org/~rreck/pp_overview

19:17:59 <rreck> ok, ill look at it

19:18:05 <AaronSw> G:Part of the Semantic Web Research Group at [MIND SWAP|http://www.mindswap.org/]

19:18:05 <dc_rdfig> added comment G1

19:19:01 <AaronSw> G:They've got [a CVS repository|http://iama.rrecktek.com/cgi-bin/cvsweb.cgi/] too

19:19:01 <dc_rdfig> added comment G2

19:43:08 <MarkB> F:aka "just merge the graphs!" (inside WSAWG joke 8-)

19:43:08 <dc_rdfig> added comment F2

20:12:35 <AaronSw> hi sbp. wanna take a look at danbri's problem?

20:13:58 <sbp> sure

20:14:05 <danbri_> which, I have so many!

20:14:18 <danbri_> this would be my mathml / cwm / rules problem...

20:14:20 <AaronSw> i was thinking of the cwm math one

20:14:22 <danbri_> see above...

20:14:23 <danbri_> :)

20:14:28 <AaronSw> jinx

20:15:09 <danbri_> sbp, is the problem (or dataset / opportunity) reasonably clear?

20:16:47 <sbp> not really

20:16:56 <sbp> and I've read the Webpage, too

20:17:09 <sbp> i.e. http://ilrt.org/~ecdb/crp/semweb-scenario.html

20:17:17 <danbri_> yup, that's the page.

20:17:32 <danbri_> So this was in 1995, when I was working on my first website job, http://bized.ac.uk/

20:17:54 <danbri_> ...we had a 'company report profiler', which allowed people to ask for companies/years/fields for company data, and get HTML or CSV data back.

20:18:13 <danbri_> Some of the fields were in CSV files on the server, others were computed using the perl functions shown in that page.

20:18:32 <danbri_> eg IC: Interest Cover = Operating Profit / Interest

20:18:54 <sbp> O.K., that's simple enough

20:18:56 <danbri_> If we know 'Interest' and 'Operating Profit' for some foafcorp:CompanyReport, we can figure out its foafcorp:ic property

20:19:00 <danbri_> from an RDFish pov

20:19:46 <danbri_> see http://ilrt.org/~ecdb/crp/schema-content.rdf for schema fragment and http://ilrt.org/~ecdb/crp/reports.xml for schema and instance data

20:20:02 <danbri_> I *think* the instances also contain values for the computable fields

20:20:19 <danbri_> I thought it was an interesting dataset anyhow

20:20:38 <sbp> so you'll want stuff like { ?x :op ?y; :i ?z . ?x math:quotient (?y ?z) } log:implies { ?x :ic ?x } . I guess?

20:21:02 <AaronSw> hey, that's what i tried!

20:21:10 <AaronSw> it didn't work, tho

20:21:15 <sbp> ugh

20:21:47 <sbp> hang on, let me run it

20:22:06 <sbp> (I borked it a bit anyway)

20:22:09 <AaronSw> it was: { ?corp :sr ?sr . ?corp :spe ?spe . ?noe math:quotientOf (?sr ?spe)} => { ?corp :noe ?noe . } .

20:23:25 <AaronSw> oh, it worked this time. heh

20:23:43 <AaronSw> heh - i had the namespaces wrong

20:23:51 * danbri_ rummages for his cwm/swap installation

20:23:57 <sbp> you needed: $ n3s -s '{ ?x :op ?y; :i ?z . ?p math:quotientOf (?y ?z) } log:implies { ?x :ic ?p } . :x :op "7"; :i "8"' | cwm --think

20:24:09 <danbri_> :)

20:24:16 <sbp> I got:-

20:24:16 <sbp> :x :i "8";

20:24:16 <sbp> :ic "0.875";

20:24:16 <sbp> :op "7" .

20:24:30 <danbri_> does that match the fields in the raw data?

20:24:37 <AaronSw> yeah, i just forgot to use danbri's namespace

20:24:45 <sbp> aha

20:24:47 <danbri_> cd

20:24:49 <danbri_> doh :)

20:25:19 <AaronSw> hm, got some missing data, danbri

20:25:23 <AaronSw> ValueError: empty string for float()

20:25:46 <danbri_> there are some blanks, I shold remove those properties I guess

20:25:52 <danbri_> ls

20:25:57 <danbri_> doubledoh

20:25:58 <AaronSw> heh! twice!

20:26:24 <danbri_> so do I need to change my data?

20:27:51 * danbri_ checks out 2000/10/swap/

20:28:37 <AaronSw> well, getting rid of the blank properties would be nice, but i guess grep -v can do that

20:29:02 <danbri_> So would moving it elsewhere, I'm spawning too many filespaces...

20:29:19 <danbri_> its just conveniently close to the other biz/ed files (and developers)

20:31:55 <sbp> hooray!

20:31:57 <sbp> $ n3s -s '@prefix default <http://xmlns.com/foaf/corp#> . { ?x sr ?y; spe ?z . ?y log:notEqualTo "" . ?z log:notEqualTo "" . ?p math:quotientOf (?y ?z) } => { ?x noe ?p }' > danbri-test.n3; wincwm --rdf http://ilrt.org/~ecdb/crp/reports.xml --n3 --filter=danbri-test.n3

20:32:04 <sbp> gives a big list of items

20:32:30 <sbp> note the use of notEqualTo... I'm glad that CWM does that builtin before the math ones

20:33:52 <sbp> (that command line has an implicit "some editing of llyn.py later..." before it)

20:34:44 * danbri_ regenerates the data suppreessing blank properties

20:34:53 <danbri_> heheh

20:35:39 * sbp is wondering how many more shameless n3s plugs he can do before midnight

20:36:51 <Seth> sean, that last one was a dozzie !

20:37:39 <danbri_> So cwm-handhold me, if you don't mind. I have checked out 2000/10/swap/, and see cwm.py

20:37:50 <danbri_> danbri@fireball:~/public_html/remote-cvs/w3.org/WWW/2000/10/swap > ./cwm.py

20:37:50 <danbri_> /usr/bin/python: can't open file '/devel/WWW/2000/10/swap/cwm.py'

20:38:13 <danbri_> Is this usual? Or serves me for using non-bundled version? I could just chuck in a symlink I guess...

20:38:27 <danbri_> n3s: looked handy...

20:39:41 <sbp> Seth: doozie: heh, thanks (after checking my understanding of that particular term...)

20:40:09 <sbp> Hmm... that's odd, actually

20:40:18 <AaronSw> danbri, try `python cwm.py`

20:40:57 <danbri_> grr ImportError: No module named Crypto.Util.randpool

20:41:19 <sbp> ooh, that's my fault

20:41:29 <sbp> set "USE_PKC" to 0 in cwm_crypto.py

20:43:11 <danbri_> What's a good trick for having a 'cwm' utility available as a script on my linux box?

20:43:20 <danbri_> is ln -s enough?

20:43:46 <sbp> Hmm... I just put an sh script in my bin

20:43:55 <hBijan> (My trick is to ping sbp and have him run it on his machine)

20:44:01 <hBijan> (I have a little script to do this for me)

20:44:03 <AaronSw> danbri_, alias is good

20:44:11 <AaronSw> (ooh, i want a copy!)

20:44:25 <sbp> heh, heh

20:44:30 <AaronSw> alias cwm="python /somedirectory/cwm.py"

20:45:13 * danbri_ symlinks

20:53:23 <danbri_> logger, pointer?

20:53:23 <danbri_> See http://ilrt.org/discovery/chatlogs/rdfig/2002-04-16#T20-53-23

20:59:21 * danbri_ fiddles with n3s

21:02:04 <AaronSw> sbp, you can use newlines on the commandline

21:02:18 <AaronSw> using quotes or backslashes

21:02:34 * danbri_ has trouble with n3s example...

21:02:55 <danbri_> #!/bin/sh

21:03:02 <danbri_>

21:03:02 <danbri_> n3s -s '@prefix default <http://xmlns.com/foaf/corp#> . { ?x sr ?y; spe ?z . ?y log:notEqualTo "" . ?z log:notEqualTo "" . ?p math:quotientOf (?y ?z) } => { ?x noe ?p }' > danbri-test.n3;

21:03:02 <danbri_>

21:03:02 <danbri_> #cwm --rdf http://ilrt.org/~ecdb/crp/reports.xml --n3 --filter=danbri-test.n3

21:03:02 <danbri_> Shouldn't this work? pretty muh copied from above.

21:03:05 <danbri_> TypeError: 'in' or 'not in' needs sequence right argument

21:03:51 <AaronSw> what isn't working? the cwm line?

21:03:53 <AaronSw> is it commented out?

21:04:02 <danbri_> nah, the prev one.

21:04:21 <AaronSw> works for me. it puts:

21:04:23 <AaronSw> this log:forAll :p , :z , :y , :x .

21:04:23 <AaronSw> { :x <http://xmlns.com/foaf/corp#sr> :y;

21:04:23 <AaronSw> <http://xmlns.com/foaf/corp#spe> :z .

21:04:23 <AaronSw> :y log:notEqualTo "" .

21:04:23 <AaronSw> :z log:notEqualTo "" .

21:04:24 <danbri_> n3s. sbp's vanished, right in the middle of an n3s early adopter hitting a but...

21:04:24 <AaronSw> :p math:quotientOf (:y :z) } log:implies { :x <http://xmlns.com/foaf/corp#noe> :

21:04:26 <AaronSw> p } .

21:04:29 <AaronSw> in danbri-test

21:04:29 <danbri_> s/but/prob/

21:04:42 <AaronSw> n3s is just an abbreviation syntax for n3

21:04:52 <danbri_> odd. I copied n3s into /usr/local/bin/ as a python script

21:05:11 <AaronSw> er, it's a python script that preprocesses abbreviated n3

21:05:17 <danbri_> yup, understood

21:05:40 <AaronSw> might want to just use the expanded n3 if n3s isn't working for ya

21:06:05 <danbri_> I can't expand it without n3s, not being an N3 weenie...

21:06:12 * danbri_ copies from above :)

21:06:23 <AaronSw> remove the newlines, tho

21:06:46 * DanCon can't work... my ssh tunnel keeps falling over, totally destroying my train of thought as I review email

21:07:06 <AaronSw> { :x <http://xmlns.com/foaf/corp#sr> :y; <http://xmlns.com/foaf/corp#spe> :z . :y log:notEqualTo "" . :z log:notEqualTo "" . :p math:quotientOf (:y :z) } log:implies { :x <http://xmlns.com/foaf/corp#noe> :p } .

21:07:48 <AaronSw> don't you cache the email locally?

21:08:26 <DanCon> evolution does some caching.

21:09:08 <DanCon> but it does some totally rude stuff: like: I select a message from a folder; it displays the message for half a second. then, for some reason, it notices the IMAP connection has gone done, and blanks the message display and folder windows.

21:09:18 <danbri_> Er, Aaron can you msg me the complete N3... I'm trying variants of the above and its mussing stuff

21:09:29 <danbri_> er missing, but that too

21:09:30 <AaronSw> sure

21:09:32 <danbri_> ta

21:10:03 <AaronSw> blanks the message display: ouch

21:10:31 <AaronSw> does it have an offline mode?

21:11:49 <AaronSw> replay: does it have an offline mode?

21:11:57 <AaronSw> hm, so much for that

21:24:00 * DanC_ is having nasty network problems.

21:25:20 <DanCn_> so... on what HTTP uris can denote... TimBL suggested { ?x log:uri [ str:startsWith "http:"; str:notContains "#" ] } log:implies { ?x a doc:Work }.

21:25:29 <DanCn_> fair enough... he can say what doc:Work means.

21:26:02 <mhgrove> DanC, are you around?

21:26:03 <AaronSw> I was trying to figure out sandro's position and I realized he doesn't believe in change.

21:26:04 <danbri_> The only plausible definition I've seen is 'those things that can be named with http: URIrefs without #'...

21:26:23 * danbri_ commits the biz/ed test examples to w3.org

21:26:24 <jonb> :-)

21:27:18 <mhgrove> DanC, mind if I ask you a quick question about those rdf test cases?

21:27:30 * jonb remains unconvinced that there is any point in trying to parse a URIref

21:28:48 * DanCon attempts to switch ISPs

21:29:36 * dajobe pokes danbri_ with an RDF Schema WD

21:29:41 <mhgrove> actually is anyone familiar with the rdf test cases?

21:29:53 <dajobe> yeah, what

21:29:56 * dajobe is co-editor

21:30:44 <mhgrove> i'm trying to write an rdf parser because I got too frustrated with jena, and I was considering using the rdf test cases as tests for my parser...

21:30:55 <AaronSw> You got frustrated with ARP?

21:30:58 <dajobe> frustrated with jena? !

21:31:05 <mhgrove> is that surprising?

21:31:10 <dajobe> yes, IMHO

21:31:26 <dajobe> what was the Q anyway?

21:31:33 <mhgrove> it seemed it suited for what i was trying to do with it...

21:31:46 <AaronSw> does ARP have a separate interface from Jena?

21:31:53 <mhgrove> and i noticed in some cases there are classes are referenced to a URL that doesn't exist, like foo

21:31:59 <dajobe> AaronSw: um, I think so

21:32:17 <mhgrove> i meant ill suited there...doh

21:32:21 <AaronSw> ah, so it can: http://www.hpl.hp.co.uk/people/jjc/arp/apidocs/com/hp/hpl/jena/rdf/arp/StatementHandler.html

21:32:41 <dajobe> mhgrove: eh? can you give me the url of the test case you mean?

21:33:09 <mhgrove> so the way i was doing a parser, when you were using some object at some url, if that url didn't exist, or the object isnt there, i throw an exception...is that not an error?

21:33:16 <mhgrove> yeah, one sec.

21:33:26 <dajobe> Note, RDF doesn't require resolving URLs

21:33:33 <dajobe> so example.org etc are used with abandon

21:33:56 <AaronSw> altho example.org does work these days...

21:34:17 <dajobe> ooh, what fun

21:34:25 <mhgrove> so there is a difference between legal syntax and valid identifiers (like a real URL)

21:34:46 <AaronSw> yeah

21:34:47 <dajobe> mhgrove: sorry? legal syntax of what? URis?

21:34:59 <dajobe> what's a real URL?

21:34:59 <mhgrove> legal syntax of RDF...

21:35:31 <mhgrove> so you can have a rdf file that parses fine...but that doesn't mean the objects it uses exist?

21:35:35 <dajobe> the identifier syntax must be valdi (they are URIs) but no resolving is required

21:35:47 <AaronSw> i.e. the urls can 404

21:35:58 <mhgrove> hrm...gotta rethink some stuff then...

21:36:00 <dajobe> mhgrove: where objects=things that the uris point to. Yes

21:36:27 <mhgrove> yeah sorry, object oriented terms are hard wired into my brain thanks to 4 years of higher educatin

21:36:28 <AaronSw> Hm, why does n-triples talk about character encoding?

21:36:31 <mhgrove> er education...

21:36:45 <dajobe> AaronSw: so it can express all Unicode chars; this is old news no?

21:37:03 <AaronSw> er, i meant for URIs, sorry

21:37:32 <dajobe> hmm

21:37:50 <dajobe> that section might need updating to allow \uXXXX inside URI-refs

21:37:51 <danbri_> dc_rdfig:view

21:37:51 <dc_rdfig> C: SW application scenarios: Company Accounts (http://ilrt.org/~ecdb/crp/reports.xml)

21:37:52 <dc_rdfig> D: "World Wide Web Consortium Issues P3P 1.0 as a W3C Recommendation" (http://www.w3.org/2002/04/p3p-pressrelease)

21:37:53 <dc_rdfig> E: A mentograph of annotation (http://robustai.net/mentography/annotation.gif)

21:37:54 <dc_rdfig> F: Web Service Face to Face - Aligning With the Semantic Web, Eric Prud'hommeaux & David Booth (April 8-12, 2002) (http://www.w3.org/2002/Talks/0411-ws-f2f-sweb/)

21:37:55 <dc_rdfig> G: Status of Parka Database Software for RDF (http://www.mindswap.org/~rreck/pp_overview.shtml)

21:38:17 <mhgrove> hey cool, our site, mindswap =)

21:38:56 <danbri_> C:I've now checked this into w3.org alongside some N3 example queries (thanks Sean, Aaron :), see [http://www.w3.org/2002/04/corpinfo/semweb-scenario|SW application scenarios: Company Accounts].

21:38:56 <dc_rdfig> added comment C4

21:39:47 <mhgrove> thanks for answering my questions dajobe and AaronSw...i'm gonna go stumble back to the drawing board

21:39:47 <danbri_> dajobe, could you point me at an example of N-Triples with embedded chunk of XML literal markup? (multiline?)

21:40:36 <dajobe> mhgrove: oh. I'm not sure what I've done :) ARP is a fine RDF/XML parser, but feel free to write something new :)

21:40:45 <danbri_> MaxF was asking re his XSLT parser (and also we were wondering how to hook up the RDF test cases to do checking... maybe using JanG or Jeremy's graph match tools?)

21:40:48 <dajobe> danbri_: I doubt it

21:40:58 <mhgrove> is ARP really different from Jena?

21:41:18 <dajobe> mhgrove: it is part of Jena; you should be able to extract just the parser if you take the right tree of classes

21:41:22 <danbri_> OK. We'll continue with a made up version for now

21:41:27 <danbri_> BTW how's the Ruby redland interface?

21:41:39 <dajobe> you tell me :)

21:41:43 <danbri_> I tried to build from CVS a day or few ago, something barfed, possibly my fault...

21:41:52 <dajobe> or mine

21:41:53 <danbri_> I got RDF query working, thought I could try it against Redland RDFdb...

21:42:02 <danbri_> I didn't investigate much...

21:42:19 <mhgrove> hrm...i might have to look into that then...before i get into this parser much further, i'm already near the point of no return

21:42:20 <danbri_> It _won't_ be efficient, but thought would be an interesting experiment...

21:42:21 <dajobe> what would be a good embedded xml in rdf/xml? mathml?

21:42:26 * danbri_ takes a break

21:42:31 <danbri_> mathml is good example, sure

21:42:46 <danbri_> XHTML in RSS 1.0 feeds (via an extension property) is another.

21:43:04 <dajobe> the content module

21:43:12 <dajobe> DanC is really having problems...

22:04:34 <connolly> sigh... typical lack of pointers to context: http://www.vonage.com/help/international.html

22:34:28 <JosD>http://project.kahosl.be/swa2002/

22:34:28 <dc_rdfig> H: http://project.kahosl.be/swa2002/ from JosD

22:35:03 <JosD> H:|Semantic Web and Applications 2002

22:35:03 <dc_rdfig> titled item H

22:35:58 <JosD> H:This seminar took place on March 6th 2002 at het Pand in Gent. (Belgium)

22:35:58 <dc_rdfig> added comment H1

22:36:26 <AaronSw> hey, jang!

22:36:26 <AaronSw> err JosD

22:37:05 <JosD> H:Further on this page you find more details about the subject of the seminar, the programme (with downloadable presentations!)

22:37:05 <dc_rdfig> added comment H2

22:37:32 <JosD> Hi Aaron, how are you doing?

22:38:16 <AaronSw> Pretty good. Did you get DSL installed at home?

22:39:02 <JosD> yes Aaron, how did you find out?

22:39:16 <JosD> proof technology?

22:39:22 <AaronSw> heh, heh

22:39:25 <AaronSw> from your IP: JosD (amdus@adsl-64317.turboline.skynet.be)

22:40:04 <JosD> o I see -- it's a big improvement!

22:40:32 <AaronSw> heh, i'd expect so!

22:40:53 <AaronSw> it's ironic, though. DanC's DSL line flaked out just before you joined

22:41:46 <JosD> really? any reason?

22:42:53 <AaronSw> not that i know

22:43:47 <JosD> Aaron, are you in Chicago now?

22:43:55 <AaronSw> Yep... well, pretty close.

22:47:07 <JosD> that was a interesting remark Seth, ...the darkness of triples is relative...

22:47:51 <Seth> thanks .. do you believe it?

22:48:53 <JosD> yes, see http://lists.w3.org/Archives/Public/www-webont-wg/2002Apr/0087.html

22:49:42 <JosD> (point 2)

22:51:10 <Seth> i think your point is much deeper than mine

22:51:59 <JosD> so, what is your point, the simpler the better

22:52:47 <Seth> i was just saying that if you have a sub graph, that is dark, then the darkness of the subgraph is relative to the graph that contains it, not some absolute kind of darkness. In other words, were one to put a flag in the triple to say that it was dark, it would be an absurbity.

22:54:08 <Seth> of course we need to establish that a graph can contain another graph

22:55:01 <Seth> ... i kind of dealt with this in http://robustai.net/mentography/context_lex_knows.gif

22:55:17 <JosD> that's sound I think, indeed, but contain? I would rather say dereferecable or some such...

22:56:27 <Seth> in that exanmple 'Lex destroy World' is dark to the outer graph, but not dark to the innermost graph

22:57:18 <JosD> can you point to that example (so that I can dereference ;-)

22:58:16 <Seth> url to the example was given above ... not sure what kind of pointer you want?

22:58:23 <JosD> oops, you did, sorry, looking ...

23:02:14 <JosD> thinking about PatH's words : "More generally, however, I would suggest that we take care to keep

23:02:14 <JosD> functionally distinct aspects of the language as distinct as

23:02:14 <JosD> possible, and that referring to/pointing to/whatever some ontology

23:02:14 <JosD> ought to one thing, and any speech act (assenting, asserting,

23:02:14 <JosD> denying, questioning, expressing doubt about, saying it is connected

23:02:15 <JosD> to foo, saying it entails foo....) involving it should be something

23:02:17 <JosD> else."

23:03:18 <JosD> so I wonder about that in your representation...

23:03:57 <Seth> yep, so one ontology that points out another ontology, doen't assert it ... all the triples in the designated ontology are dark relative to the ontology that pointed it out.

23:03:58 <danbri_>http://lists.w3.org/Archives/Public/www-archive/2002Apr/0040.html

23:03:59 <dc_rdfig> I: http://lists.w3.org/Archives/Public/www-archive/2002Apr/0040.html from danbri_

23:04:11 <danbri_> I:|RDF query by example, example

23:04:11 <dc_rdfig> titled item I

23:04:16 <Seth> .. that all im saying ... not a very deep observation

23:04:49 <danbri_> I:Idea was to decorate graphs with variable names as properties, as a syntax for RDF query test cases (that maps to Squish, RDQL, Algae etc.)

23:04:49 <dc_rdfig> added comment I1

23:05:00 <danbri_> I:Comments, feedback, flames etc welcomed...

23:05:00 <dc_rdfig> added comment I2

23:05:09 * danbri_ waves, heads off

23:05:12 <JosD> all right Seth

23:05:33 <danbri_> hi JosD btw, welcome to the world of proper TCP/IP access :)

23:06:38 <JosD> hi DanB (if I may call you like that) how are things going these days?