Semantic Web Interest Group IRC Chat Logs for 2003-05-22

03:03:27 <sethl_lap> most complicated query tested so far: SELECT ?x WHERE (?x, <rdf:type>, <cp:Parameter>), (?x, <cp:parameterName>, ?y), (?x, <cp:parameterValue>, ?z) AND ?z > 20 and ?y == \"TTL\" USING rdf FOR <http://www.w3.org/1999/02/22-rdf-syntax-ns#>, cp FOR <http://www.brivo.com/2002/12/controlpanel#>

03:03:52 <sethl_lap> with new RDQL engine for rdflib

03:05:14 <sethl_lap> I have to wire in the rdf-query testcases now

03:05:23 <sethl_lap> I'm relatively confident that the engine is working

06:45:25 <deluxe> deluxe is now known as deluxe|bbl

06:46:57 <lasse> hmm.. lots of people on this channel but not very active

06:47:43 <lasse> I'm doing my masters thesis (MSc) about Semantic Web in Knowledge Sharing. Hints and tips are very welcome :)

09:14:36 <edd-lap_> edd-lap_ is now known as edd

09:36:43 * DanC looks around for sandro and EricM

09:41:35 <danbri> i saw ericm talking to dave beckett outside w3ctrack room

11:09:33 <tim-patria> tim-patria is now known as tim-budapest

11:10:48 * tim-budapest eonders about commandeering the channel for the rdfq bof

11:11:50 <crysflame> oo

11:12:04 <crysflame> there's always rhizo/#bots

11:14:06 <tim-budapest> DanC, which channel were you on?

11:16:02 <dajobe>http://www.tbray.org/ongoing/When/200x/2003/05/21/RDFNet

11:16:03 <dc_rdfig> A: http://www.tbray.org/ongoing/When/200x/2003/05/21/RDFNet from dajobe

11:16:10 <dajobe> A:|The RDF.net challenge - Tim Bray

11:16:11 <dc_rdfig> Titled item A.

11:16:30 <dajobe> A:/me stops wasting his life dealing with syntax

11:16:30 <dc_rdfig> Added comment A1.

11:19:14 <edd> bah, largely recycled from his earlier rants

11:20:09 <dajobe> yes, and now I have to respond

11:21:35 <dajobe> and I have to do it on the web

11:23:04 * edd sighs

11:23:14 <edd> it seems a bit ego feeding for him

11:23:21 <edd> do you really have to make a response?

11:25:08 * edd hopes the query bof going well :)

11:25:53 <tim-budapest> XQuery being adapted to SQL by cacnnonical rep'n of RDBs

11:30:15 <dajobe> rdf query bof...

11:30:37 <dajobe> andys: dusing xml query - be daft to duplicate the work there. take the XPath out and put an rdf graph matcher in

11:30:38 <tim-budapest> Take Xpath out and RDFpath in - you can return stuff which can be RDF, hich is the right way to work (agreement)

11:30:41 <dajobe> xml query returns results that can be formated

11:30:51 <dajobe> so it would be ths same for rdf, want rdf back

11:31:01 <dajobe> Andys^

11:31:13 <dajobe> emiller:

11:31:29 <dajobe> w3c is working very heavily in the area of query

11:31:35 <dajobe> ... lots of mmebers working on thuis

11:31:48 <dajobe> and is importatnt to understand the relationship to that going on in the field with that in the rdf

11:32:00 <dajobe> as long as what ytou described solves this problem in directed labeleg gtraphs

11:32:06 <dajobe> can use the existing framework, seesms symbiotic

11:32:21 <tim-budapest> What paul said to EricM was that a lot of people are working on XQuery..

11:32:24 <dajobe> emiller reproted what paul ? said

11:32:45 <dajobe> ericP: how many peole like the idea?

11:32:54 <dajobe> xquery -xpath +an rdf path

11:32:59 <dajobe> proposal

11:33:03 <dajobe> counting heads in the room

11:33:19 <DanC> poll: "Take Xpath out and RDFpath in - you can return stuff which can be RDF, hich is the right way to work" worth exploring

11:33:20 <dajobe> consensus pretty much that it was worth exploring

11:33:21 <crysflame> mm, xquery with an rdfpath. interesting.

11:33:22 <DanC> many in favor

11:33:25 <dajobe> 2/3 in favour

11:33:36 <dajobe> timbl: i've done that

11:33:44 <dajobe> ... if you take xml query FOR, LET ...

11:33:53 <dajobe> ... WHERE constraints RETURN something

11:34:05 <dajobe> xmlq is based on orderede xml 'FLOWR'

11:34:14 <dajobe> can iterate recursively over the thing, extremely ordered

11:34:24 <DanC> tutorial on cwm paths (and lists) http://www.w3.org/2003/Talks/0520-www-tf1-b-more/

11:34:26 <dajobe> and all process ordered as xml is ordered

11:34:30 <dajobe> DanC: corrects tim on orderering

11:34:42 <dajobe> timbl: if you take out the flowrc clause and remove ordering contrain

11:34:47 <dajobe> then the FOR, LET, WHERE are identical

11:34:58 <dajobe> then you get a ??? you get a WHERE ... RETURN extremely like rdf query

11:35:07 <dajobe> so if you start with xl query, then you gfet up this FLW thing

11:35:20 <dajobe> I'm worried that rdf will head in the same place as the xml syntax

11:35:23 <dajobe> [sigh!!!!]

11:35:32 <dajobe> ... when all we really needs ia where clause

11:35:57 <dajobe> take advanteage of xml query apps, optimising

11:36:02 <dajobe> ... we should develop rdf things

11:36:04 <dajobe> danbri:

11:36:10 <dajobe> <brilliance/>

11:36:34 <dajobe> bwm:

11:37:03 <dajobe> nick (gibbons) was wrong - it is possible to query an rdf graph with xml query??

11:37:18 <dajobe> it may be possible to go beyond that and do some inferenceing

11:37:36 <dajobe> bwm: more important Q - asks tim. Seems it is a no brainer to build bridges back to the xml community

11:37:59 <dajobe> ... you almost seem to be arguing that we shouldn't be looking at all of xml query

11:38:08 <sandro> (that perhaps you can use any query language with inference underneath and the language doesnt change)

11:38:12 <dajobe> timbl: we should look at it and especially look at the functions

11:38:21 <dajobe> ... and not duplicate that. We do need an rdf path.

11:38:41 <dajobe> ericP: when we take out xmlpathj from xmlq, and add an rdfpath

11:38:50 <dajobe> and strip down to FLW or when, wheren, then we have a graph query language

11:39:00 <dajobe> that we can use without that rdfpath - one thing to implemnent

11:39:16 <dajobe> and then above that add xpath to get at some other parts of the document [??]

11:39:22 <dajobe> ... peeking into the literals

11:39:32 <dajobe> We should do the rdf path part and then reuse XMLQuery if we want to

11:39:38 <dajobe> on the reimaining bits such as querying lierals.

11:39:42 <dajobe> benG:

11:40:07 <dajobe> one way to do this is to consider a subset of xml query

11:40:17 <dajobe> and then backend it with an xml query app

11:40:25 <dajobe> rather than deal with low level detail again

11:40:30 <dajobe> such as datatypes

11:40:43 <dajobe> and their functions, list ops etc.

11:41:09 <dajobe> want to work with a free-nough language that optimisers

11:41:12 <dajobe> can deal with

11:41:13 <tim-budapest> q+ to say - mapping RDFQ to XMLQ is straightforward given a canonical (arbitarty) representatoih fo RDF in XML.

11:41:43 <dajobe> some capabilities that are really to do with graphs and not trees

11:41:55 <dajobe> that really won't work with existing xquery impls since they really don't do it

11:42:01 <dajobe> and won't optimise well.

11:42:41 <dajobe> DanC: inx ml query you query the syntax

11:42:47 <dajobe> in rdf, you deal with the domain of discourse

11:42:59 <dajobe> ... most interested in the testing format, can it be expressed in test cases

11:43:14 <dajobe> query WG process can then be measured against those tests - when done these, you are at last call

11:43:29 <dajobe> ... want to continue discussing these tests

11:43:39 <dajobe> and ask about 'can you do these in XML Query'

11:43:46 <dajobe> and see which ones can/cannot be done

11:43:48 <dajobe> .

11:44:02 <dajobe> strawpoll about DanC point above

11:44:07 <dajobe> approx 1/4 in favour

11:44:09 <dajobe> danbri:

11:44:16 <dajobe> good job connecting to xml & xmlq

11:44:19 <dajobe> ^on

11:44:50 <dajobe> translation - the ijndustry is taking things expressed in els & attrs and translatiing into things in the backened database

11:45:03 <dajobe> rdf query needs a good story

11:45:12 <libby>http://lists.w3.org/Archives/Public/public-qt-comments/2002Sep/0020.html

11:45:12 <dc_rdfig> B: http://lists.w3.org/Archives/Public/public-qt-comments/2002Sep/0020.html from libby

11:45:13 <dajobe> bullet points -

11:45:18 <dajobe> - querying datatypes & xsd datatypes

11:45:26 <dajobe> - using xmlq functions and operators spec (WD)

11:45:33 <dajobe> (use the whole thing probably)

11:45:39 <dajobe> - use the chunk of xml query that does templating

11:45:41 <libby> B:|RDF Use Cases - message from Jonathan Robie

11:45:41 <dc_rdfig> Titled item B.

11:45:47 <dajobe> maybe we can use the whole thing?

11:46:02 * DanC notes we have 15 minutes left

11:46:04 <dajobe> by passing the rdf query around in the xmlq

11:46:05 <libby> B:"This is a response to Libby Miller's comment [1] on RDF query use cases."

11:46:05 <dc_rdfig> Added comment B1.

11:46:16 <dajobe> andys:

11:46:24 <dajobe> going from rdf to xmkl database sounds like a good way to encrypt the query

11:46:38 <dajobe> going from rdf straight to dht e store is a mmuch better and more optimisable way

11:46:38 <DanC> B:|XML Query response to comment on RDF query use cases

11:46:39 <dc_rdfig> Titled item B.

11:46:50 <dajobe> when I said use the xml query frameowkr, I meant use the conceptual model

11:46:58 <dajobe> as modified for the rdf domain

11:47:31 <dajobe> AndyS: does their approach handling query seem apoproriate for the rdf domain

11:47:34 <edd-lap_> edd-lap_ is now known as edd

11:47:36 <dajobe> then going to the details

11:47:52 <dajobe> danbri: we still may be able to use some of the details

11:48:02 <dajobe> timbl: web page soemwhere on comparing the funcitons [??]

11:48:33 <dajobe> for rdf in an canoical xml form,a ny rdf query can be expressed in an xmlq unless it is incomplete

11:48:44 <dajobe> but the messyiness of the xmlq out depends on your input, mapping

11:48:56 <dajobe> might be worth going through the canonical thing and then

11:49:07 <DanC> on comparing XML Query functions and operators with cwm built-ins, by timb: http://www.w3.org/2000/10/swap/doc/Functions

11:49:08 <dajobe> do the query so that if you want to use your xmlq engine, you can do it

11:49:20 <gromgull> Sorry

11:49:23 <dajobe> timbl: I don't think, like andy, that this will be pretty.

11:49:41 <libby> sorry?

11:49:51 <gromgull> wrong window , sorry :)

11:49:54 <libby> :)

11:49:58 <dajobe> ericP: next item

11:50:59 <dajobe> what to discuss chat

11:51:37 <dajobe> meet again this week?

11:51:54 <dajobe> maybe friday evening?

11:52:16 <dajobe> action ericp: find a room for friday evening

11:52:38 <dajobe> room details to be recorede on the message board by 12:00 friday

11:52:42 <dajobe> and on the ESW wiki page

11:52:55 <DanC> let's meet Friday; EricP to coordinate where; notice to message board, wiki page

11:53:08 <tim-budapest> (uri?)

11:53:09 <dajobe> ericP: 5 mins left

11:53:41 <dajobe> DanC: rs:size

11:54:03 <dajobe> ... only tests I've seen use the files to scope the data

11:54:12 <dajobe> ... in the query language, do you ahve to say this?

11:54:23 <dajobe> the rs:size param is a func of the manifest?

11:54:38 <dajobe> AndyS: some people want to know size of result table since some people want to know that for simple testing

11:54:55 <dajobe> DanC: is it ok, straighforward to take cwm, run q and then not use the count?

11:55:08 <dajobe> AndyS: I don't, I do a graph compare; several others don't either (in theroom)

11:55:11 <dajobe> DanC: great

11:55:17 <dajobe> sandro: are these answers correct?

11:55:21 <dajobe> AndyS: yes, they are

11:55:26 <DanC> I asked if it's ok to compute the count *outside* the query engine

11:55:29 <dajobe> ... with the cobjunctive pattern

11:55:46 <dajobe> sandro: concerned with dup answers

11:56:01 <dajobe> ha sanyone with DQL answers looked at this, notion of equality I expect is different

11:56:03 <dajobe> ?

11:56:27 <dajobe> DanC: no inference allowed?

11:56:55 <dajobe> [too many people talking]

11:57:01 <DanC> (Danbri draws on the whiteboard; anybody willing to take a picture and put a pointer here?)

11:57:18 <dajobe> danbri think we can expand the manifest format

11:57:32 <dajobe> ... to extend to add it to cover when we do rules and inference

11:57:47 <dajobe> AndyS: tricky territory, diff inference engines have different answers from them

11:58:01 <dajobe> timbl: first roll out for the graph matchinging quiery for the test cases

11:58:08 <dajobe> ... then can be used for anyone who defines a particular logic

11:58:20 <dajobe> tests can start with this data and results are *that8

11:58:43 <dajobe> and maybe pick up thje OLW ones

11:58:47 <dajobe> nmg: same as tim

11:58:56 <dajobe> DanC: if you pick up the OWL stuff, you will have infinite counts

11:58:59 <dajobe> ericp:

11:59:15 <dajobe> I bet people who've done this in XSB have thought about this more

11:59:25 <dajobe> [what hafve they done?]

11:59:38 <dajobe> ... for testing if any facts they have... to test their soundness and completness

11:59:56 <dajobe> beng: soundness of the impl or anyh particular rulesset...

11:59:59 <dajobe> ericP: impls

12:00:16 <dajobe> beng: look on sourceforge?

12:00:23 <dajobe> jeen:

12:00:41 <dajobe> beng: they've been around, done some tof this

12:00:49 <dajobe> jeen: sesame does have an RDF MT inferencer, no benchmark on the query level

12:00:53 <dajobe> we do have a test suite on the api level

12:01:08 <dajobe> puts ina bunch of triople, close, computes closure, checks stuff

12:01:10 <dajobe> at api level

12:01:17 <dajobe> smh: 3store does this too

12:01:22 <dajobe> jeen: on ource sourceforge site

12:01:47 <dajobe> ari: typed graph matching people

12:01:54 <dajobe> "sub graph isomorphism" into google

12:01:59 <dajobe> have done this, similar problems

12:02:06 <dajobe> they may have graph datasets

12:02:11 <dajobe> with tests

12:02:20 <dajobe> ari: somebody at NYU

12:02:23 <dajobe> [hunt for google clues]

12:02:36 <dajobe> they must have test suites

12:02:45 <dajobe> ericP: thanks for all - friday evening

12:02:48 <dajobe> ----- end of meeting -----

12:02:49 <tim-budapest> - http://www.cs.sunysb.edu/~algorith/implement/nauty/implement.shtml

12:02:51 <tim-budapest> eg

12:03:07 <DanC> BenG: Cormen worte a book graph alrorithms

12:03:25 <DanC> & _ & Rivest

12:04:37 <DanC>http://www.cs.nyu.edu/cs/faculty/shasha/papers/agm.html

12:04:37 <dc_rdfig> C: http://www.cs.nyu.edu/cs/faculty/shasha/papers/agm.html from DanC

12:04:51 <DanC> C:|Graphdiff: Approximate Graph Matcher and Clusterer

12:04:52 <dc_rdfig> Titled item C.

12:05:40 <Wack> hmm, does rdfs support multiple inheritance? (cant seem to find it in the rdfs wd)

12:06:10 <crysflame> wack: rss modules seem like a .. kind .. of rdf inheritance, at least from a beginner standpoint here on my side.

12:13:16 <shellac> rdfweb.org, xmlns.org & foafnaut.org are about to disappear from the internet for ~30min

12:20:42 <dajobe>http://ilrt.org/discovery/chatlogs/rdfig/2003-05-22.html#T11-30-15

12:20:42 <dc_rdfig> D: http://ilrt.org/discovery/chatlogs/rdfig/2003-05-22.html#T11-30-15 from dajobe

12:20:51 <dajobe> D:|RDF Query Bof transcript - partial

12:20:51 <dc_rdfig> Titled item D.

12:23:06 <dajobe> BLURB: Reminder, WWW2003 conference chat in #www2003, please join us

12:23:06 <dc_rdfig> E: Reminder, WWW2003 conference chat in #www2003, please join us from dajobe

12:23:39 <dajobe> E:[http://www2003.xmlhack.com/|WWW2003 community coverage]

12:23:40 <dc_rdfig> Added comment E1.

12:26:54 <dajobe>http://internetalchemy.org/2003/05/rpvSyntaxIsUgly.html

12:26:55 <dc_rdfig> F: http://internetalchemy.org/2003/05/rpvSyntaxIsUgly.html from dajobe

12:27:01 <dajobe> F:|RPV Syntax is Ugly

12:27:01 <dc_rdfig> Titled item F.

12:27:21 <dajobe> F:Ian Davis in reply to Tim Bray

12:27:26 <dc_rdfig> Added comment F1.

13:02:04 <nmg_> nmg_ is now known as nmg

13:20:52 <libby_> libby_ is now known as libby

14:02:27 <sethl> morning

14:03:02 <libby> sethl, have you been doing query stuff?

14:03:24 <sethl> yup! implementation of RDQL for rdflib, I was going to hand off to eikeon today

14:03:35 <sethl> though, been having hard time defining what RDQL actually is :)

14:03:41 <libby> nice one!

14:03:47 <sethl> it was fun :)

14:03:50 <libby> really? there's a bnf somewhere I think

14:03:55 <sethl> I was looking at SeRQL, too

14:03:58 <libby> are you goign to use the testcase?

14:04:01 <sethl> yeah, inside Jena

14:04:02 <libby> testcases?

14:04:07 <libby> cool

14:04:15 <sethl> I was just looking at them last night, I'm going to try

14:04:24 <sethl> that way, if I pass those, I'll know I'm good w/ the spec

14:04:40 <sethl> i found many things I think are the testcases

14:04:56 * sethl goes to find the URI

14:06:10 <libby> we havve soem query ones - http://swordfish.rdfweb.org/rdfquery/tests/

14:06:19 <libby> though there are lots in jena dist too

14:07:13 <sethl> I found the Jena ones, I wasn't sure which are the "blessed" test cases (query test cases seems a bit overloaded)

14:07:27 <sethl> found this: http://www.w3.org/2003/03/rdfqr-tests/rdf-query-testcases.html

14:07:41 <libby> right

14:08:01 <libby> the actual cases are at the swordfish address (we'll move 'em later)

14:08:15 <sethl> nice, ok, the pieces are coming together...

14:08:23 <sethl> right now I have my own unit tests passing

14:09:05 <sethl> thanks libby, you read my mind :)

14:09:26 <libby> hm, fancy donating them?

14:10:24 <sethl> sure, no prob (they are in pyunit)

14:11:04 <lasse> are there any cool sites on the Internet which one could demonstrate the power of SemWeb/RDF easily?

14:11:19 <lasse> with

14:11:20 <sethl> I see squishtests and rdqltests... are they the same, just different query languages?

14:11:20 <sandro> sure would be nice

14:11:45 <sethl> lasse: lots of people use CWM (python app) to show nifty demos

14:12:26 <lasse> which platform does it run on or do I need a python compiler?

14:12:34 <sandro> there is no cwm web demo, alas. (I tried to write one for this conference, but didnt finish)

14:12:44 <sandro> cwm needs python, yes.

14:12:46 <sethl> any python (2.1 or >) I think is fine

14:12:59 <lasse> wonder if there's python for os x

14:13:04 <sethl> included!

14:13:11 <lasse> darn :)

14:13:16 <sethl> (at least, w/ 10.2)

14:13:27 <sandro> see http://www.w3.org/2000/10/swap/doc for documentation on cwm, including links to installation instructions

14:13:27 <lasse> thanks a bunch, must check it out

14:13:36 <sethl> fun little tool

14:13:57 <libby> thanks sethl

14:14:27 <libby> squish is basically a vaiant of rdql (or the other way around) bit less powerful, slightly simpler to write

14:14:46 <libby> s/vaiant/variant/

14:15:13 <libby> - http://swordfish.rdfweb.org/rdfquery/squish-bnf.html

14:15:17 <crysflame> heh

14:15:34 <crysflame> Parse::BNF was very helpful for translating that BNF into a grammar.

14:15:36 <sethl> ok, just wanted to make sure I'm running the right thing. so, there are a suite of tests for each query language it seems

14:15:44 <crysflame> ^ (different)

14:16:01 <libby> hm

14:16:15 <sethl> but the suites test the same things, yes?

14:16:18 <libby> the plan has been to convert the queries to ntriples, using bnodes as variables

14:16:43 <sethl> how will you express things like rdql's AND in ntriples?

14:16:49 <libby> at the moment, they test probbaly different things - we havn't worked if it's a comprehensive set

14:16:56 <sethl> gotcha

14:17:08 <libby> sethl, we don;t, just the simplest queries for now

14:17:13 <sethl> k

14:17:20 <libby> so oyu may find that some are in ntriples is all I meant

14:18:35 <libby> sethl, tell me if theyare useful to you, or if not, why not :)

14:18:46 <sethl> will do! I'll start w/ http://swordfish.rdfweb.org/rdfquery/tests/tests/rdql-tests-2003-04-10/

14:19:45 <mattb> i just started looking at rdflib a couple of days back

14:19:49 <mattb> anticipating doing some python soon

14:19:54 <sethl> I find it very easy

14:20:00 <sethl> interface is a lot like redland's

14:20:08 <mattb> yeah, seemed nice, python idiomatic

14:20:09 <mattb> any idea when the query engine will be ready?

14:20:27 <mattb> although redland is a lot faster of course

14:20:36 <sethl> you can try it out today I guess. only tested w/ my internal tests, though

14:20:50 <sethl> so real world usage is good

14:21:04 <sethl> I'll write up some docs and send it to eikeon today

14:21:09 <mattb> i'd be happy to have a play as soon as i can find time

14:21:14 <mattb> prolly in some quiet conf session

14:21:19 <sethl> then I'm going to try these test cases

14:21:59 <sethl> the main difference is my implementation of the AND filter. I use python's eval() to do it, so any valid python is valid there.

14:22:16 <sethl> so, you can write AND ?y > 20 and ?x == "foo"

14:22:39 <sethl> or AND ?y > 20 and (?x == "foo" or ?x == "bar")

14:22:45 <sethl> stuff like that

14:22:51 <mattb> interesting stuff

14:23:05 <mattb> it's based on top of rdflib, or a patch to it?

14:23:11 <sethl> it's on top

14:23:25 <mattb> if it uses the simple triple matching api calls, presumably it could use redland too?

14:24:07 <sethl> yeah, next step is to make it work against an interface, and swap out rdf implementations. since redland has the python layer, that will be cool to test.

14:24:17 <mattb> very nice

14:24:36 <sethl> there are only a few integration points, so it should be quick.

14:24:47 <sethl> would that be more useful that wiring in test cases?

14:24:57 <sethl> (there are already pyunit tests for it, though)

14:25:09 <mattb> i'd go for the testing first

14:25:14 <mattb> prove the engine then make it more portable

14:25:30 <sethl> will do, I can always refactor the interface in as I work the tests

14:25:31 <mattb> helps everyone else using the testcases, too

14:25:36 <sethl> true

14:25:40 <mattb> since they are new

14:25:49 <sethl> I gotta say,python makes it fun

14:26:04 <eikeon> ronwolf also wrote a rdflib backend that uses redland... another option for mixing and matching.

14:26:23 <sethl> cool

14:26:36 <dajobe> yeah, I've not seen that code yet from him

14:26:40 * mattb on a big back-to-python swing at the moment

14:26:41 * dajobe summons ronwalf

14:27:53 <eikeon>http://rdflib.net/cvs/rdflib/rdflib/store/RedlandStore.py

14:27:53 <dc_rdfig> G: http://rdflib.net/cvs/rdflib/rdflib/store/RedlandStore.py from eikeon

14:28:01 <dajobe> <eikeon>ronwolf also wrote a rdflib backend that uses redland... another option for mixing and matching.

14:28:11 <dajobe> ah, that's nice

14:28:19 <ronwalf> Yeah, I have to work on that backend a bit

14:28:27 <ronwalf> Some impedance matching to do

14:29:17 <edd> Morbus: yeah, me :)

14:29:19 <ronwalf> Redland has duplicate triples, RDFlib does not

14:29:37 <Morbus> edd: priv msg?

14:29:38 <ronwalf> Redland has contexts associated with each node

14:29:51 <dajobe> G:|RedlandStore.py - python interface for rdflib to Redland

14:29:51 <dc_rdfig> Titled item G.

14:30:05 <ronwalf> RDFLib has returns a list of contexts if you ask it about a node

14:30:14 <dajobe> does it?

14:30:16 <ronwalf> Those are the biggies

14:30:51 <ronwalf> If I remember right, it's something like store.contexts(statement)

14:30:59 <ronwalf> Er, switch node for statements earlier

14:31:31 <eikeon> Sounds right.

14:31:59 <dajobe> where?

14:32:12 <dajobe> oh, redland has ..

14:32:30 <sethl_2> doh, sorry

14:40:33 <chaalsBUD_> grrr network

14:40:36 <chaalsBUD_> libby?

14:40:44 <chaalsBUD_> query which no worky:

14:40:49 <chaalsBUD_> select ?mbox where (http://xmlns.com/foaf/0.1/mbox ?per ?mbox) (http://purl.org/net/inkel/rdf/schemas/lang/1.1#masters ?per it )

14:41:14 * chaalsBUD_ isn't sure if "it" should be quoted, but didn't work either way anyway

14:42:41 <libby> looking now...

14:43:41 <chaalsBUD_> seems to worky now, but doesn't get a result

14:51:00 <libby> chaals - this works: select ?mbox, ?ii where (http://xmlns.com/foaf/0.1/mbox ?per ?mbox)

14:51:13 <libby> it should work in the query itswlf tho

14:51:16 <libby> maybe a bug

14:51:55 * chaalsBUD_ doesn't get what your query does.

14:52:07 <libby> hm, this works too: select ?mbox where (http://xmlns.com/foaf/0.1/mbox ?per ?mbox) (http://purl.org/net/inkel/rdf/schemas/lang/1.1#masters ?per "en")

14:52:18 <libby> I think there was something screwed up with the jsp

14:52:29 <libby> ok now

14:53:04 <libby> hm, no the "it" version doesn;t work

14:53:07 <libby> weird

14:53:53 <libby> oh, no "it" in teh database

14:54:47 <chaalsBUD_> second query no worky for me :(

14:55:01 <libby> which one?

15:42:08 <danq> danq is now known as DanQ

19:22:44 <_joshua> shh

20:34:30 <sethl_2> sethl_2 is now known as sethl

22:41:58 <fulruf> !commands

23:08:39 <ziobelo> hi

23:08:53 <ziobelo> list

23:09:10 <ziobelo> commands

23:09:24 <ziobelo> !list

23:09:36 <ziobelo> !lista