07:18:53 RRSAgent has joined #dawg 07:18:53 is logging to http://www.w3.org/2005/01/20-dawg-irc 07:19:07 AlbertoR has joined #dawg 07:20:16 SteveH has joined #dawg 07:23:20 Zakim, remind me in 150 minutes 07:23:37 Zakim, help 07:23:40 kendall has joined #dawg 07:24:14 /invite zakim 07:24:22 Zakim has joined #dawg 07:24:39 Zakim, help 07:24:39 Please refer to http://www.w3.org/2001/12/zakim-irc-bot for more detailed help. 07:24:42 Some of the commands I know are: 07:24:44 xxx is yyy - establish yyy as the name of unknown party xxx 07:24:47 if yyy is 'me' or 'I', your nick is substituted 07:24:50 xxx may be yyy - establish yyy as possibly the name of unknown party xxx 07:24:52 I am xxx - establish your nick as the name of unknown party xxx 07:24:55 xxx holds yyy [, zzz ...] - establish xxx as a group name and yyy, etc. as participants within that group 07:24:58 xxx also holds yyy - add yyy to the list of participants in group xxx 07:25:00 who's here? - lists the participants on the phone 07:25:02 who's muted? - lists the participants who are muted 07:25:05 mute xxx - mutes party xxx (like pressing 61#) 07:25:08 unmute xxx - reverses the effect of "mute" and of 61# 07:25:10 is xxx here? - reports whether a party named like xxx is present 07:25:12 list conferences - reports the active conferences 07:25:14 this is xxx - associates this channel with conference xxx 07:25:15 excuse us - disconnects from the irc channel 07:25:17 I last learned something new on $Date: 2005/01/20 15:39:57 $ 07:25:19 Zakim, remind us in 150 minutes about lunch 07:25:19 ok, JanneS 07:30:21 AndyS has joined #dawg 07:30:44 (we revised the agenda a bit) 07:30:45 RRSAgent, pointer? 07:30:45 See http://www.w3.org/2005/01/20-dawg-irc#T07-30-45 07:30:50 ==== Protocol Spec 07:31:09 http://www.w3.org/2001/sw/DataAccess/ftf4.html 1.19 07:31:26 DONE: EricP/KendallC to publish "SPARQL Protocol for RDF" pending approval from TomA and AndyS 07:37:16 DanC: Abstract protocol has testing implications: should have more than one binding to test it 07:38:23 Kendall: Who might do the SOAP protocol? 07:38:47 ... links have been to HTTP id so far 07:39:45 ... gives weight to the "need 2 to test" arguement 07:40:28 Eric: SOAP has mappings to HTTP GET 07:40:59 ... WSDL decription to same GET; SOAP returns 07:41:47 Kendall: Assumed there will be a SOAP version ... want to make it RESTful via abs. protocol 07:44:40 Hmm... soap encoding for bindings? http://www.w3.org/TR/2003/REC-soap12-part2-20030624/#datamodel 07:44:57 ACTION EricP: inviestigate WSDL/SOAP encodings of SPARQL protocol 07:45:47 DanC: There is a risk of untested hooked (abstract protocol) 07:47:07 protocolRootReferent: No query string returns what? 07:47:59 Service-y thing: a description ; graph-centric world: all the graph 07:48:55 Kendall: wanted to be neutral 07:49:16 Steve: returns an HTML to query the graph 07:50:16 Kendall: serviceDescription => HTTP OPTIONS or GET on some URI 07:56:27 (maybe the root referent URI) 07:56:47 http://www.w3.org/Protocols/rfc2616/rfc2616-sec9.html describes http options 07:57:09 DanC: can't bookmark and link to OPTIONS 07:58:13 " Responses to this method are not cacheable." is directly from the spec 07:59:20 DaveB: suggests "GET /uri => service description" 08:00:12 Alberto: Does it work with firewall? 08:00:37 DaveB: Used by webdav : MS does OPTIONS on the root 08:01:23 Kendall: suggest we talk about what is in the service description 08:02:24 Kendall: wants client to adapt: e.g. RDQL query => SPARQL query used 08:02:44 ... based on server capabilities 08:04:00 DaveB: ATOM style: description doc has URIs to service end points for various operations 08:04:56 Kendall: How we cover GET on a graph to fetch (a representation of) the graph? 08:06:02 DanC: service discovery ... go find triples ... e.g. type SPARQLservice . 08:06:25 V1: other triples at same place 08:06:39 V2: go to service URI to get more triples 08:07:45 Kendall: What's the minimum here? 08:08:07 DanC: What's in the service description? 08:09:17 Maybe contains: ... 08:09:25 "Good at FOAF" 08:09:30 OWL inference level 08:09:45 what DAWG ext:functions() I support 08:10:53 Graphs you can query (DanC wishes to return here) 08:11:16 e.g. CIA fact book 08:12:05 DanC: Service description could say what's in the background graph 08:12:42 will it be the URI of the graph? 08:13:33 or the URIs of the "(automatically) loaded" graphs? 08:14:05 Kendall: Uses "WSDL 2 RDF" here 08:14:24 ... we (DAWG) are further ahead 08:14:41 ... we should not make up a vocabulary 08:15:28 DanC: we should do something 08:16:32 DanC: Formalize what we know we want: e.g. RDQL and SPARQL 08:16:43 ... RDFS closure 08:17:22 Homework: describe your service 08:17:25 or do here. 08:19:22 Choosing HTTP params names : why not just choose? 08:20:26 DanC: Who gets to choose? Do we take away a choice (degree of freedom) 08:25:23 See also: http://www.faqs.org/rfcs/rfc2396.html section 2.2 08:25:40 And http://www.faqs.org/rfcs/rfc1738.html section 3.3 08:30:36 [[Group vocabulary design]] 08:32:00 DanC: Service URI contains language in URI , not query string, http://.../service/sparql/... 08:32:50 for inspiration about SPARQL HTTP bindings, see the OGC WFS and WMS specs http://www.opengeospatial.org/specs/?page=specs - they defined a clear HTTP GET interface with well-known parameters 08:34:53 AndyS: Would require any client to interrogate a service description to use a service: can'tgo straight there. 08:35:20 ... this makes writing a JDBC driver have an extra point of failure (getting the service description) 08:36:59 some sample implementation (see URLs for example GetMap ) - http://www.asemantics.com/showcase/wms.html - simple enough for users - http://www.asemantics.com/showcase/zoom.html 08:37:37 DanC: Suggest: "/service?sparqlParam=SELECT * .." to avoid needing a SPARQL URI in the query 08:41:16 AndyS: This is the same as saying "?lang=sparql" - both are short names needing the service description to resolve 08:45:10 Eric: we spec a few short names to avoid needing service look up (we speced the param name) 08:49:56 +1 Eric 08:57:12 The param "sparqlParam" => languages, and "I comply to SPARQL protocol doc" 08:59:34 ACTION: DanC: write up "sparqlParam" approach 09:00:30 DanC: next: How do we say [Background graph is always RDFS closed]? 09:01:22 Eric: services will want to specify sets of OWL based on some properties 09:01:41 .... e.g Annotea: closure over a few properties that were useful 09:05:54 ACTION: SteveH: Write up a service description of features supported for his service 09:06:41 DanC: Next: specifying the graph(s) I can query over 09:06:56 by default? 09:07:10 or maximum list? 09:07:39 ... meaning the background graph 09:08:28 ... meaning the background graph or maybe the named graphs as well 09:09:06 Idea: :queriesAgainst , :queriesOver, ... 09:09:43 ACTION: Kendall: Add section for service descriptions and put graph specifying in it to start it off 09:09:55 will it be a URI or just a hadle? 09:10:08 s/hadle/handle/ 09:13:24 s/view/due/ 09:15:11 ACTION: DanC: Write up definitive service concept 09:19:09 ACTION: Kendall: Announce the protocol spec 09:19:34 ACTION: KendallC: think about predicate to relate html-forms and svc descriptions, and how this goes into the spec 09:19:57 kendall has joined #dawg 09:23:16 break for 30mins 09:42:25 PROPOSAL: add explicit yes/no in result set; keep boolean in protocol spec; yes, ASK syntax and section 10.4 Asking "yes or no" questions v1.133 17-Nov-04 09:43:10 [ Protocol doc has boolean results (amoung others). XML results set has no boolean result. Query doc assumes a boolean results. ] 09:44:28 janne: there will be an example in the query spec that shows the boolean result 09:44:44 (actually expressed as a question) 09:45:09 so RESOLVED. ABSTAIN: Connolly 09:45:17 ABSTAIN: SteveH 09:46:28 ======== protocolRootReferent --> http://www.w3.org/2001/sw/DataAccess/issues#protocolRootReferent 09:46:28 Janne - there already is an example : does http://www.w3.org/2001/sw/DataAccess/rq23/#ask work for you? 09:48:53 Action: KendallC to propose a solution to protocolRootReferent 09:49:26 Andy: wanna talk w/ me @ lunch about protocolRootReferent? 09:49:35 OK 09:50:14 works in the sense that the link works :-) The result is ok, I'm just wondering (I think choice of words Dave had) why the result set format cannot have if not the root element, then the 1st child is 09:50:38 ========= DESCRIBE --> http://www.w3.org/2001/sw/DataAccess/issues#DESCRIBE 09:51:53 punted 09:52:05 ========= schedule discussion 09:52:50 DanC: please think about cascadedQueries and DESCRIBE 09:53:38 (and update the wbs thingy) 09:55:20 JanneS, you asked to be reminded at this time about lunch 09:57:01 JanneS, Zakim wants to know if you're hungry 10:04:30 [ discussion of removing abstract syntax from protocol vs. adding a soap binding vs. untested abstract syntax ] 10:05:30 oh, yes, lunchtime is NOW 10:05:36 DanC: how about Last Call (query, protocol, use case) 17 March ? 10:05:39 Zakim has left #dawg 10:57:50 AndyS has joined #dawg 11:06:30 Yoshio has joined #dawg 11:23:15 [ examining f2f for LC issues resolution 23-24 May ] 11:24:25 ACTION AlbertoR: propose a f2f meeting in Leiden 11:31:48 near xtech http://www.xtech-conference.org/ 11:33:27 "I don't know" -- for the record! :> 11:36:01 ========= DESCRIBE --> http://www.w3.org/2001/sw/DataAccess/issues#DESCRIBE 11:37:51 ACTION DanC: (with EricP) communicate WBS user experience to systems team 11:41:30 Alberto: we offer a Z3950 interface that leans on DESCRIBE 11:41:45 q+ to ask how we write tests for DESCRIBE 11:41:49 hmm 11:41:55 ohwell 11:44:58 [ disucssion of danbri's proposal ] 11:45:50 Janne: Issue of the lifetime of the ?doc URI in danbri proposal 11:48:59 the doc could be too big a file... user should be guided and that what DESCRIBE should do... no? 11:49:57 HiroyukiS has joined #DAWG 11:51:17 [ are empty DESCRIBE results valid answers? ] 11:51:58 I think so, eric, with Yes-No information 11:52:39 DanC: we could say "any valid RDF document is an answer" 11:53:19 AndyS: we could go further, we could presribe, for the purposes of this test, this is the answer 12:05:18 PROPOSED: keep the DESCRIBE syntax from SPARQL Query 1.3 12:05:43 OBJECTIONS: DanC 12:05:57 JanneS has joined #dawg 12:05:59 Abstentions: KendallC, SteveH, DaveB 12:06:04 RESOLVED 12:07:19 http://librdf.org/query?language=sparql;uri=http%3A%2F%2Fwww.ilrt.bristol.ac.uk%2Fpeople%2Fcmdjb%2Fwebwho.xrdf;query=PREFIX%20foaf%3A%20%3Chttp%3A%2F%2Fxmlns.com%2Ffoaf%2F0.1%2F%3E%0ASELECT%20%3Fname%20%3Fnick%0AWHERE%0A%20%20%28%3Fx%20rdf%3Atype%20foaf%3APerson%29%0A%20%20%28%3Fx%20foaf%3Aname%20%3Fname%29%0A%20%5B%20%28%3Fx%20foaf%3Anick%20%3Fnick%29%20%5D%0A 12:07:28 http://esw.w3.org/topic/DawgShows 12:08:40 ACTION 12:09:15 ACTION AndyS: propose test case re optionals and ordering 12:10:53 WITHDRAWN: # 12:10:53 ACTION: KendallC expose our walking tour data to SPARQL querying clients 12:10:57 action -11 12:12:23 KendallC: NASA is going to have a big interop. SPARQL interop may be the best thing to demo. Rough deadline someplace in August. 12:13:06 ... For SPARQL, it will probably mean turning our tests into some sort of conformance suite. 12:13:19 ... (similar for OWL) 12:14:36 ... RDFXML will work, but it's trivial and boring 12:14:52 s/will work/may work/ -- daveb! :> 12:15:01 ======== # handling public comments 12:15:18 DanC: we used to have TomA. is chaos sufficient? 12:15:34 DaveB: i just joined the list 12:16:58 KendallC: Re Tucana, the Tucari developers are continuing at sourceforge. they claim ITQL is a superset of SPARQL and to have faced stuff that we'll be facing soon. 12:22:39 ======= SPARQL QL spec 12:22:56 Zakim has joined #dawg 12:22:59 Zakim, pick a scribe 12:22:59 sorry, DanC_lap, I don't know what conference this is 12:23:20 http://www.w3.org/2005/01/yacker/uploads/sparql_n3?lang=perl 12:26:49 strawpoll, likes RDQL: 2 12:27:00 +1 (JanneS) 12:29:07 likes N3: 3 12:29:14 example, in preparation to straw poll [[ 12:29:27 PREFIX foaf: 12:29:27 SELECT ?me ?you 12:29:27 WHERE { ?me foaf:knows ?you. ?you foaf:knows } 12:29:27 PREFIX foaf: 12:29:27 SELECT ?me ?you 12:29:28 WHERE { ?me foaf:knows ?you. ?you foaf:knows } 12:29:30 PREFIX foaf: 12:29:32 SELECT ?me ?you 12:29:34 WHERE { ?me foaf:knows ?you. ?you foaf:knows } 12:29:36 ]] 12:31:06 oppose: RDQL 2 / N3 3 12:34:11 s p1 o1; p2 o2, o3. 12:34:22 turtle example 1: http://www.w3.org/2005/01/yacker/uploads/sparql_n3?lang=perl&text=SELECT+*%0D%0AWHERE+%7B+%3Fp+%3Fs+%3Fo+%7D&action=validate+text 12:34:51 DB: I can see doing: 1. () to {} 2. period between triples. 3. ; 4. , 12:35:06 http://www.w3.org/2005/01/yacker/uploads/sparql_n3?lang=perl&text=SELECT+%3Fp+%3Fo%0D%0AWHERE+%7B+%3Fs+%3Fp+%3Fo+.+%3Fo+%3Fp2+%3Fo2%7D&action=validate+text 12:36:01 http://www.w3.org/2005/01/yacker/uploads/sparql_n3?lang=perl&text=PREFIX+foo%3A+%3Chttp%3A%2F%2Fwww.w3.org%2F%3E%0D%0ASELECT+%3Fp+%3Fo%0D%0AWHERE+%7B+%3Fs+%3Fp+%3Fo+.+%3Fo+%3Fp2+%3Fo2%7D&action=validate+text 12:40:43 COSNTRCUT example: http://www.w3.org/2005/01/yacker/uploads/sparql_n3?lang=perl&text=PREFIX+foo%3A+%3Chttp%3A%2F%2Fwww.w3.org%2F%3E%0D%0ACONSTRUCT+%7B+%3Fs+foo%3Abop+%3Fo2+.%0D%0A++++++++++++%3Fo2+foo%3Abing+%3Fo+%7D%0D%0AWHERE+%7B+%3Fs+foo%3Abar+%3Fo+.%0D%0A++++++++%3Fo+foo%3Abaz+%3Fo2+%7D&action=validate+text 12:41:29 [[[ 12:41:30 PREFIX foo: 12:41:31 CONSTRUCT { ?s foo:bop ?o2 . 12:41:31 ?o2 foo:bing ?o } 12:41:31 WHERE { ?s foo:bar ?o . 12:41:32 ?o foo:baz ?o2 } 12:41:34 ]]] 12:46:57 DB: I can see doing: 1. () to {} 2. period between triples. 3. ; 4. , 12:47:06 ^ n3 style 12:48:34 N3 style: 1. (a b c) (d e f) becomes { a b c . d e f } 2. semi ; and 3. , 12:48:50 else RDQL/1.171 12:49:05 pref RDQL: 3 12:49:14 pref N3: 2 12:49:26 oppose RDQL: 1 12:49:36 oppose N3: 2 12:51:46 PROPOSED: adopt syntax of 1.171 SPARQL draft 12:52:19 RESOLVED: abstain KC AS. EP objecting. 13:02:45 ----- accessing collections ------ 13:04:47 I was reading http://www.aifb.uni-karlsruhe.de/WBS/pha/rdf-query/ 13:05:47 DaveB: only one path language does it 13:05:53 ericP: plus algae 13:07:27 no? so I kind of prefer having some querying method (over postponing), though having no idea so far 13:07:43 going round the table.... 13:08:02 kendall: we dont have anything 13:08:16 danc: conflicted, think we should postpone 13:08:41 janne: havent needed it, wont miss it right away, whats a typical use case? find simple solution? 13:09:25 ericp: what i did in algae is a hack, if theres a certain flag set and you ask for something that points to a list it searches for anything in that list - works for testcases, but never had RW scenario 13:09:34 ericp: no thoughts on what WG should do 13:09:55 yoshio: if we dont have mehtod to do this its hard, prefer having something 13:10:16 hiruki?: not an aspect for query right now 13:10:40 alberto: collents not in problem space, not interesting in solution, should eventually use DESCRIBE for this 13:10:55 s/hiruki/hiroyuki/ 13:11:00 daveb: i dont use, cant rush into it, needs a lot of thought, shouldn't start now as no proposal 13:11:28 andy: not fatal not to do it, will miss out the community, chikcen and egg, dont use because you cant access them easily 13:11:48 andy: two solutions, member-of property - would be 1st prop we have to create 13:12:05 andy: special binding function member_of() 13:12:31 andy: we have to make concious solution, will come up as a comment 13:13:13 SH: I'm pretty sure I'd object to any proposals at this point; I don't have time to get experience to make me confident about any proposals in the near term 13:13:40 ACTION DanC: propose to postpone 13:14:14 in a week we'll either have good reason to talk more or we won't. 13:16:02 graphSolutionMapping 13:16:20 was angenls on head of pin thing, put it on issues list 13:16:41 andy doesnt see any problems 13:17:24 RESOLVED to close graphSolutionMapping per SPARQL 1.162/1.171. 13:17:45 RESOLVED to accept wording 13:17:45 int rq23 13:19:37 nestedOptionals 13:20:06 danc: how does it extend the expressive capability of the language 13:25:13 http://www.w3.org/2001/sw/DataAccess/rq23/#OptionalMatchingNested where does the 'mbox' column come from to the result? magic 13:29:20 PREFIX foaf: 13:29:20 PREFIX vcard: 13:29:20 SELECT ?foafName ?gname ?fname 13:29:20 WHERE ( ?x foaf:name ?foafName ) 13:29:20 [ ( ?x foaf:mbox ?mbox ) ] 13:29:21 [ ( ?x vcard:N ?vc ) ( ?vc vcard:Given ?gname ) 13:29:23 [ ( ?vc vcard:Family ?fname ) AND BOUND ?mbox ] 13:29:40 ---- oops 13:30:42 PREFIX foaf: 13:30:42 PREFIX vcard: 13:30:42 SELECT ?foafName ?gname ?fname 13:30:42 WHERE ( ?x foaf:name ?foafName ) 13:30:42 [ ( ?x foaf:mbox ?mbox ) ] 13:30:43 [ ( ?x vcard:N ?vc ) ( ?vc vcard:Given ?gname ) ] 13:30:45 [ ( ?vc vcard:Family ?fname ) AND BOUND ?mbox ] 13:31:37 -------- oops 13:31:48 PREFIX foaf: 13:31:49 PREFIX vcard: 13:31:49 SELECT ?foafName ?gname ?fname 13:31:49 WHERE ( ?x foaf:name ?foafName ) 13:31:49 [ ( ?x foaf:mbox ?mbox ) ] 13:31:49 [ ( ?x vcard:N ?vc ) ( ?vc vcard:Given ?gname ) ] 13:31:51 [ ( ?vc vcard:Family ?fname ) AND BOUND ?vc ] 13:39:33 PROPOSED: to adopt sparql rq23 v 1.171 as addressing nestedOptionals 13:40:09 objections: SH, KC, AR; abstaining Yoshio 13:40:25 in favor: AS, EP, HS 13:40:34 abstain: DB 13:40:53 in favor: JS 13:41:06 no carried. 13:42:02 KC: too hard to read/understand. (syntax discussion was curtailed). benefits outweighed by the cost. 13:45:20 So I'd be willing to merely abstain if the syntax were less confusing/ugly/hard-to-teach 13:48:43 would be easier if we had notion of local vaiables, which I don't expect to be adopted... 13:49:42 options: rq23 1.171 / side condition on the grammar: nested optionals not allowed. 13:50:00 pref nested as per 1.171: 3 13:50:10 pref side side: 4 13:51:42 opposes nested optionals as per 1.171: l3 13:52:00 oppose no nested optional: 2 13:52:03 s/13/3/ 13:54:48 consUnboundVars 14:04:15 what if the variable is not optional? is it still successfull? 14:05:20 general feeling towards "may return a warning" 14:05:21 warniings in the rdf/xml comments, x-http-, n3 comments 14:05:34 for the case that the user gets unexpectedly less triples with construct 14:05:44 unbound variable are deliberte - user used optionals or UNION with different path 14:06:05 intended not to have some triples be realisable 14:06:10 PROPOSED: to address consUI add "... may issue a warning" in section 14:06:47 PROPOSED: to address consUnboundVars by adding "... may issue a warning" in section 10.2 Constructing an Output Graph, and discussing warnings in protocol and result set specs (xml comments, http headers, etc.) 14:08:13 PROPOSED: to address consUnboundVars by noting "... may issue a warning... " in section 10.2 Constructing an Output Graph, and discussing warnings in protocol and result set specs (xml comments, http headers, etc.) 14:09:16 PROPOSED: to address consUnboundVars by changing "... may issue an error or warning... " to "... may issue a warning..." in section 10.2 Constructing an Output Graph, and discussing warnings in protocol and result set specs (xml comments, http headers, etc.) 14:10:04 so RESOLVED. ACTION AndyS, DaveB, KendallC 14:10:29 consWithBnodes 14:12:10 andy: re sandro's comment 14:12:26 andy: my usecase is vcard, needs bnodes 14:12:56 typo on http://www.w3.org/2001/sw/DataAccess/rq23/#construct 3rd paragraph "predciate position" -> "predicate position" 14:19:10 PROPOSED: to close consUnboundVars ala section 10.2 Constructing an Output Graph 14:19:10 1.133 / 1.171 14:19:43 so RESOLVED. 14:21:31 kendall has joined #dawg 14:24:33 bijan has joined #dawg 14:28:58 Zakim has left #dawg 14:31:24 Topic: disjunction 14:31:30 ... review of history... 14:31:36 ... discussion of SQL union ... 14:36:11 steve, SQL UNION http://lists.w3.org/Archives/Public/public-rdf-dawg/2004OctDec/0294.html 14:39:56 DanC: asks about interaction with SOURCE (now GRAPH) 14:43:16 PROPOSED: to resolve disjuction issue ala v 1.171 of rq23 section 6 More Pattern Matching – Alternatives 14:44:09 OBJECTION: DaveB; ABSTAINING: SH, YF 14:44:19 so RESOLVED 14:45:17 (pretty sure earlier decisions addressed useMentionOp) 14:46:36 I dug up an algae union disjunction test --> http://dev.w3.org/cvsweb/perl/modules/W3C/Rdf/test/Disjunction1-alg.sh?rev=HEAD&content-type=text/x-cvsweb-markup 14:49:01 EP reviews algae federation use case... 14:51:42 what if a variable is bound to a b-node? 14:51:46 if you have SELECT ?foo FROM WHERE....() --- cascade ---> FROM WHERE (...) -- cascade ---> FROM WHERE (...) makes sense though 14:52:15 or combinations of LOAD/FROM in new syntax 14:53:06 PROPOSED: to postpone cascadedQueries; while federation use cases are interesting, the designs don't seem mature and the use cases are not urgent. 14:54:14 ABSTAIN: KC. so RESOLVEd. 14:54:42 Topic: valueTesting 15:00:15 ok - I implemented (only few fully tested though) most of the SPARQL 1.71 doc ( in perl - code at http://cvs.sourceforge.net/viewcvs.py/rdfstore/rdfstore/lib/DBD/RDFStore.pm?rev=1.79&view=markup the numerical and date ranges can be pushed down to DB level ) 15:00:45 dateTimes with timezones, too? 15:00:56 ( package DBD::RDFStore::st::constraints; in particular ) 15:01:10 JanneS: not yet - 15:02:26 referring to http://www.w3.org/2001/sw/DataAccess/rq23/#tests part of the SPARQL query spec 15:03:06 functions and operators available in SPARQL: 15:03:08 (6.2.7) op:numeric-unary-plus 15:03:08 (6.2.8) op:numeric-unary-minus 15:03:08 (6.3.1) op:numeric-equal 15:03:08 (6.3.2) op:numeric-less-than 15:03:11 (6.3.3) op:numeric-greater-than 15:03:13 (6.2.1) op:numeric-add 15:03:16 (6.2.2) op:numeric-subtract 15:03:18 (6.2.3) op:numeric-multiply 15:03:21 (6.2.4) op:numeric-divide 15:03:23 (7.2.3) op:comapre 15:03:26 (7.2.4) op:matches 15:03:53 what about sparql: prefix into v. 1.71 of document ? 15:07:47 ep: numeric types: ... 15:08:09 as: owl uses integers (arbitrary precision)... 15:08:36 as: double, decimal, integer 15:11:31 PROPOSED: date less-than, date-greater-than, date-equal; string ops per 1.171; numeric double, integer, decimal, 9 numeric ops above 15:13:06 "For xs:decimal values the number of digits of precision returned by the numeric operators is ·implementation-defined·. I" 15:13:09 DB: xquery says implementation defined precision, not arbitrary precision 15:13:26 just before end of 6.2 http://www.w3.org/TR/xpath-functions/#op.numeric 15:13:34 for dates I think we could do with 10.3.10 op:date-equal 15:13:34 10.3.11 op:date-less-than 15:13:35 10.3.12 op:date-greater-than 15:13:43 from http://www.w3.org/TR/xpath-functions/ that is 15:14:02 or in fact 15:14:03 10.3.7 op:dateTime-equal 15:14:03 10.3.8 op:dateTime-less-than 15:14:03 10.3.9 op:dateTime-greater-than 15:14:11 that includes time component with date 15:14:15 +1 janneS - we do that 15:14:43 + less/grater-than-equal 15:16:04 PROPOSED: date less-than, date-greater-than, date-equal; string ops per 1.171; numeric double, integer, decimal, 9 numeric ops above, per xquery f&o 15:16:12 AS: in any case, no duration 15:18:06 PROPOSED: datetime, date, time; date less-than, date-greater-than, date-equal; string ops per 1.171; numeric double, integer, 9 numeric ops above, per xquery f&o 15:18:50 http://www.w3.org/TR/xpath-functions/#datatypes 15:18:56 EP and others (KC) express reservations about deciding today. EP offers to research and propose 15:20:57 See http://www.w3.org/TR/xpath-functions/#casting-from-primitive-to-primitive 15:21:00 EP and others (KC) express reservations about deciding today. EP offers to research and propose 15:21:03 PROPOSED: datetime, date, time; date less-than, date-greater-than, date-equal; string ops per 1.171; numeric double, integer, 9 numeric ops above, per xquery f&o 15:22:18 ABSTAIN, EP, KC, SH. so RESOLVED. 15:25:15