IRC log of dawg on 2005-01-20

Timestamps are in UTC.

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