IRC log of dawg on 2005-02-01

Timestamps are in UTC.

14:31:20 [RRSAgent]
RRSAgent has joined #dawg
14:31:20 [RRSAgent]
is logging to http://www.w3.org/2005/02/01-dawg-irc
14:31:30 [dajobe-jango]
dajobe-jango has joined #dawg
14:31:32 [DanC]
agenda + XML Result set
14:31:41 [DanC]
agenda + test update
14:31:49 [DanC]
agenda + SPARQL Protocol Spec
14:31:56 [AlbertoR]
AlbertoR has joined #dawg
14:31:58 [DanC]
agenda + SPARQL QL (rq23) update
14:32:03 [Zakim]
+JosD
14:32:06 [Zakim]
+[ASemantics]
14:32:08 [Zakim]
+DanC
14:32:16 [AndyS]
zakim, who is on the phone?
14:32:16 [Zakim]
On the phone I see ??P0, Yoshio, Kevin, ??P2, JosD, [ASemantics], DanC
14:32:28 [AndyS]
zakim, ??P0 is AndyS
14:32:28 [Zakim]
+AndyS; got it
14:32:31 [DanC]
Zakim, ??P2 is DaveB
14:32:31 [Zakim]
+DaveB; got it
14:32:58 [DanC]
Zakim, ??P0 is AndyS
14:32:58 [Zakim]
I already had ??P0 as AndyS, DanC
14:33:03 [Zakim]
+??P5
14:33:05 [DanC]
Zakim, take up agendum 1
14:33:05 [Zakim]
agendum 1. "Convene, take roll, review records and agenda http://lists.w3.org/Archives/Public/public-rdf-dawg/2005JanMar/0076.html" taken up [from DanC]
14:33:11 [SteveH]
Zakim, ??P5 is SteveH
14:33:11 [Zakim]
+SteveH; got it
14:33:43 [AlbertoR]
regrets Kendall Clark
14:33:45 [DanC]
Regrets: KendallC
14:33:51 [DanC]
Regrets+ Howard
14:34:17 [DanC]
RRSAgent, please draft minutes
14:35:07 [DanC]
Scribe: AlbertoR
14:35:26 [Zakim]
+EricP
14:35:29 [DanC]
not done: http://www.w3.org/2001/sw/DataAccess/ftf4.html $Revision: 1.40 $
14:36:20 [DanC]
6 of 25 items summarized
14:36:46 [DanC]
next meeting: Tuesday, February 8, 2005
14:36:54 [DanC]
partial regrets AndyS
14:36:58 [DanC]
next scribe: EricP
14:36:58 [AlbertoR]
EricP will scribe 08/02/2005
14:37:20 [DanC]
ACTION DanC: to (with EricP) communicate WBS user experience to systems team.
14:37:21 [Zakim]
+PatH
14:38:03 [DanC]
RRSAgent, please draft minutes
14:38:14 [patH]
patH has joined #dawg
14:38:35 [DanC]
Meeting: RDF Data Access WG Teleconference
14:38:38 [DanC]
Chair: DanC
14:38:41 [DanC]
RRSAgent, please draft minutes
14:39:18 [DanC]
Zakim, next item
14:39:18 [Zakim]
agendum 2. "ftf meetings " taken up [from DanC]
14:39:32 [DanC]
- ACTION AlbertoR: to propose a f2f meeting in Leiden http://lists.w3.org/Archives/Public/public-rdf-dawg/2005JanMar/0072.html
14:40:20 [DaveB-lap]
www2005 is 10-14 May
14:40:33 [DanC]
"10-14 May 2005 Chiba, Japan
14:40:33 [DanC]
WWW2005" -- http://www.w3.org/2001/sw/CG/#future
14:41:08 [AlbertoR]
ok proposed f2f 6th will be in between www2005 and XTech2005
14:41:30 [DanC]
ACTION DanC: consider Asemantics hosting proposal w.r.t. WG schedule
14:41:58 [AlbertoR]
SteveH was also willing to offer to host next face-to-face
14:42:38 [DanC]
ACTION SteveH: propose to host in Southampton right after XTech
14:42:38 [AlbertoR]
ACTION: SteveH will make another offer for next face-to-face
14:42:57 [DanC]
ACTION PatH: propose to host in Florida
14:43:03 [AlbertoR]
ACTION -4
14:44:27 [DanC]
Zakim, next item
14:44:27 [Zakim]
agendum 3. "valueTesting, XML Schema datatypes" taken up [from DanC]
14:44:42 [DanC]
XML Schema Datatypes and the Semantic Web Jeremy Carroll (Monday, 31
14:44:42 [DanC]
January)
14:44:42 [DanC]
http://lists.w3.org/Archives/Public/public-rdf-dawg/2005JanMar/0071.html
14:44:56 [DanC]
http://www.w3.org/2001/sw/BestPractices/XSCH/xsch-sw-20050127/
14:46:33 [AlbertoR]
SWBP requested review of the XSCH document: EricP and PathH
14:46:54 [AlbertoR]
ACTION: EricP and PathH will review XSCH draft document
14:48:54 [AlbertoR]
the two reviewers will first report to DAWG - comments will most propably arrive after publication
14:49:02 [DanC]
Zakim, next agendum
14:49:02 [Zakim]
agendum 4. "accessing collections" taken up [from DanC]
14:49:19 [DanC]
Zakim, who's on the phone?
14:49:19 [Zakim]
On the phone I see AndyS, Yoshio, Kevin, DaveB, JosD, [ASemantics], DanC, SteveH, EricP, PatH
14:50:02 [AlbertoR]
Jos test-case http://lists.w3.org/Archives/Public/public-rdf-dawg/2005JanMar/0075.html
14:50:20 [AlbertoR]
oppss..http://lists.w3.org/Archives/Public/public-rdf-dawg/2005JanMar/0077.html
14:51:26 [DaveB-lap]
for me ,postpone
14:51:30 [SteveH]
+1
14:51:37 [AlbertoR]
+1
14:51:42 [DanC]
postpone, pick an o:item URI, make syntax
14:51:52 [DanC]
^options
14:52:17 [DanC]
yoshio, which do you prefer/like? postpone, pick an o:item URI, make syntax
14:52:49 [SteveH]
zakim, mute me
14:52:49 [Zakim]
SteveH should now be muted
14:53:05 [AlbertoR]
about to postpone accessing-collections:
14:53:57 [SteveH]
zakim, unmute me
14:53:57 [Zakim]
SteveH should no longer be muted
14:54:04 [SteveH]
zakim, mute me
14:54:04 [Zakim]
SteveH should now be muted
14:54:31 [AlbertoR]
postpone: 6
14:54:34 [DanC]
postpone: 1+1+1+1+1+1 -1; pick uri: 1+1
14:54:34 [AlbertoR]
owl:item: 2
14:55:00 [Yoshio]
0
14:57:02 [SteveH]
not really realted
14:58:22 [AlbertoR]
DanC: cwm seems/is able to deal with collections nicely - it might be stated into the service description
15:00:28 [AlbertoR]
EricP: mentioned a possible issue related to the interaction of collections (or inferred triples) and result-set format
15:00:51 [AndyS]
There is also the &ns:func() mechanism
15:02:11 [SteveH]
I'm interested, but still think its early
15:03:07 [AlbertoR]
JosD and PatH (Andy parially?) agree on access-collection importance -
15:04:26 [DanC]
Zakim, take up item 7
15:04:26 [Zakim]
agendum 7. "SPARQL Protocol Spec" taken up [from DanC]
15:04:41 [DanC]
ACTION KendallC: Announce the protocol spec
15:04:49 [DanC]
ACTION EricP: to inviestigate WSDL/SOAP encodings of SPARQL protocol
15:05:03 [AlbertoR]
Kendall, EricP actions continued
15:05:06 [DanC]
ericp: preempted by datatype work
15:05:15 [DanC]
ACTION KendallC: to propose a solution to protocolRootReferent
15:05:24 [DanC]
. ACTION DanC: write up "sparqlParam" approach
15:06:09 [DanC]
ACTION DanC: write up "sparqlParam" approach (reply to kendallC)
15:06:47 [DanC]
Some protocol & service description issues http://lists.w3.org/Archives/Public/public-rdf-dawg/2005JanMar/0046.html
15:07:12 [SteveH]
no
15:07:14 [DanC]
ACTION SteveH: Write up a service description of features supported for his service
15:07:33 [DanC]
ACTION KendallC: Add section for service descriptions and put graph specifying in it to start it off (some progress, it seems. continues)
15:07:45 [DanC]
ACTION DanC: Write up definitive service concept
15:08:26 [DanC]
ACTION KendallC: to think about predicate to relate html-forms and svc descriptions, and how this goes into the spec
15:08:55 [DanC]
ACTION DanC: discuss accessingCollections in service description context
15:09:05 [DanC]
Zakim, close this agendum
15:09:05 [Zakim]
agendum 7 closed
15:09:06 [Zakim]
I see 4 items remaining on the agenda; the next one is
15:09:08 [Zakim]
4. accessing collections [from DanC]
15:09:12 [DanC]
Zakim, close item 4
15:09:12 [Zakim]
agendum 4 closed
15:09:13 [Zakim]
I see 3 items remaining on the agenda; the next one is
15:09:14 [Zakim]
5. XML Result set [from DanC]
15:09:21 [DanC]
Zakim, next item
15:09:21 [Zakim]
agendum 5. "XML Result set" taken up [from DanC]
15:09:30 [DanC]
ACTION DaveB: discuss options for column ordering and pick one
15:10:03 [SteveH]
+1 to query order
15:10:05 [DanC]
DaveB: seems "unconstrined" is unpopular, so I'm down to 2
15:10:39 [SteveH]
* would be mention order wouldn't it?
15:12:17 [SteveH]
unconstrained order is inconvienient for clients
15:12:26 [AndyS]
What's u2070?
15:12:27 [SteveH]
and little more work for servers
15:13:06 [AndyS]
But yes, the var name is currently NCNAMES - any problem there?
15:13:10 [AlbertoR]
AndyS: picked up one from BNF (ramdonly) - just wondering about Unicode and ordering in the result-set
15:13:12 [ericP]
intellectually, i prefer DanC's. aestheticaqlly, i prefer andy's
15:13:30 [SteveH]
SELECT order is very useful in some cases
15:13:32 [AndyS]
Not sort order!
15:13:39 [SteveH]
especially if you have SELECT expressions!
15:13:48 [AlbertoR]
+1 SteveH
15:13:56 [ericP]
+1 SteveH
15:14:49 [AndyS]
+1 to SELECT order if SELECT ?a ?z ?c
15:15:12 [AndyS]
If SELECT * - don't care but can't see value of mention order
15:15:37 [SteveH]
AndyS, why not? its helpful sometimes
15:15:51 [SteveH]
I think SQL does specify, but havent checked
15:16:03 [AndyS]
It's work - assumes (1) you parsed it and (2) you parsed in order :-)
15:16:26 [SteveH]
AndyS, yes, but easier for users - ie. predictable webween systems
15:16:29 [DanC]
PROPOSED: order in result set is: if SELECT a,b, c then a,b,c; if SELECT * then unconstrained
15:16:30 [SteveH]
*between
15:16:51 [SteveH]
mildly
15:16:58 [SteveH]
I prefer mention order
15:17:06 [SteveH]
Zakim, unmute me
15:17:06 [Zakim]
SteveH should no longer be muted
15:17:56 [DanC]
SteveH argues for first mentioned order in SELECT *
15:18:09 [DanC]
EricP: mysql seems to use schema order in case of SELECT *
15:19:45 [DanC]
SteveH notes that lexicographic order is complicated for expressions
15:20:11 [AlbertoR]
if SELECT *would be sorted in query order - which one? (considering optimizers and arbitrary grap-patterns processing)
15:20:17 [ericP]
why is nothing easy?
15:20:41 [DanC]
PROPOSED: order in result set is: if SELECT a,b, c then a,b,c; if SELECT * then unconstrained
15:21:11 [DanC]
so RESOLVED. SteveH abstaining
15:21:18 [ericP]
ok, i guess that was comparitively easy
15:21:37 [DaveB-lap]
http://lists.w3.org/Archives/Public/www-rdf-interest/2005Jan/0103.html
15:21:48 [AlbertoR]
ACTION -16
15:21:52 [DanC]
DONE: ACTION DaveB: to post announcement about XML result set on appropriate mailing lists.
15:22:02 [DanC]
ACTION DanC: ask Liam, via SemWeb CG, which XML groups might be interested in SPARQL result set WD
15:22:03 [SteveH]
Zakim, mute me
15:22:03 [Zakim]
SteveH should now be muted
15:22:42 [SteveH]
i've implemeted much of it
15:23:04 [DanC]
several seem to have implemented it, or started anyway
15:23:54 [DanC]
Turtle Tuples: Turtle-based query result format Arjohn Kampman (Monday, 24 January) http://lists.w3.org/Archives/Public/public-rdf-dawg-comments/2005Jan/0019.html
15:24:02 [SteveH]
I have a turtleish result for in the current version of my store - its good for low bandwidhth but not hugely important
15:24:45 [DanC]
Zakim, next agendum
15:24:45 [Zakim]
agendum 6. "test update" taken up [from DanC]
15:24:52 [SteveH]
its very like arjohns
15:25:14 [AndyS]
I had a sim format - and dropped it 'cos of maintenance overhead for now
15:25:16 [DaveB-lap]
bound tests in http://lists.w3.org/Archives/Public/public-rdf-dawg/2005JanMar/0080.html
15:25:22 [DanC]
re ACTION DaveB: to draft tests for BOUND (based on ericp's examples)
15:25:25 [DaveB-lap]
AndyS found a bug in the 0080
15:26:26 [AlbertoR]
DaveB: will make a new direcotry in the test cases rep for BOUND
15:26:40 [AlbertoR]
All other actions continued
15:26:43 [DanC]
ACTION SteveH: to propose tests for str() and =~
15:26:50 [DanC]
ACTION SteveH: to revise test manifest w.r.t. "background" and named graphs
15:26:55 [DanC]
ACTION DaveB: to propose source test to approve
15:26:59 [DanC]
ACTION AndyS: to propose test case re optionals and ordering
15:27:04 [DanC]
Zakim, next agendum
15:27:04 [Zakim]
agendum 8. "SPARQL QL (rq23) update" taken up [from DanC]
15:27:41 [DanC]
http://www.w3.org/2001/sw/DataAccess/rq23/ $Revision: 1.187 $ of $Date: 2005/02/01 10:56:24 $
15:28:41 [DanC]
ACTION EricP: to update rq23 w.r.t. str()
15:29:13 [DanC]
ACTION EricP: to update rq23 w.r.t. BOUND/UNSAID (some progress; UNSAID removed)
15:30:39 [DanC]
EricP seems to have new info re funcs/ops. will send.
15:31:19 [AlbertoR]
AndyS: op:anyURI is literal with a URI datatype
15:31:36 [DanC]
ScribeNick: DanC
15:31:39 [DanC]
ScribeNick: AlbertoR
15:33:18 [AlbertoR]
ACTION AndyS: to revise "unnamed" terminology
15:33:30 [AlbertoR]
ACTIOn -24
15:33:36 [AlbertoR]
ACTION -24
15:33:47 [DanC]
hmm... "8.4 GRAPH and a single, unnamed graph"
15:34:07 [DanC]
ACTION AndyS: to revise "unnamed" terminology (continues after all)
15:34:47 [DanC]
ADJOURN.
15:34:49 [Zakim]
-EricP
15:34:52 [Zakim]
-DaveB
15:34:53 [SteveH]
bye
15:34:56 [Zakim]
-SteveH
15:34:56 [DanC]
RRSAgent, please draft minutes
15:35:06 [Zakim]
-Yoshio
15:36:27 [ericP]
W3C has 3 recommendations in the area of selecting from XML infosets: XPath, XQuery and XSLT. While RDF query addresses a different data domain (the RDF graph), any query mechanisms defined by this group must leverage off, and, where possible, interoperate with, the related XML recommendations.
15:36:29 [AlbertoR]
ok I will
15:36:32 [ericP]
XPath
15:36:34 [ericP]
The XPath language defines a non-XML syntax for addressing subtrees from XML documents. It factors out the selection functionality common to XQuery and XSLT. It is hoped that RDF Query will provide analogous functionality for addressing subgraphs. The working group will adopt, or define relationships with, applicable portions of the XPath data model.
15:36:39 [ericP]
XQuery
15:36:42 [ericP]
The Relationship to XQuery section defines the interactions with XQuery.
15:36:44 [ericP]
XPath relies on XSD to define the datatypes on which the functions and operators work
15:36:53 [Zakim]
-[ASemantics]
15:36:56 [Zakim]
-JosD
15:37:33 [Zakim]
-Kevin
15:37:35 [Zakim]
-AndyS
15:37:41 [KevinW]
KevinW has left #dawg
15:38:47 [Zakim]
-DanC
15:40:27 [AndyS]
AndyS has left #dawg
16:35:00 [Zakim]
disconnecting the lone participant, PatH, in SW_DAWG()9:30AM
16:35:02 [Zakim]
SW_DAWG()9:30AM has ended
16:35:05 [Zakim]
Attendees were Yoshio, Kevin, JosD, [ASemantics], DanC, AndyS, DaveB, SteveH, EricP, PatH
16:59:08 [DaveB]
I note ConstructGroup is not used in rq23 grammar
17:16:52 [AndyS]
DaveB - noted - it was a late removal (Alberto suggested) and clearly I didn't remove it all
17:17:06 [AndyS]
I have been trying the bound test case
17:17:14 [DaveB]
so have I
17:17:15 [DaveB]
it doesn't work
17:17:31 [AndyS]
Removed ConstructGroup
17:17:57 [AndyS]
I get a different answer.
17:18:11 [AndyS]
-------------
17:18:11 [AndyS]
| a | c |
17:18:11 [AndyS]
=============
17:18:11 [AndyS]
| :c2 | :f |
17:18:11 [AndyS]
| :a2 | :c2 |
17:18:12 [AndyS]
-------------
17:18:38 [AndyS]
which looks correct but even on this small example it's hard to be sure
17:18:59 [AndyS]
I fixed the manifest, data and query. Presumably you have too.
17:20:23 [DaveB]
not yet
17:20:39 [DaveB]
the engine's getting the wrong answer (3 items) but does give the 2 correct ones
17:21:28 [DaveB]
the ones above
17:22:02 [AndyS]
I had to fix a few things first - my engine was unhappy about unbound variables in a more serious way than intended.
17:26:57 [DaveB]
let me add the test to a dir 'bound'
17:27:23 [AndyS]
Please fix manifest/data/query for synatx first please
17:27:37 [DaveB]
what?
17:28:08 [DaveB]
manifest.n3
17:28:14 [DaveB]
mf:action
17:28:14 [DaveB]
[ qt:query <bound1.rq> ;
17:28:14 [DaveB]
qt:data <data.n3> ] ;
17:28:14 [DaveB]
mf:result <bound1-result-001.n3>
17:28:25 [DaveB]
hmm, that's wrong too
17:28:34 [AndyS]
That is not as per email.
17:29:17 [AndyS]
e.g "qt:query <bound1-001>"
17:29:30 [DaveB]
yeah, well it's broken
17:29:40 [DaveB]
let me cvs add something see what it is
17:30:36 [DaveB]
try cvs up
17:30:45 [DaveB]
I hardly expect the web to work ACL-wise, but I'll check
17:31:08 [DaveB]
seems to http://www.w3.org/2001/sw/DataAccess/tests/data/bound/
17:32:56 [Zakim]
Zakim has left #dawg
17:33:12 [DaveB]
RRSAgent, off