minutes: RDF Data Access WG 19Oct (non-IRC version)

On 23/10/2004, at 4:19, Dan Connolly wrote:

> http://www.w3.org/2004/10/19-dawg-irc
> If we don't get a record better than that for
> last week's telcon by the time I do the agenda
> Monday morning, I'm likely to just use the IRC log.
> (Simon, poke poke, again.)

Shamefacedly and much belatedly, I offer the following:

Minutes of RDF Data Access Working Group  
(http://www.w3.org/2001/sw/DataAccess/) teleconference for Tuesday  
2004-10-19T14:30:00Z, for review.

Supplementary chat log from irc://irc.w3.org:6665/dawg available at  

Scribe: Simon Raboczi

1. Convene, take roll, review record and agenda

roll call

   Tom Adams
   Dave Beckett
   Kendall Clark
   Dan Connolly
   Yoshio Fukushige
   Steve Harris
   Pat Hayes
   Eric Prud'hommeaux
   Simon Raboczi
   Alberto Reggiori
   Hiroyuki Sato
   Andy Seaborne

   Jos De Roo
   Howard Katz
   Farrukh Najmi

PROPOSED: to accept minutes for 10-12-2004 meeting
as a true record of our meeting 2004-10-12
RESOLVED to adopt  
with two amendments:
1) the re-addition of Kendall's ACTION to produce a protocol draft
2) Alberto's amendments from  

next meeting: Tuesday, October 26, 2004
Tom Adams tentatively volunteers to scribe, with Alberto Reggiori as  

Sundry ACTIONs not warranting their own agenda:

ACTION KendallC: expose walking tour data to SPARQL querying clients
CONTINUES without discussion, ETA before F2F4

ACTION EricP: find logistics re F2F5 at tech plenary in Boston March
  28 Feb - 4 Mar (2 days).  Some WG preference to Mon/Tue of that week.
EricP responds in
28 February - 4 March 2005 (Monday - Friday)
meetings Monday/Tuesday 28 Feb-1 March and Thursday/Friday 3-4 March
Hyatt Harborside Hotel  
Boston, MA, USA

ACTION DanC: talk with Kendall about issues list maintenance

2. Publication of SPARQL design and Data Access UC&R

Thanks offered to EricP, AndyS, Kendall and all.

  SPARQL Query Language for RDF

  RDF Data Access Use Cases and Requirements

AndyS suggests another WD circa December, before the January meeting.
EricP: W3C December holidays publishing moritorium: last request on  
20-Dec, last pub: 22-Dec

3. Feedback on the SPARQL design

  [Fwd: Coments on first working draft of SPARQL]
  Seaborne, Andy (Sunday, 17 October)
and a few others.

Peter Patel-Schneider commented that "bound" is something procedural --  
AndyS to discuss in email.
Peter P-S raised an issue about the definition of subgraph.
DanC suggests asking for a test case to clarify the subgraph issue.

ACTION: PatH review SPARQL def'ns post-publication

ACTION Dirk: provide details about DB interfaces, re '?' and '$'
CONTINUES for one more week -- proposed to drop if no progress by then

To what extent are these issues addressed by the 2004-Oct-12 drafts?

ACTION AlbertoR: owns issue 'DESCRIBE'
AlbertoR hasn't gone through the draft yet to be able to determine  
whether the DESCRIBE issue is addressed by the 12Oct draft. AlbertoR  
expects he'll take 2-3 weeks to get to it.
AndyS believes that there is more work to be done on DESCRIBE. AndyS  
intends to post to the list.  AndyS would like the retract the CBD  
suggestion in section 11.3 of the WD because CBD is still fluid.

ACTION SteveH: own (i.e. propose resolution to) disjunction issue
SteveH on disjunction: he's currently only happy with removing  
patH: Disjunction is logically troublesome unless stated veeeery  
AndyS is inclined to re-include disjunction.

ACTION SteveH: owns issue 'nested optionals'
DaveB: end of section 5 has a nested optional example

ACTION: Tom to send sample iTQL queries and result formats for
subqueries vs. disjunction

ACTION DanC: owner of issue 'yes or no questions'
DanC said he's satisfied with the current spec, however Kendall doesn't  
feel the current ASK definition is satisfactory.
Eric proposed that this will be best handled with a protocol document  
on the table next to the query lang.

Kendall asserted the the present SPARQL ql draft satisfied him re:  
PREFIX, which he was issue owner of.
DaveB however then raised issues with prefix: prefix can current appear  
in two places in the grammar according to the production Query ::=  
PrefixDecl* ReportFormat PrefixDecl* and it's not specified in which  
order repeated instances override each other.
DaveB took an ACTION to illustrate prefix interaction details by  
DanC requests for all to be ready to resolve next week the PREFIX issue.

AndyS has done writing on value testing which he'd like to get comments  

AndyS is inclined to drop the UNSAID feature, based on it being  
Kendall and SimonR both have uses within their respective organizations  
for UNSAID and were reluctant to see it dropped.

4. Web Services Constraints and Capabilities

ACTION EricP: draft UC on overlap between RDF query and web service
constraints with respect to WS-Policy
DaveB requested a summary of the WS requirements (e.g. conjunction now,  
disjunction eventually), which Eric agreed to supply.
Eric noted that neither policies not capabilities are expressed in RDF.


ACTION: SteveH to turn his SOURCE tests into SPARQL ones

dawg-source-simple-* Steve Harris (Friday, 15 October)

ACTION DaveB: Update the source section 9, add more formal links,
update the examples, try to think about extra constraints as EricP
proposed (SOURCE ?s and ?s only in in SELECT). Look at various
people's source test cases.

DanC assessed that we have progress on SOURCE tests, but there's still  
haggling in progress on the actual definition.
DaveB noted that the current syntax contains all the syntax for named  
DanC asks whether the various different designs for SOURCE are indeed  
equivalent or not?
SteveH replied negatively, that he can produce a test case that would  
behave differently between his implementation and AndyS's.
SteveH: Andy and I disagree about  
Discussion continued on until after the meeting about whether the  
current behavior in the tests was tenable (i.e. that returning  
duplicate rows in a result is different from a result with just one  
copy of the same row)

Meeting adjourned at 15:38:00Z.

Recapping, for next time:
- Dan wants us to be able to resolve PREFIX at the next teleconference.
- Andy wants feedback on his value testing material.

Received on Monday, 25 October 2004 16:31:14 UTC