IRC log of dawg on 2005-06-07

Timestamps are in UTC.

14:30:58 [RRSAgent]
RRSAgent has joined #dawg
14:30:58 [RRSAgent]
logging to
14:31:06 [DaveB]
Zakim, who's on the phone?
14:31:06 [Zakim]
On the phone I see PatH, ??P19, ??P20
14:31:07 [Zakim]
14:31:08 [Zakim]
14:31:08 [patH]
14:31:13 [DaveB]
Zakim, ??P19 is DaveB
14:31:13 [Zakim]
+DaveB; got it
14:31:14 [AndyS]
??P20 is AndyS
14:31:20 [AndyS]
zakim, ??P20 is AndyS
14:31:20 [Zakim]
+AndyS; got it
14:31:21 [DanC]
a day or two is fine, pat
14:31:27 [patH]
14:31:44 [DanC]
DanC has changed the topic to: RDF Data Access 7 Jun. scribe: PatH
14:31:51 [kendall]
zakim, mute me
14:31:51 [Zakim]
Kendall_Clark should now be muted
14:31:53 [Zakim]
14:31:55 [Zakim]
14:32:02 [DanC]
Zakim, take up item 1
14:32:02 [Zakim]
agendum 1. "Convene, take roll, review records and agenda" taken up [from DanC]
14:32:06 [kendall]
It will make the sound of pending extinction :<
14:32:09 [DanC]
Zakim, who is on the phone?
14:32:09 [Zakim]
On the phone I see PatH, DaveB, AndyS, Kendall_Clark (muted), DanC, ??P24, EricP
14:32:17 [kendall]
zakim, unmute me
14:32:17 [Zakim]
Kendall_Clark should no longer be muted
14:32:54 [Zakim]
14:33:01 [DanC]
Regrets: Howard, Yoshio Fukushige
14:33:34 [JanneS]
JanneS has joined #dawg
14:33:36 [Yoshio]
Yoshio is on IRC
14:34:16 [DanC]
Zakim, ??P24 is JanneS
14:34:16 [Zakim]
+JanneS; got it
14:34:41 [kendall]
sorry I generally avoid talking about politics, but mentioning Friedman -- who never met an official ass he wouldn't kiss -- is provocation too much! :>
14:34:48 [DanC]
RRSAgent, pointer?
14:34:48 [RRSAgent]
14:34:53 [SteveH]
DanC, sorry, phone sortage, skyping asap
14:34:59 [DanC]
14:35:28 [DanC]
PROPSED: to accept as a true record
14:35:37 [Zakim]
14:35:47 [DanC]
Zakim, IPcaller is SteveH
14:35:47 [Zakim]
+SteveH; got it
14:35:49 [SteveH]
Zakim, mute me
14:35:49 [Zakim]
SteveH should now be muted
14:35:55 [DanC]
14:36:20 [SteveH]
I rrobably cant make 14 june
14:36:24 [SteveH]
14:36:39 [DanC]
agenda + next meeting
14:36:49 [kendall]
zakim, mute me
14:36:49 [Zakim]
Kendall_Clark should now be muted
14:37:02 [patH]
I will have to miss 14th also.
14:37:37 [DanC]
Zakim, next agendum
14:37:37 [Zakim]
agendum 2. "issue wsdlAbstractProtocol" taken up [from DanC]
14:37:59 [DanC]
my suggestion
14:38:13 [kendall]
zakim, unmute me
14:38:13 [Zakim]
Kendall_Clark should no longer be muted
14:38:30 [patH]
Kendall knows what it says.
14:38:38 [kendall]
oh, don't scribe that. :>
14:38:43 [patH]
14:39:41 [kendall]
I don't understand the motivation to avoid the union, frankly. Certainly not at the cost of doubling the number of operations. Or, if Jeen is right, trebling them. :>
14:41:34 [DanC]
splitting? 3ish
14:41:44 [DaveB]
hmm, if I get my QueryAndTransform interface, does that 2x it again?
14:41:45 [patH]
Eric+Dan+jeanne like splitting
14:42:04 [ericP]
+maybe daveb, iirc
14:42:15 [DaveB]
I'm moving to neutral really
14:42:16 [patH]
Thanx, jeen
14:43:44 [patH]
jeen: aligns with our implementation, will send info
14:44:41 [patH]
DanC: Yes, Kendall, issue is abstract, editor's call
14:45:09 [DaveB]
14:45:40 [kendall]
well, my point about "code smell" is that it *could* be not entirely arbitrary, in which case we should know that.
14:46:41 [DaveB]
14:47:53 [kendall]
I have too FWIW
14:48:02 [patH]
DaveB worries about number of operations, Kendall agrees? but isnt worried. XSLT is considered but nobody cares too much.
14:49:49 [DanC]
PROPOSED: to close wsdlAbstractProtocol ala 1.51 2005/06/06 18:35:00 , i.e. one interface, one operation; HTTP keywords query, default-graph-uri, named-graph-uri
14:49:54 [patH]
kendall: design tries to look like ordinary Web service
14:49:56 [ericP]
+1 to KendallC
14:50:49 [patH]
DanC: close discussion; DaveB Second; no obj; DaveB abstains
14:50:49 [DanC]
RESOLVED, DaveB Jeen, SteveH abstaining
14:51:16 [AndyS]
The WSDL XML is broken apparently (
14:51:40 [kendall]
AndyS: there's a bare "&". oops. I've fixed it in my draft.
14:51:54 [DanC]
Zakim, next agendum
14:51:54 [Zakim]
agendum 3. "issue: fromUnionQuery" taken up [from DanC]
14:51:56 [AndyS]
OK - just so it does not get lost.
14:52:01 [kendall]
no, i saw it.
14:52:03 [kendall]
14:52:27 [DanC]
Zakim, take up item 4
14:52:27 [Zakim]
agendum 4. "issue valueTesting" taken up [from DanC]
14:52:56 [DanC]
section 1.373 $ of $Date: 2005/06/06 14:45:18 $
14:53:16 [kendall]
(oh, i forgot to mention that I have to leave by 11:45 my time; i.e., 75 minutes total)
14:53:17 [DaveB]
Yoshio: we skipped it till later
14:53:20 [SteveH]
zakim, unmute me
14:53:20 [Zakim]
SteveH should no longer be muted
14:53:28 [Yoshio]
14:53:30 [kendall]
till later in the telcon, i think :>
14:54:12 [ericP]
14:54:37 [DanC]
SH: not too happy about NaN being false. and ""
14:54:37 [DaveB]
SteveH some things are false, such as NaN values
14:54:37 [SteveH]
""@fr is false
14:54:41 [patH]
Steve not happy with some parts, empty strings
14:55:31 [patH]
with lang tags
14:56:37 [ericP]
14:56:48 [ericP]
14:57:18 [kendall]
Are we under attack?
14:57:26 [patH]
Dan might be.
14:58:05 [DaveB]
(near table 11.2 a broken link, the "Xquery F&O Primitive...[17]" link is 4040 )
14:59:06 [kendall]
I'm behind on this issue, FWIW.
14:59:12 [kendall]
er, FYI
14:59:29 [ericP]
loc:location [ r:type loc:geodetic ;
14:59:29 [ericP]
loc:rotational "30.300000"^^loc:latitude ;
14:59:37 [ericP]
loc:location [ r:type loc:EarthCenteredEarthFixed ;
14:59:37 [ericP]
loc:rotational "-7420507.1"^^loc:ECEF_X ;
14:59:55 [patH]
some values neither = not =/=
15:00:09 [patH]
15:01:06 [patH]
DaveB: solution is given, so?
15:01:18 [DaveB]
^- AndyS
15:01:28 [patH]
15:02:31 [patH]
The example is puzzling all round.
15:02:45 [DanC]
PROPOSED: to address valueTesting ala 1.373 2005/06/06 14:45:18 and
15:03:02 [DanC]
(approval of not for now)
15:03:07 [DaveB]
seems extendedType-eq-pass-result.n3 needs fixing
15:03:17 [DaveB]
should be identical to extendedType-eq-fail-result.n3
15:03:22 [DanC]
Zakim, who's on the phone?
15:03:22 [Zakim]
On the phone I see PatH, DaveB, AndyS, Kendall_Clark, DanC, JanneS, EricP, Jeen_Broekstra, SteveH
15:04:34 [patH]
AndyS likes to follow Xpath
15:06:35 [DanC]
RESOLVED, KC, PatH abstaining
15:07:15 [DaveB]
ericP: eq-pass has a typo - air: is not declared
15:07:41 [DanC]
ACTION EricP: finish extendedType-eq-pass-result.n3 test
15:07:53 [DaveB]
ericP: also eq-fail!
15:08:26 [ericP]
DaveB, roger
15:09:24 [SteveH]
zakim, mute me
15:09:24 [Zakim]
SteveH should now be muted
15:09:47 [DanC]
AndyS notes IRI/URI stuff still TODO editorially in section 11
15:09:56 [DanC]
Zakim, next agendum
15:09:56 [Zakim]
agendum 3. "issue: fromUnionQuery" taken up [from DanC]
15:10:42 [kendall]
zakim, mute me
15:10:42 [Zakim]
Kendall_Clark should now be muted
15:10:43 [kendall]
zakim, unmute me
15:10:43 [Zakim]
Kendall_Clark should no longer be muted
15:10:54 [kendall]
(coughing fit, sorry)
15:11:15 [patH]
Lots of mail. DanC summarises: 2 designs, no 3 designs.
15:11:42 [ericP]
DaveB, done -- tx
15:13:07 [patH]
First 2 designs in ed. draft: is closure queried or lower bound?
15:13:16 [DanC]
"9 Specifying RDF Datasets" has "@@Add text to say that specifying the dataset overrides an service dataset."
15:13:41 [patH]
3rd design given in email, aggregate graph.
15:13:48 [ericP]
15:14:29 [patH]
aggregate contains merge of named graphs.
15:15:39 [patH]
All servers would have to do things right.
15:18:24 [patH]
Yoshio, ignore that, I wasnt following the point.
15:18:58 [DanC]
OPTIONS: (a) without FROM/FROM_NAMED, dataset is unconstrained; with FROM/FROM_NAMED, dataset is bounded from below by given references. (b) like (a) but FROM/FROM named completely specify the dataset, (c) datasets have "aggregate graph" rather than background/default graph, and it always contains the merge of the named graphs
15:20:03 [patH]
ericP: agg. graph design requires giving name to trusted source/graph
15:20:20 [kendall]
(Clark's Conjecture: Semantic Web would *leap* ahead, as if by magic, if we could ban use of the word "trust".)
15:20:29 [DanC]
Zakim, who's on the phone?
15:20:29 [Zakim]
On the phone I see PatH, DaveB, AndyS, Kendall_Clark, DanC, JanneS, EricP, Jeen_Broekstra, SteveH (muted)
15:22:10 [Yoshio]
Does aggregate graph start empty or certain default graph?
15:22:25 [SteveH]
zakim, unmute me
15:22:25 [Zakim]
SteveH should no longer be muted
15:23:37 [patH]
agg. graph always contains the 'default' graph in the datatset, the G in {G, <n1,G1>...}
15:23:41 [kendall]
It may not matter, but I don't grok that claim, Dan.
15:24:06 [Yoshio]
thanks pat
15:24:25 [DanC]
SH: +agg -others
15:25:07 [DanC]
JB: +a ... RDFS ...
15:25:30 [DanC]
15:25:50 [DanC]
EP: +c, =a, =b
15:26:06 [DanC]
JS: -b, =a, =c
15:26:12 [kendall]
KC: +++a, object to b, don't understand c
15:26:42 [DanC]
AS: +a, =b ...
15:26:55 [DanC]
DB: +c =a =b
15:27:09 [DanC]
PH: +a =b =c
15:28:27 [DanC]
a and c seem to have no strong opposition; b is opposed by SH, JS, KC
15:28:42 [AndyS]
The authors revised it:
15:28:43 [jeen]
and by me as well
15:28:47 [JanneS]
* I will leave irc but will continue by phone for 10 more mins.
15:28:54 [kendall]
I heard Jeen oppose (b), I think.
15:28:55 [kendall]
ah, yes
15:29:03 [DanC]
b is opposed by several
15:29:12 [Yoshio]
15:29:29 [DanC]
counting... c a c a a c a . 3-3 tie
15:30:14 [DanC]
15:31:16 [DaveB]
I only saw 2x +c, me and eric, both =a =b
15:31:31 [jeen]
Allow me to summarize: option c is basically a more constrained version of option a. A leaves it up to the implementation/publisher what is in the default graph. C requires that all named graphs are in the default graph (or aggregate graph as it is called there). correct?
15:31:54 [patH]
Seems right, jeen.
15:32:14 [SteveH]
jeen, thats a matter of perspective, in some ways a is more constrinaing
15:33:00 [patH]
DaveB: RDFS wont work with c.
15:33:24 [DanC]
(my understanding is that Steve's implementation does C and does RDFS)
15:33:25 [ericP]
s/RDFS wont work with c/convince me that RDFS wont work with c/
15:33:27 [patH]
jeen: aaargh.
15:33:34 [kendall]
I oppose C, fwiw, because I think it should be entirely up to the provider of the graph to determine what goes in the graph. C, as I understand it, constrains the publisher of the graph with regard to what the publisher can, well, publish.
15:34:48 [ericP]
KendallC, C says that the default query (with no GRAPH constraints) queries the merge of all the named graphs
15:34:51 [AndyS]
Jeen: Fair summary
15:34:52 [DanC]
(sounds like sesame is meets the constraints in C too)
15:35:24 [patH]
DaveB: c not a, becasuz its simpler?
15:35:56 [patH]
jeen: yes, simpler; but issue concerns trust?
15:36:01 [kendall]
not all inputs are equal
15:36:09 [SteveH]
zakim, mute me
15:36:09 [Zakim]
SteveH should now be muted
15:36:41 [patH]
Enjoy dinner, Yoshio
15:36:55 [SteveH]
can we cut the noise, I cant hear
15:37:22 [jeen]
Zakim, mute me
15:37:22 [Zakim]
Jeen_Broekstra should now be muted
15:37:30 [kendall]
I don't think or claim that giving notice will change the outcome. I just care about the appearance of fairness.
15:37:32 [jeen]
Zakim, unmute me
15:37:32 [Zakim]
Jeen_Broekstra should no longer be muted
15:37:36 [kendall]
zakim, mute me
15:37:36 [Zakim]
Kendall_Clark should now be muted
15:37:36 [patH]
Pat locally muted
15:37:37 [Zakim]
15:37:46 [Zakim]
15:37:46 [AndyS]
zakim, mute me
15:37:48 [Zakim]
AndyS should now be muted
15:37:52 [AndyS]
zakim, unmute me
15:37:52 [Zakim]
AndyS should no longer be muted
15:38:04 [kendall]
It's the Voice of God commanding compliance with design A!
15:38:14 [jeen]
15:38:17 [kendall]
But, as is typical with "the voice of god", it's damn ambiguous! :>
15:38:37 [kendall]
zakim, unmute me
15:38:37 [Zakim]
Kendall_Clark should no longer be muted
15:38:52 [SteveH]
zakim, unmute me
15:38:52 [Zakim]
SteveH should no longer be muted
15:39:33 [DanC]
Zakim, who's on the phone?
15:39:33 [Zakim]
On the phone I see PatH, DaveB, AndyS, Kendall_Clark, DanC, JanneS (muted), EricP, Jeen_Broekstra, SteveH
15:40:41 [DanC]
c a c a a c
15:40:53 [DanC]
janne, got any advice/preference?
15:43:45 [patH]
kendall: basically a web-social question of constraining what publishers do.
15:44:08 [kendall]
a view I take, FWIW, precisely because of webarch.
15:44:57 [DanC]
PROPOSED: to go option (a) without FROM/FROM_NAMED, dataset is unconstrained; with FROM/FROM_NAMED, dataset is bounded from below by given references.
15:45:07 [patH]
Pat sees Kendalls point.
15:46:20 [kendall]
procedurally i guess you only need one objection :>
15:46:21 [DanC]
SH objects. abstaing: EricP, DaveB
15:46:30 [DanC]
15:47:02 [kendall]
sarcasm has to be constrained somehow! :>
15:47:47 [DaveB]
this closes fromUnionQuery?
15:48:11 [kendall]
the devil is definitely the left side. well done, eric.
15:48:19 [DanC]
ACTION AndyS: revise rq23 to suit fromUnionQuery design
15:48:32 [patH]
Ive never been called a deep blue sea before.
15:49:09 [DanC]
Zakim, next agendum
15:49:09 [Zakim]
agendum 5. "punctuationSyntax" taken up [from DanC]
15:49:28 [DanC]
Zakim, take up agendum 6
15:49:28 [Zakim]
agendum 6. "pending Comments, useMentionOp" taken up
15:51:09 [DanC]
ACTION DanC: respond to timbl on behalf of the WG, after looking at what ericP wrote
15:51:20 [DanC]
yes, WG decision on useMentionOp sticks
15:51:37 [DanC]
Zakim, next agendum
15:51:37 [Zakim]
agendum 5. "punctuationSyntax" taken up [from DanC]
15:52:01 [DanC]
Zakim, take up item 7
15:52:01 [Zakim]
agendum 7. "Last Call for SPARQL QL" taken up [from DanC]
15:52:47 [kendall]
i've been responding to protocol comments, FWIW, since the last publication.
15:53:36 [patH]
Handling comments at last call, it all gets formal. Rather like a dance at a 17th century ball.
15:53:42 [AndyS]
AndyS notes:
15:53:50 [AndyS]
LC comment =>
15:54:16 [AndyS]
(1) editor attempts to explain (no bytes changing)
15:54:58 [kendall]
Ah, but the great century for formal balls was the 18th and into the bow-end of the 19th. :>
15:55:21 [AndyS]
15:56:02 [patH]
Yes, Kendall, but in the 17th they were trying hard to be formal but there was still straw on the floor. Its more like that.
15:56:14 [AndyS]
(2) Repeat comment but now a LC comment. Chair can quote decision on email.
15:56:29 [kendall]
*ah*, excellent. It's formal, but the pig is liable to wander through...!
15:56:34 [patH]
15:56:58 [AndyS]
(3) New information ; WG member persuded to need bytes chnage
15:57:14 [AndyS]
(3a) editorial => editor can pass to WG as such
15:57:22 [kendall]
can or must, Andy?
15:57:25 [AndyS]
Chair decides.
15:57:28 [AndyS]
15:57:33 [kendall]
15:57:35 [kendall]
er, right
15:58:06 [AndyS]
(3b) non-editorial: formal process to agree change
15:58:54 [kendall]
do it
15:59:27 [AndyS]
I agree WG should do an internal review
15:59:51 [DanC]
KC, SH volunteer to do a last call review
15:59:57 [DanC]
JB too
16:01:45 [Zakim]
16:02:24 [patH]
Pat also has to leave v. soon.
16:03:04 [ericP]
EricP's yacker grammar:
16:03:59 [AndyS]
Directory: SyntaxFull
16:04:05 [DanC]
PROPOSED: to address punctuationSyntax ala the 70-ish tests in syntax/full
16:04:12 [DaveB]
16:04:13 [DanC]
PROPOSED: to address punctuationSyntax ala the 70-ish tests in SyntaxFull
16:04:23 [ericP]
16:04:31 [ericP]
16:04:31 [ericP]
16:04:51 [DanC]
RESOLVED, PatH and EricP abstaining
16:05:42 [Zakim]
16:05:43 [DanC]
RESOLVED to meet 14 Jun with KendallC to scribe.
16:05:43 [Zakim]
16:05:43 [Zakim]
16:06:00 [Zakim]
16:06:38 [Zakim]
16:06:44 [DanC]
steve, I didn't explicitly action anybody to mark those 70 tests approved. can you do that?
16:06:53 [DanC]
i.e.. would you please?
16:07:08 [SteveH]
DanC, sure
16:07:12 [DanC]
16:07:14 [DaveB]
the bulk approval was as good as we could get. I also didn't like some of them
16:07:35 [DanC]
I'll be somewhat liberal about re-considering individual tests.
16:08:03 [SteveH]
DaveB, why deidnt you abstain or so? I was wondering what the state of the '.'s is, im not really happy with what I understand from the doc, but dont know it well enough to abstain
16:09:49 [DaveB]
I don't think all the state of '.'s is yet tested
16:14:24 [Zakim]
16:14:27 [DaveB]
DaveB has joined #dawg
16:14:28 [Zakim]
16:14:31 [jeen]
jeen has left #dawg
16:15:06 [patH]
16:25:23 [DanC]
RRSAgent, make logs world-access
16:25:40 [DanC]
RRSAgent, bye
16:25:40 [RRSAgent]
I see 3 open action items:
16:25:40 [RRSAgent]
ACTION: EricP to finish extendedType-eq-pass-result.n3 test [1]
16:25:40 [RRSAgent]
recorded in
16:25:40 [RRSAgent]
ACTION: AndyS to revise rq23 to suit fromUnionQuery design [2]
16:25:40 [RRSAgent]
recorded in
16:25:40 [RRSAgent]
ACTION: DanC to respond to timbl on behalf of the WG, after looking at what ericP wrote [3]
16:25:40 [RRSAgent]
recorded in