IRC log of dawg on 2005-04-19

Timestamps are in UTC.

14:23:13 [RRSAgent]
RRSAgent has joined #dawg
14:23:13 [RRSAgent]
logging to
14:23:25 [DaveB]
DaveB has joined #dawg
14:23:50 [DanC]
agenda + news from SparqlingDays
14:24:45 [Zakim]
SW_DAWG()10:30AM has now started
14:24:53 [Zakim]
14:26:31 [howardk]
wow, your guys' clock is fast! I show 7:25 am here. concurs. but I'll join you anyway. :-)
14:26:48 [Zakim]
14:26:50 [jeen]
i was a bit early
14:26:54 [AndyS]
zakim, ??P26 is AndyS
14:26:54 [Zakim]
+AndyS; got it
14:28:22 [Zakim]
14:28:29 [DaveB]
Zakim, ??P27 is DaveB
14:28:29 [Zakim]
+DaveB; got it
14:28:37 [Zakim]
14:28:39 [Zakim]
14:28:43 [Zakim]
14:29:25 [DanC]
agenda order is 1, 10
14:29:26 [Zakim]
14:29:46 [DanC]
Zakim, next agendum
14:29:46 [Zakim]
agendum 1. "Convene, take roll, review records and agenda" taken up [from DanC]
14:30:25 [Zakim]
14:30:34 [DanC]
andys asks: why separate ops for comparing integers and decimals?
14:30:56 [DanC]
(not quite convened)
14:30:57 [ericP]
can anybody hear me?
14:31:07 [DaveB]
ericP: I can't
14:31:09 [DanC]
nope, ericp
14:31:20 [Zakim]
14:31:25 [Zakim]
14:31:28 [ericP]
weak, worked two hours ago
14:31:29 [DanC]
regrets: Yoshio, KendallC
14:31:31 [Zakim]
14:31:39 [Zakim]
14:31:54 [KevinW]
KevinW has joined #DAWG
14:32:14 [Zakim]
14:32:32 [AlbertoR]
AlbertoR has joined #dawg
14:33:19 [DanC]
Zakim, take up item 6
14:33:19 [Zakim]
agendum 6. "issue: valueTesting" taken up [from DanC]
14:33:54 [DanC]
Zakim, who is on the phone?
14:33:54 [Zakim]
On the phone I see Jeen_Broekstra, AndyS, DaveB, DanC, HowardK, EricP, ??P31, Kevin, AlbertoR
14:33:56 [DanC]
Zakim, take up item 1
14:33:56 [Zakim]
agendum 1. "Convene, take roll, review records and agenda" taken up [from DanC]
14:34:40 [DanC]
regrets: Yoshio, KendallC, Jos De Roo
14:35:17 [DanC] minutes
14:35:57 [jeen]
minutes approved
14:36:01 [DanC]
next meeting: 26 Apr
14:36:07 [DanC]
regrets 2005-04-26: Jos De Roo
14:36:36 [jeen]
scribe next week ericP
14:37:20 [DanC]
Zakim, who's talking?
14:37:31 [Zakim]
DanC, listening for 10 seconds I heard sound from the following: DanC (41%)
14:37:47 [DanC]
Zakim, next item
14:37:47 [Zakim]
agendum 10. "news from SparqlingDays" taken up [from DanC]
14:37:54 [DanC]
Zakim, who's on the phone?
14:37:54 [Zakim]
On the phone I see Jeen_Broekstra, AndyS, DaveB, DanC, HowardK, EricP, JanneS, Kevin, AlbertoR
14:38:32 [jeen]
AS has student looking into mapping sparql to sql
14:38:35 [DaveB]
14:39:34 [DanC]
DaveB: I don't think limit/offset belong in the protocol
14:40:18 [SteveH]
i agree limit+offset aren't protocol features
14:41:30 [jeen]
EP made an algea client
14:41:38 [DanC]
I saw
14:41:58 [DaveB]
most things can be found off
14:42:21 [ericP]
algae CGI tool linked from
14:42:37 [ericP]
actual script is at
14:42:49 [jeen]
alberto: the idea of the sparqling days was to get together and try to set up some examples and demos
14:43:00 [howardk]
andy: shows the four 'basic' numeric types for function operands; what you're saying about subtype substitution is correct: anything that takes xs:decimal can take xs:integer as well (so I'm not sure why xs:integer is singled out separately, unless it's for efficiency issues.
14:43:11 [ericP]
(some style probs 'cause the proxy breaks stylesheet links)
14:43:19 [Zakim]
14:43:48 [AlbertoR]
14:43:54 [patH]
patH has joined #dawg
14:44:04 [DanC]
Zakim, what item is this?
14:44:04 [Zakim]
agendum 10. "news from SparqlingDays" has been taken up, DanC
14:44:06 [DaveB]
also in progress, I wished for a sparql logo, and lo, one appeared...
14:44:22 [jeen]
alberto: simple website where we want to collect some ideas about sparql, applications and demos, and how to connect, basically a way to showcase 'new web'
14:44:25 [DaveB]
the new web site is
14:44:26 [AlbertoR]
14:44:38 [Zakim]
14:44:43 [AlbertoR]
14:44:47 [SteveH]
Zakim, NickG is SteveH
14:44:47 [Zakim]
+SteveH; got it
14:44:54 [SteveH]
zakim, mute me
14:44:54 [Zakim]
SteveH should now be muted
14:45:27 [AlbertoR]
14:45:30 [DanC]
(lo... )
14:45:52 [AlbertoR]
14:46:19 [AlbertoR]
Zavisa Bjelogrlic
14:46:31 [DaveB]
logo creator is nick "binzac" on
14:46:56 [AlbertoR]
everybody is welcome to join the experiment - just send email to - soon there will be a public mailing list
14:47:23 [DanC]
Zakim, next item
14:47:23 [Zakim]
agendum 2. "issue: fromUnionQuery" taken up [from DanC]
14:48:02 [DanC]
. ACTION AndyS: add syntax to specify datasets in the query language
14:48:31 [jeen]
AndyS the query language does nothing about merging
14:49:04 [DanC]
AndyS: the FROM/FROM NAMED stuff only has one URI for the background graph. no merging. I'd like feedback please
14:49:38 [DanC]
ACTION KendallC: draft protocol section on specifying datasets (fromUnionQuery) including WSDL types and HTTP examples
14:50:12 [AlbertoR]
14:50:20 [jeen]
DaveB it's scary putting wsdl 2.0 semantics in the protocol doc
14:50:28 [DaveB]
at the start
14:50:36 [DanC]
Zakim, who's talking?
14:50:44 [AlbertoR]
HTTP bindings should go first eventually
14:50:46 [Zakim]
DanC, listening for 10 seconds I heard sound from the following: Pat_Hayes (10%)
14:51:03 [Zakim]
14:51:10 [DanC]
nope, it's not you, pat
14:51:38 [DanC]
ACTION EricP: try out the "2 list of URIs" datasets design from
14:51:45 [Zakim]
14:52:31 [jeen]
(why not list of uris for bg graph?)
14:52:57 [jeen]
andys: the next step is people want/expect merging, etc. etc.
14:53:22 [AndyS]
I've implmented 2 lists (the foreground is a merge, the background is one name per graph)
14:53:44 [DanC]
Zakim, next agendum
14:53:44 [Zakim]
agendum 3. "issue: wsdlAbstractProtocol" taken up [from DanC]
14:53:44 [AndyS]
Could have FROM NAMED <u1> <u2> AS <u3>
14:54:21 [ericP]
If we do named graphs, I prefer what Andy just proposed.
14:54:37 [DanC]
ACTION: KendallC to own wsdlAbstractProtocol (based on WSDL description added to protocol to editor's draft).
14:54:40 [DanC]
Zakim, next agendum
14:54:40 [Zakim]
agendum 3 was just opened, DanC
14:54:46 [DanC]
Zakim, close item 3
14:54:46 [Zakim]
agendum 3 closed
14:54:47 [Zakim]
I see 6 items remaining on the agenda; the next one is
14:54:48 [Zakim]
4. issue: xmlAbstractSyntax [from DanC]
14:54:50 [DanC]
Zakim, next agendum
14:54:50 [Zakim]
agendum 4. "issue: xmlAbstractSyntax" taken up [from DanC]
14:55:03 [DanC]
ACTION: KendallC to work out remaining details in sparqlx.{rnc,xsd,rng}: expression syntax, order-by stuff (xmlAbstractSyntax)
14:56:26 [DanC]
DanC: what's the defined term corresponding to SPARQL query?
14:56:43 [jeen]
andys: it's called Query
14:57:05 [jeen]
dc: that's concrete, want defs
14:57:30 [jeen]
andys: there isn't a term for that
14:59:21 [DaveB]
14:59:23 [DanC] 1.6
15:00:08 [DaveB]
the recursive term is query-pattern
15:01:04 [DanC]
ACTION AndyS: define a term for abstract form of SPARQL query
15:01:05 [DaveB]
although in the Query WD, I'd have expected it to be graph-pattern
15:01:13 [DanC]
Zakim, next agendum
15:01:13 [Zakim]
agendum 5. "protocol publication candidate" taken up [from DanC]
15:01:14 [AndyS]
In rq23 the recursive bit is "graph pattern"
15:01:42 [DanC]
ACTION: KendallC to incorporate EricP's privacy text, salting to taste
15:01:56 [jeen]
AndyS: will read it, can't commit to review
15:01:57 [DanC]
DanC solicits reviewers for upcoming pub
15:02:04 [DanC]
Zakim, next agendum
15:02:04 [Zakim]
agendum 5 was just opened, DanC
15:02:10 [DanC]
Zakim, close this item
15:02:10 [Zakim]
agendum 5 closed
15:02:12 [Zakim]
I see 4 items remaining on the agenda; the next one is
15:02:12 [DanC]
Zakim, next agendum
15:02:14 [Zakim]
6. issue: valueTesting [from DanC]
15:02:15 [Zakim]
agendum 6. "issue: valueTesting" taken up [from DanC]
15:02:26 [DanC]
ACTION: DaveB to work with EricP to clarify valueTesting proposal
15:02:39 [DanC]
. ACTION: HowardK to seek clarification on valueTesting w.r.t. differenced from XQuery
15:02:57 [jeen]
HK made progress. not sure when to claim victory, wants feedback from andy and ericP
15:03:21 [DaveB]
discussing hk's doc
15:05:23 [DanC]
ack danc
15:05:23 [Zakim]
DanC, you wanted to encourage xref checking ala
15:06:13 [jeen]
DanC - check links from our spec to xquery spec re types and comps
15:07:03 [jeen]
DanC how close to done?
15:08:19 [jeen]
ericP some fuzzy issues left:
15:10:34 [jeen]
- whether plain rdf literals because they are the same term or because they have the same value
15:10:34 [DaveB]
this seems to be the rdf term equality operator -
15:13:33 [jeen]
DanC when do you go to second bit in OR?
15:13:47 [DaveB]
I'd need a test for this potential OR, can't see ericP's problem.
15:13:50 [SteveH]
in rdf-term-equal one could be unbound
15:14:25 [jeen]
andyS: idea: table in document with example comparisons
15:15:19 [DaveB]
DanC was asking "1.0"^^xsd:integer compared to "1"^^xsd:integer
15:15:25 [DaveB]
different terms, would compare equal
15:15:34 [ericP]
bindings same type-specific
15:15:34 [ericP]
?x ?y node equivilence
15:15:34 [ericP]
<foo> <foo> 1 error
15:15:34 [ericP]
_:a _:a 1 error
15:15:34 [ericP]
"123" "123" 1 error
15:15:36 [ericP]
"123"^^int "123"^^int 1 1 =(int, int)
15:15:39 [ericP]
"123"^^int "123"^^float 0 1 =(float, float)
15:15:41 [ericP]
"11:01"^^dt "11:01"^^dt 1 1 =(dateTime, dateTime)
15:16:58 [jeen]
DanC asks Janne if this is a problem in his impl
15:17:27 [jeen]
Janne - haven't considered it in detail yet
15:17:59 [AndyS]
Don't understand :: "123" "123" 1 error
15:18:32 [SteveH]
AndyS, "123" doesnt have a type, so its realsonable that a type specific test fails
15:18:46 [ericP]
agae impl:
15:19:15 [ericP]
look for 2nd instance of "sub numericOrDTCompare {
15:19:16 [ericP]
15:19:33 [jeen]
DanC cwm background graph is always normalized
15:19:51 [jeen]
Dave is not coding this part of the design
15:19:56 [DaveB]
15:19:58 [SteveH]
I've done some of it, but not all
15:20:02 [DanC]
Zakim, next agendum
15:20:02 [Zakim]
agendum 7. "issue: sort" taken up [from DanC]
15:20:11 [ericP]
all atoms in algae impl aer normalized
15:20:19 [DanC]
. ACTION JeenB: Write a few sort test cases.
15:20:29 [AndyS]
I think that we extend the type space: XSD types + plain literal.
15:20:48 [DanC]
15:21:36 [patH]
it would make sense to treat plain literals as though they werre typed with xsd:string.
15:21:59 [DanC]
<rs:solution rdf:parseType="Resource">
15:22:00 [DanC]
<rs:index rdf:datype="">1</rs:index>
15:22:19 [patH]
this equality is specified in the RDF semantics as normative.
15:22:23 [DaveB]
the rdf/xml is invalid
15:22:56 [DanC]
15:23:43 [AndyS]
PatH: yes
15:24:24 [DaveB]
rdf plain literal is an xsd:string literal + language tag. need to consider both in equality
15:25:40 [patH]
I was presuming no lang tag. Lang tagged literalks are different from tagged, must have both tag and literal string-equal.
15:26:38 [AndyS]
We going to enforce non-case etsts
15:26:49 [AndyS]
We going to enforce non-case tests : lang tag: "en" = "EN" ?
15:27:37 [patH]
Hmm, never considered that. Suggest we consult external authorites :-)
15:28:57 [SteveH]
zakim, unmute me
15:28:57 [Zakim]
SteveH should no longer be muted
15:29:17 [SteveH]
zakim, mute me
15:29:17 [Zakim]
SteveH should now be muted
15:29:45 [SteveH]
zakim, unmute me
15:29:45 [Zakim]
SteveH should no longer be muted
15:30:02 [SteveH]
zakim, mute me
15:30:02 [Zakim]
SteveH should now be muted
15:31:27 [jeen]
DanC close sort issue?
15:32:18 [DanC]
PROPOSED: that v 1.307 addresses the sort issue
15:33:22 [jeen]
DaveB not happy with wording, many issues came up with LIMIT, OFFSET etc
15:33:27 [ericP]
i have not addressed this
15:34:33 [DanC]
(darn. wind noise)
15:34:36 [DaveB]
too much noise
15:34:59 [DanC]
15:35:06 [DaveB]
if 10.1 said "Query patterns generate a sequence of solutions, and each solutoon...
15:35:09 [DaveB]
that'sd be better
15:35:25 [Zakim]
15:35:26 [DaveB]
"passed through a stage to control" sounds like implementation
15:35:58 [DaveB]
and the bulleted list should be an order; controls are processed in the following order 1. DISTINCT ...
15:36:09 [Zakim]
15:36:18 [ericP]
15:36:38 [Zakim]
15:36:38 [AndyS]
I can't find Solution Sets in the doc.
15:36:47 [DaveB]
ignore SOlution Sets, I said Soluton sequences
15:37:28 [ericP]
i muted JanneS and it went away immediately
15:37:31 [Zakim]
15:37:53 [AndyS]
"Solution Sets vs Solution Sequences" in 0038
15:38:03 [DaveB]
well I did say I was confused
15:38:13 [DaveB]
clearly solutions can't be sets; as you can distinct them
15:38:14 [SteveH]
zakim, unmute me
15:38:14 [Zakim]
SteveH should no longer be muted
15:38:34 [jeen]
Steve: thought about it, not implemented it
15:38:45 [jeen]
Steve: feature I really want but looks....
15:39:14 [jeen]
DaveB: only feedback I got was 'please do DISTINCT and LIMIT'
15:39:43 [DanC]
Zakim, who's on the phone?
15:39:43 [Zakim]
On the phone I see Jeen_Broekstra, AndyS, DaveB, DanC, HowardK, EricP, JanneS (muted), Kevin, AlbertoR, SteveH, Pat_Hayes
15:40:23 [AlbertoR]
some ORDER_BY is important and needed today -
15:41:31 [DanC]
poll: who thinks postponing sort is a bad idea? 3ish
15:41:50 [DaveB]
I'm neutral
15:42:20 [SteveH]
I'm also nuetral
15:43:00 [jeen]
AR: i think it's very important to have some kind of ordering
15:43:50 [jeen]
AR: would rather spend more time implementing what the spec says than not having any kind of sort
15:44:11 [DanC]
ACTION JeenB: Write a few sort test cases.
15:44:15 [DanC]
Zakim, next agendum
15:44:15 [Zakim]
agendum 8. "SPARQL QL publication candidate" taken up [from DanC]
15:44:41 [jeen]
DanC this afternoon likely
15:44:52 [DanC]
ACTION: PatH to internal review of rq23
15:45:16 [DanC]
Zakim, next agendum
15:45:16 [Zakim]
agendum 9. "test suite maintenance (time permitting)" taken up [from DanC]
15:45:36 [jeen]
andys - still have comments from bijan that i haven't processed yet
15:45:55 [DanC]
. ACTION: EricP to pair with SteveH on making the HTML test results page
15:46:08 [DanC]
ACTION: EricP to pair with SteveH on making the HTML test results page
15:46:31 [DanC]
. ACTION: SteveH to to revise test manifest w.r.t. "background" and named graphs
15:46:34 [DanC]
15:46:40 [DanC]
ACTION: DaveB to to propose source test to approve
15:46:57 [DanC]
ACTION: SteveH to prepare test cases for publication as WG Note (no deadline/urgency)
15:47:25 [Zakim]
15:47:25 [DanC]
. ACTION: AndyS to add the above graph test cases (analagous to valueTesting test cases) (don't expect quick delivery)
15:47:30 [DanC]
ACTION: AndyS to add the above graph test cases (analagous to valueTesting test cases) (don't expect quick delivery)
15:47:37 [DanC]
ACTION: DanC to follow up re optional test based on op:dateTime triple
15:47:50 [DanC]
. ACTION: AndyS to propose a test re "x"^^mytype != "y"^^mytpe
15:49:10 [DanC]
ACTION: EricP to propose a test re "x"^^mytype != "y"^^mytpe
15:49:49 [DanC]
15:49:49 [Zakim]
15:49:53 [Zakim]
15:49:54 [Zakim]
15:50:02 [DanC]
RRSAgent, make logs world-access
15:50:07 [howardk]
eric: wanna chat about getting together ??
15:50:08 [Zakim]
15:50:10 [DanC]
RRSAgent, pointer?
15:50:10 [RRSAgent]
15:50:22 [ericP]
howard, sure
15:50:44 [Zakim]
15:50:46 [Zakim]
15:51:23 [jeen]
jeen has left #dawg
15:51:38 [DanC]
regrets + Thompson, Bryan B
15:52:37 [Zakim]
15:53:05 [AndyS]
AndyS has left #dawg
15:58:51 [Zakim]
16:07:16 [ericP]
netstorm -- will ask you to repeat a lot
16:08:46 [ericP]
fallback to irc
16:09:05 [howardk]
16:09:15 [howardk]
i hear you fine
16:09:44 [howardk]
2 hours ??
16:09:49 [howardk]
by phone?
16:09:55 [howardk]
sounds good
16:10:09 [ericP]
still broken
16:10:09 [howardk]
broken ... too bad
16:10:19 [howardk]
sounds good eric.
16:10:43 [howardk]
16:10:43 [howardk]
ok, 1 hour ...
16:10:43 [howardk]
good ...
16:10:43 [howardk]
16:10:43 [howardk]
16:10:43 [howardk]
16:10:43 [Zakim]
16:10:52 [Zakim]
16:10:53 [ericP]
ouch! my ear bud hurts
16:10:54 [Zakim]
SW_DAWG()10:30AM has ended
16:10:55 [Zakim]
Attendees were Jeen_Broekstra, AndyS, DaveB, DanC, HowardK, EricP, Kevin, AlbertoR, JanneS, Pat_Hayes, SteveH
16:11:16 [DanC]
RRSAgent, stop