14:22:24 <RRSAgent> RRSAgent has joined #dawg
14:22:24 <RRSAgent> logging to http://www.w3.org/2007/03/06-dawg-irc
14:22:49 <SimonR> Admittedly, it's only the simplest demonstration of the difference.  The problem becomes visible if you try to translate a result set into a logical expression.
14:24:21 <LeeF> Meeting: RDF DAWG Weekly
14:24:27 <LeeF> Chair: LeeF
14:24:57 <LeeF> Scribe: SimonR
14:25:00 <LeeF> Agenda: http://lists.w3.org/Archives/Public/public-rdf-dawg/2007JanMar/0129.html
14:25:10 <LeeF> Regrets:
14:25:20 <LeeF> agenda + convene, time zone change
14:25:30 <LeeF> agenda + review action items
14:25:51 <LeeF> agenda + unexpected/auto DISTINCT
14:25:58 <LeeF> agenda + Normative and informative, formal and informal parts of rq25
14:26:05 <LeeF> agenda + Test suite
14:26:14 <LeeF> agenda + rq25 status and reviews, last call timing
14:26:53 <AndyS> no zakim
14:27:03 <LeeF> as in...?
14:27:37 <AndyS> I phoned in and nothing heppneded here
14:27:46 <AndyS> zakim, this is dawg
14:27:46 <Zakim> ok, AndyS; that matches SW_DAWG()9:30AM
14:27:54 <AndyS> zakim, who is on the phone?
14:27:54 <Zakim> On the phone I see Andy_Seaborne, [IBMCambridge]
14:28:03 <LeeF> zakim, IBMCambridge is LeeF
14:28:03 <Zakim> +LeeF; got it
14:28:24 <AndyS> I limited to 60 minutes today
14:28:33 <Zakim> +??P4
14:28:42 <SimonR> Zakim, ??P4 is me
14:28:42 <Zakim> +SimonR; got it
14:29:31 <Zakim> +[MIT528]
14:29:41 <LeeF> hi ericP
14:29:44 <ericP> Zakim, MIT528 is me
14:29:44 <Zakim> +ericP; got it
14:29:47 <ericP> heya
14:29:48 <LeeF> MIT528 != Paris :-(
14:29:56 <ericP> not so much, no
14:30:26 <LeeF> SteveH, curious if you happen to be around today?
14:30:40 <SteveH> LeeF, yes hi
14:31:04 <LeeF> Any chance you have the time to dial in for today's call?
14:31:11 <SteveH> LeeF, yes, I think so...
14:31:17 <cgi-irc> cgi-irc has joined #dawg
14:32:29 <LeeF> Well, we're going to get started in a couple of minutes - main topics are the auto-distinct issue and publishing last call, so whenever you can join would be great
14:32:34 <Zakim> +Souri_Das
14:33:00 <Zakim> + +0238055aaaa
14:33:22 <SteveH> Zakim, +0238055aaaa is SteveH
14:33:22 <Zakim> +SteveH; got it
14:33:25 <SteveH> Zakim, mute me
14:33:25 <Zakim> SteveH should now be muted
14:34:44 <LeeF> zakim, who's on the phone?
14:34:44 <Zakim> On the phone I see Andy_Seaborne, LeeF, SimonR, ericP, Souri_Das, SteveH (muted)
14:35:07 <LeeF> Minutes from the 20th Feb: http://lists.w3.org/Archives/Public/public-rdf-dawg/2007JanMar/att-0127/2007-02-20-dawg-minutes.html
14:35:20 <LeeF> SimonR seconds
14:35:21 <LeeF> approved
14:35:36 <LeeF> Minutes from 27th Feb, plus Jeen's regrets:
14:35:36 <LeeF> http://lists.w3.org/Archives/Public/public-rdf-dawg/2007JanMar/att-0123/27-dawg-minutes.html
14:37:10 <LeeF> seconded, ericP
14:37:11 <LeeF> approved
14:38:21 <SimonR> I like UTC.  It stays put.  (And a 0130 local time start will be punishing.)
14:42:07 <cgi-irc> 9:30am US time is good for me
14:43:24 <LeeF> decision: stay at 14:30 UTC
14:43:28 <LeeF> zakim, take up agendum 2
14:43:28 <Zakim> agendum 2. "review action items" taken up [from LeeF]
14:43:45 <LeeF> ACTION: LeeF to talk to SteveH and JeenB about auto distinct behavior in implementations [DONE]
14:43:56 <LeeF> ACTION: AndyS to add text clarifying the prohibition on blank node labels in multiple BGPs to rq25 [CONTINUES]
14:44:03 <LeeF> ACTION: EricP to run the yacker tool over and annotate the existing tests [CONTINUES]
14:44:18 <SimonR> (Discussion with Steve and Jeen at: http://lists.w3.org/Archives/Public/public-rdf-dawg/2007JanMar/0128.html )
14:44:18 <AndyS> AndyS's action is done.
14:44:21 <LeeF> ACTION: EricP to add text to spec noting that ORDER BY comparisons may use  extended implementations of < that operate on types beyond what's given in the operator table [CONTINUES]
14:44:34 <LeeF> ACTION: LeeF to remember that the wee, lost filter tests should be put  [CONTINUES]
14:44:41 <LeeF> action -2
14:44:48 <LeeF> ACTION: AndyS to add text clarifying the prohibition on blank node labels in multiple BGPs to rq25 [DONE]
14:44:56 <LeeF> ACTION: LeeF or EliasT to reply to Bjoern regarding (not) POSTing application/sparql-query documents [CONTINUES]
14:45:29 <Souri> Souri has joined #dawg
14:45:44 <LeeF> zakim, Souri_Das is Souri
14:45:44 <Zakim> +Souri; got it
14:45:59 <LeeF> zakim, close agendum 1
14:45:59 <Zakim> agendum 1, convene, time zone change, closed
14:46:00 <Zakim> I see 5 items remaining on the agenda; the next one is
14:46:01 <Zakim> 2. review action items [from LeeF]
14:46:05 <LeeF> zakim, close agendum 2
14:46:05 <Zakim> agendum 2, review action items, closed
14:46:06 <Zakim> I see 4 items remaining on the agenda; the next one is
14:46:07 <Zakim> 3. unexpected/auto DISTINCT [from LeeF]
14:46:11 <LeeF> zakim, take up next agendum
14:46:11 <Zakim> agendum 3. "unexpected/auto DISTINCT" taken up [from LeeF]
14:47:34 <Zakim> -ericP
14:47:44 <SimonR> q+
14:48:04 <Zakim> +[MIT528]
14:48:13 <ericP> Zakim, MIT258 is me
14:48:13 <Zakim> sorry, ericP, I do not recognize a party named 'MIT258'
14:48:19 <ericP> Zakim, MIT528 is me
14:48:19 <Zakim> +ericP; got it
14:49:02 <SimonR> LeeF: Last week's call tended to favor exact cardinality results.  Dissenters were SimonR (during the call) and SteveH in subsequent communication to LeeF.  Jeen said could work with either.
14:49:12 <LeeF> ack SimonR
14:49:40 <EliasT> EliasT has joined #dawg
14:50:16 <patH> patH has joined #dawg
14:50:41 <SimonR> SimonR: Why do we care about cardinality?  What is its logical meaning?
14:50:55 <EliasT_> EliasT_ has joined #dawg
14:51:25 <Zakim> +PatH
14:51:30 <LeeF> hi, Pat
14:51:36 <EliasT_> EliasT_ has joined #dawg
14:51:37 <SimonR> EricP: FredZ told horrors stories about the history of SQL, in which migration to "non-capricious DISTINCT" happened painfully.
14:51:45 <patH> hi, sorry im late
14:52:35 <SteveH> Zakim, unmute me
14:52:35 <Zakim> SteveH should no longer be muted
14:53:19 <AndyS> There 3 positions: (1) auto DISTINCT (2) defined # duplicates (3) anything in between is OK (loose)
14:53:37 <SimonR> SteveH: Agrees with EricP's account of what FredZ said, but concerned that we're limiting possible indexing schemes.
14:55:22 <SimonR> SteveH: What happens if a triple occurs in two graphs?  AndyS: One for each graph.
14:55:25 <LeeF> <s> <p> ?o 
14:55:25 <LeeF> occuring in multiple graphs - get 1 solution per graph (w/ the same ?o)
14:59:15 <LeeF> (3a) No way to say that auto-distinct is ok
14:59:15 <LeeF> (3b) A "distinctable" keyword to allow 1 < #solutions < n 
14:59:56 <ericP> CONDENSED
15:00:26 <AndyS> I'd prefer to provide defined output (for aggregates later).  If impls choose to be indeterminate - fine but not in spec.
15:01:00 <Souri> MIN, ALL, SOME
15:01:15 <Souri> w.r.t cardinality
15:04:11 <SimonR> It certainly becomes harder to do testing, if we permit multiple correct answers -- but we already have that problem.  (eg w/o ORDER BY, etc)
15:05:31 <LeeF> ericP: In favor of DISTINCT, ALL, and reducible as the default
15:05:41 <LeeF> AndyS: not in favor of that because most people will write the default case
15:05:41 <SimonR> EricP: Now in favor of having an ALL keyword, with the default being reducible.
15:06:56 <patH> Might this be handled by deft use of language? condesable is MUST but some clear option we choose is SHOULD, or somethng like that?
15:08:11 <AndyS> It affects testing and wouldn't we'd need conformance language?
15:09:12 <SteveH> zakim, mute me
15:09:12 <Zakim> SteveH should now be muted
15:09:52 <SteveH> zakim, unmute me
15:09:52 <Zakim> SteveH should no longer be muted
15:14:33 <LeeF> 1) default is strict counting as per the algebra, no keyword to loosen the counting
15:14:33 <LeeF> 2) default is strict counting as per the algebra, add a DISTINCTABLE keyword to loosen the counting
15:14:33 <LeeF> 3) default is loose counting, no keyword to force strict counting
15:14:33 <LeeF> 4) default is loose counting, add an ALL keyword to force strict counting
15:15:07 <LeeF> zakim, who's on the phone?
15:15:07 <Zakim> On the phone I see Andy_Seaborne, LeeF, SimonR, Souri, SteveH, ericP, PatH
15:15:45 <LeeF> 5) always distinct
15:16:24 <LeeF> Andy: ARQ has external flag for "always distinct" - I think those sorts of issues are outside the language
15:17:04 <LeeF> Andy: +1 on #1
15:18:02 <LeeF> Simon: -1 on strict counting; +1 #5, +0.5 #3
15:18:23 <AndyS> Andy's example: SELECT sum(?salary) { ?x :hasSalary ?salary }
15:19:17 <LeeF> Souri: +1 #2 
15:21:42 <SteveH> SteveH: +1 #4, +0.9 #2, -1 #5   #1 is acceptable
15:22:21 <SimonR> Response to Andy's query: SUM(?salary, SELECT ?person ?salary { ?person :salary ?salary })
15:23:04 <AndyS> Simon: I wanted the minutes to record my example.
15:23:15 <SimonR> It's a good example.
15:24:01 <SimonR> You see the point of the response, though?  I choose the properties that serve to distinguish the salary values, in this case the people drawing them.
15:25:04 <AndyS> And my response was that the burden is now on the app writer. This is for the minutes.
15:26:12 <LeeF> LeeF has joined #dawg
15:26:22 <LeeF> LeeF has left #dawg
15:26:29 <LeeF> LeeF has joined #dawg
15:28:11 <SimonR> LeeF: Unlikely to arrive at a decision before Andy needs to leave.  Not quite sure how to proceed.  EricP: Nothing that'll work in the next 10 minutes.  :)
15:29:44 <SimonR> SteveH: What if we required "loose" counting to be either DISTINCT or ALL, but not in between?  EricP: Doesn't really help implementers.
15:30:41 <patH> vote on 1-5: Pat: no very strong opinion, marginally like 2 best.
15:31:22 <AndyS> Hmm - can change the CONSTRUCT answers!
15:31:48 <SimonR> LeeF: Hoping to have QL document ready (minus perhaps just this issue) -- so reviews of rq25 must come in promptly.
15:31:53 <ericP> vote: 1) +1, 2) +1, 3) -1, 4) +1
15:31:54 <AndyS> e.g. CONSTRUCT { [] :salary ?sal } WHERE { ?x :salary ?sal }
15:33:29 <SteveH> AndyS, as [] is a bNode, I don't think it changes anything
15:33:39 <SimonR> AndyS: What are we proposing for next week re: LC   LeeF: Proposion that we make the decision of normative parts of document and DISTINCT issues.
15:34:02 <patH> I have to go for 5 mins, will leave the phone connected.
15:34:08 <AndyS> It does!  A new one gets generated each template :-) So it, not lean, it counts the results!
15:34:37 <SteveH> AndyS, but leaning is perfectly acceptable
15:34:54 <Souri> Zakim, mute me
15:34:54 <Zakim> Souri should now be muted
15:35:03 <SteveH> -> _:a :salary 100000 . _:b :salary 100000 .
15:36:06 <AndyS> Yep - leaning is good here.
15:38:52 <AndyS> SELECT sum(?salary) { ?x a :person . ?x :hasSalary ?salary }
15:42:02 <Zakim> -Andy_Seaborne
15:42:03 <SteveH> G1 { <person-a> :salary 100000 }  G2 { <person-a> :salary 100000 }
15:48:34 <SteveH> G1 { [ :salary 100000 ; :id 12 ] }   G2 { [ :salary 100000 ; :id 12 ] }
15:48:35 <ericP> G1 { [ :employeeId 21; :salary 100000 ] }  G2 { [ :employeeId 21; :salary 100000 ] }
15:48:49 <ericP> by a scant second
15:50:30 <ericP> SELECT ?g ?id ?salary WHERE { GRAPH ?g { :employeeId ?id; :salary ?salary ] }
15:50:34 <ericP> SELECT ?g ?id ?salary WHERE { GRAPH ?g { :employeeId ?id; :salary ?salary } }
15:50:46 <AndyS> AndyS has left #dawg
15:51:02 <SimonR> SUM(?salary, SELECT ?co ?id ?salary WHERE { GRAPH ?co { ?e :employeeID ?id . ?e :salary ?salary }}
15:51:10 <ericP> SELECT ?g ?id ?salary WHERE { GRAPH ?g { ?who :employeeId ?id; :salary ?salary } }
15:51:39 <ericP> SELECT SUM(?salary) WHERE { GRAPH ?g { ?who :employeeId ?id; :salary ?salary } }
15:51:54 <ericP> 200000
15:53:24 <Souri> SELECT ?x, SUM(?sal) { ?x a :Person . ?x :hasSalary ?salary } GROUP BY (?x) HAVING COUNT(*) > 1
15:53:34 <Souri> :-)
15:54:21 <ericP> ooo, HAVING COUNT... cool
15:54:34 <patH> are we all talking the same language?
15:54:38 <ericP> SELECT SUM(?salary) WHERE { GRAPH ?g { ?who :employeeId ?id; :salary ?salary } } => 200000
15:54:59 <ericP> SELECT SUM(?salary) WHERE { GRAPH <G1> { ?who :employeeId ?id; :salary ?salary } } => 100000
15:55:16 <SimonR> patH: We're all talking our different, slightly preferred languages and trying to subvert each other.  :)
15:55:17 <patH> Seems to me this is a lot of hassle to solve a problem that shouldnt even arise in RDF anyway.
15:55:31 <patH> thanks, Simon.
15:55:40 <SteveH> G1 { [ :employeeId 21; :salary 100000; :worksFor :A ] }  G2 { [ :employeeId 21; :salary 100000; :worksFor :B ] }
15:56:44 <ericP> SELECT SUM(?salary) WHERE { GRAPH ?g { ?who :employeeId ?id; :salary ?salary; ?worksFor ?w } } => 200000
15:56:51 <ericP> SELECT SUM(?salary) WHERE { GRAPH ?g { ?who :employeeId ?id; :salary ?salary; ?worksFor :B } } => 100000
15:59:57 <SteveH> G1 { [ :employeeId 21; :salary 100000; :worksFor :A ] }  G2 { [ :employeeId 21; :salary 100000; :worksFor :B ] . [ :employeeId 21; :salary 100000; :worksFor :B ] }
16:01:01 <ericP> pretty-printing that:
16:01:02 <ericP> G1 { [ :employeeId 21; :salary 100000; :worksFor :A ] }
16:01:02 <ericP> G2 { [ :employeeId 21; :salary 100000; :worksFor :B ] . 
16:01:02 <ericP>      [ :employeeId 21; :salary 100000; :worksFor :B ] }
16:04:43 <patH> If G2 is put into RDF using bnodes as subjects (ccase 1) then you will likely have two different bnodes. If you use URIs as subjects (case 2) then the redundancy will not get into the graph.
16:05:52 <LeeF> rrsagent, draft minutes
16:05:52 <RRSAgent> I have made the request to generate http://www.w3.org/2007/03/06-dawg-minutes.html LeeF
16:05:56 <LeeF> rrsagent, make logs world-access
16:06:51 <SimonR> { [ :employeeId 21; :salary 100000; :worksFor :A; :from :G1 ] . [ :employeeId 21; :salary 100000; :worksFor :B; :from :G2 ] . [ :employeeId 21; :salary
16:07:01 <SimonR> +100000; :worksFor :B :from G2 ] }
16:09:09 <patH> case 1:  G1{ _:1 :id :employeeID21 .  _:1 :salary 10000 . _:1 :worksFor :A . }
16:09:37 <LeeF> The fact is that we as a WG have postponed counting/aggregates as an issue ( http://www.w3.org/2001/sw/DataAccess/issues#countAggregate ) . We shouldn't try to solve it without existing implementation experience, and I'm not convinced that we should let it unduly influence our decision on this issue now...
16:09:59 <patH> case 2 : { G1:employeeID21 :salary 10000 .  G1:employeeID21 :worksFor :A .}
16:10:28 <patH> Comforting thought, Lee.
16:15:31 <ericP> +!
16:15:33 <ericP> +1
16:25:18 <Souri> We do not allow bNode sharing between distinct graphs! Things get complex otherwise.
16:25:59 <Souri> URI should be the only way to correlate nodes from distinct graphs.
16:28:25 <Zakim> -Souri
16:29:32 <patH> Have to go. You guys work it out and tell the rest of us what you decide, OK?
16:30:08 <Zakim> -PatH
16:30:58 <Zakim> -SimonR
16:33:41 <LeeF> did I hear the C word?
16:42:10 <ericP> Default Graph: [ :name "Bob" ; :livesIn "Boston" ] .[ :name "Alice" ; :livesIn "Boston" ] .
16:42:48 <ericP> SELECT ?city WHERE { [ :livesIn ?city ] }
16:43:34 <ericP> vs. SELECT ?city WHERE { [ :livesIn ?city ; :name ?name ] }
16:43:41 <Zakim> -SteveH
16:43:41 <ericP> vs. SELECT ?city ?name WHERE { [ :livesIn ?city ; :name ?name ] }
16:43:43 <ericP> .
16:46:27 <ericP>   1. RDF is a ... [distinguish as a graph (not tree) data format]
16:46:27 <ericP>   2. what can an RDF query lang give you?
16:46:27 <ericP>   3. this is SPARQL, and it's cool, so there
16:46:30 <ericP> .
16:48:25 <ericP> The Resource Description Framework (RDF) is a framework for representing information in the Web.
16:49:55 <ericP>   1. RDF is a directed labeled graph language for representing information in the Web.
16:50:32 <ericP> s/language/data format/
16:51:45 <ericP> structure
16:52:17 <ericP> "A query language that uses the structure of XML intelligently can express queries across all these kinds of data, whether physically stored in XML or viewed as XML via middleware."
16:53:39 <ericP> Zakim, who is here?
16:53:39 <Zakim> On the phone I see LeeF, ericP
16:53:40 <Zakim> On IRC I see LeeF, RRSAgent, Zakim, SimonR, SteveH, iv_an_ru, ericP
17:07:24 <SimonR> D'oh.  LeeF, are you still there?  What's the URL of the draft minutes?  O_O
17:07:52 <LeeF> one sec
17:08:00 <LeeF> http://www.w3.org/2007/03/06-dawg-minutes.html
17:08:17 <SimonR> Got it, ta.
17:08:23 <ericP> SimonR, you can mail me replacement text for that document and i'll commit it in place
17:08:30 <LeeF> + updating the situation on the today's distinct issue
17:08:43 <SimonR> EricP, sure.
17:08:53 <LeeF> + addressing Kendall's conformance comments, and other comments from kendall that have been marked with an "eric?"
17:09:11 <LeeF> + taking a look at occurrences of "error" and "warning" in SPARQL QL and either linking to protocol or rewording
17:09:40 <LeeF> + proposing a shortest path to closing the openWorldValueTesting issue
17:10:07 <LeeF> + The Eric-assigned @@s enumerated by Andy
17:10:51 <LeeF> + Settling on ORDER BY text with Andy that gives semantics of multiple order keys, etc.
17:12:50 <SimonR> Incidentally, did you catch anything from that discussion about the algebra of OPTIONAL I was having with Andy just before the meeting started?  It's the main avenue through which I plan to cause trouble in the near future.  :)
17:27:09 <LeeF> http://lists.w3.org/Archives/Public/public-rdf-dawg/2007JanMar/0111.html
17:31:03 <ericP> http://www.w3.org/2001/sw/DataAccess/rq23/rq25#modOrderBy
17:34:01 <Zakim> -LeeF
17:34:02 <Zakim> -ericP
17:34:04 <Zakim> SW_DAWG()9:30AM has ended
17:34:05 <Zakim> Attendees were Andy_Seaborne, LeeF, SimonR, ericP, SteveH, Souri, PatH
19:08:32 <AndyS> AndyS has joined #dawg
19:18:51 <AndyS> EricP?  You free to sync up now (if not too long!).  Or we can set a meet time for later in the week.
19:32:19 <ericP> oops, just started a meeting
19:32:59 <AndyS> So plan B then :-)
19:34:21 <ericP> i was going to add a sentence to the ORDER BY proposed text and drop it in
19:37:10 <ericP> basically, instead of saying that "an {order function} is made of an {expression} and an optional {order modifier}" say that "ORDER BY is followed by a series of {expressions} with optional {order modifiers}. This forms an order function..."
19:37:20 <ericP> but i'm not sure how to finish that sentence
19:37:36 <ericP> generally agreeable to you?
19:53:43 <AndyS> Sorry - my dad on the phone worrying about whether his PC has been compromised.
19:54:16 <AndyS> It hadn't - but he was right to be worried.  Malware attack.
19:56:43 <AndyS> I'm happy for you to replace OrderBy section - looked better text.  My comments were that "order function" was confused - is it the overall function or a comparator.
19:57:29 <AndyS> There is a problem as to whether it is a "function" and if so, what's its domain and range.
19:57:41 <AndyS> zakim, leave us
19:57:41 <Zakim> I don't understand 'leave us', AndyS
19:57:46 <AndyS> zakim, please leave us
19:57:46 <Zakim> I don't understand 'please leave us', AndyS
19:57:55 <AndyS> zakim, leave
19:57:55 <Zakim> Zakim has left #dawg
19:58:10 <AndyS> rrsagent, please leave 
19:58:10 <RRSAgent> I see 6 open action items saved in http://www.w3.org/2007/03/06-dawg-actions.rdf :
19:58:10 <RRSAgent> ACTION: LeeF to talk to SteveH and JeenB about auto distinct behavior in implementations [DONE] [1]
19:58:10 <RRSAgent>   recorded in http://www.w3.org/2007/03/06-dawg-irc#T14-43-45
19:58:10 <RRSAgent> ACTION: EricP to run the yacker tool over and annotate the existing tests [CONTINUES] [3]
19:58:10 <RRSAgent>   recorded in http://www.w3.org/2007/03/06-dawg-irc#T14-44-03
19:58:10 <RRSAgent> ACTION: EricP to add text to spec noting that ORDER BY comparisons may use  extended implementations of < that operate on types beyond what's given in the operator table [CONTINUES] [4]
19:58:10 <RRSAgent>   recorded in http://www.w3.org/2007/03/06-dawg-irc#T14-44-21
19:58:10 <RRSAgent> ACTION: LeeF to remember that the wee, lost filter tests should be put  [CONTINUES] [5]
19:58:10 <RRSAgent>   recorded in http://www.w3.org/2007/03/06-dawg-irc#T14-44-34
19:58:10 <RRSAgent> ACTION: AndyS to add text clarifying the prohibition on blank node labels in multiple BGPs to rq25 [DONE] [6]
19:58:10 <RRSAgent>   recorded in http://www.w3.org/2007/03/06-dawg-irc#T14-44-48
19:58:10 <RRSAgent> ACTION: LeeF or EliasT to reply to Bjoern regarding (not) POSTing application/sparql-query documents [CONTINUES] [7]
19:58:10 <RRSAgent>   recorded in http://www.w3.org/2007/03/06-dawg-irc#T14-44-56