IRC log of dawg on 2004-11-16

Timestamps are in UTC.

14:28:43 [RRSAgent]
RRSAgent has joined #dawg
14:28:44 [afs]
zakim ??P0 is AndyS
14:28:49 [ericP]
zakim, please dial ericP-617
14:28:49 [Zakim]
ok, ericP; the call is being made
14:28:50 [Zakim]
14:28:54 [afs]
zakim, ??P0 is AndyS
14:28:54 [Zakim]
+AndyS; got it
14:29:08 [AlbertoR]
AlbertoR has joined #dawg
14:29:48 [Zakim]
14:30:07 [SimonR]
Zakim, ??P1 is SimonR
14:30:07 [Zakim]
+SimonR; got it
14:30:24 [Zakim]
14:30:37 [SimonR]
Zakim, ??P2 is DaveB
14:30:37 [Zakim]
+DaveB; got it
14:31:08 [SimonR]
We have regrets from Danc and Kendall so far.
14:31:14 [Zakim]
14:31:28 [SimonR]
Zakim, ??P3 is JanneS
14:31:28 [Zakim]
+JanneS; got it
14:32:00 [DanC-AIM]
DanC-AIM has joined #dawg
14:32:02 [Zakim]
14:32:58 [SimonR]
Evening, Dan. You're pretty intrepid, coming in on the Hiptop I presume!
14:33:22 [TomAdams]
TomAdams has joined #dawg
14:33:29 [DanC-AIM]
14:33:50 [DanC-AIM]
Finally got around to getting a sidekick II this week
14:34:44 [ericP]
zakim, who is here?
14:34:44 [Zakim]
On the phone I see AndyS, EricP, SimonR, DaveB, JanneS, AlbertoR
14:36:13 [Zakim]
14:36:14 [ericP]
danc, do you have a list of who is in good standing?
14:36:17 [SimonR]
Checking whether we have enough attendees to form a quorum...
14:36:26 [ericP]
i'm tring to see if we have a quorum
14:36:28 [patH]
patH has joined #dawg
14:36:34 [DanC-AIM]
There are no quorum rules
14:36:50 [TomAdams]
I can call in if we need a quorum
14:36:56 [DanC-AIM]
You have a quorum if yo say you do and nobody objects, ericp
14:37:27 [ericP]
tom, are you not planning to call in?
14:37:33 [Zakim]
14:37:50 [TomAdams]
I am, but I'm talking to my boss at the moment, I can push him off if need be.
14:38:00 [ericP]
we'll carry on
14:38:02 [ericP]
zakim, take up agendum 1
14:38:02 [Zakim]
agendum 1. "Convene, take roll, review agenda" taken up [from ericP]
14:38:08 [TomAdams]
OK. I will call in shortly.
14:38:08 [swh]
Zakim, Holbrook|Gibbins is SteveH
14:38:08 [Zakim]
sorry, swh, I do not recognize a party named 'Holbrook|Gibbins'
14:38:42 [SteveH]
Zakim, +Holbrook|Gibbins is SteveH
14:38:42 [Zakim]
sorry, SteveH, I do not recognize a party named '+Holbrook|Gibbins'
14:38:42 [ericP]
PROPOSE: to accept as a true record:
14:38:42 [ericP]
14:38:42 [ericP]
as amended by
14:38:42 [ericP]
14:38:42 [ericP]
14:39:14 [SimonR]
RESOLVED to accept minutes -- seconded, no oppositions.
14:39:25 [ericP]
propose to continue these without discussion:
14:39:25 [ericP]
ACTION DanC: owner of issue 'yes or no questions' (pending protocol doc)
14:39:25 [ericP]
ACTION AlbertoR: owns issue 'DESCRIBE'
14:39:25 [ericP]
ACTION SteveH: is willing to adapt his testing infrastructure to
14:39:25 [ericP]
generate input/output RDF/XML and typed nodes into manifest file
14:39:27 [ericP]
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.
14:39:30 [ericP]
ACTION DanC: notify SemWeb CG of EbXML possible sync point in April
14:40:02 [ericP]
14:40:10 [JosD]
JosD has joined #dawg
14:40:18 [ericP]
zakim, take up next agendum
14:40:18 [Zakim]
agendum 2. "FTF 4" taken up [from ericP]
14:40:20 [SimonR]
All miscellaneous ACTIONs continue.
14:40:31 [ericP]
ACTION DanC: Make a dawg f2f4 registration form
14:40:31 [ericP]
14:40:55 [SimonR]
We have 3 or 4 people who have registered. Everyone should register!
14:40:56 [Zakim]
14:41:13 [SimonR]
Janne tempts us with the first snow of the season in delightful Finland!
14:41:56 [ericP]
zakim, take up next agendum
14:41:56 [Zakim]
agendum 3. "Protocol" taken up [from ericP]
14:41:58 [SimonR]
Eric asks Janne whether there are any logistics issues? Janne suggests that this would be better to defer until next week, when more attendence results are in.
14:42:31 [SimonR]
Eric suggests that without Kendall we'll defer discussing protocol.
14:42:35 [ericP]
ACTION JanneS: read the protocol draft [protocol] and email a review to the WG before the next telcon 2004-11-09
14:42:38 [ericP]
14:43:02 [SimonR]
When Tom joins we'll ask whether he's read the protocol draft.
14:43:07 [ericP]
zakim, take up next agendum
14:43:07 [Zakim]
agendum 4. "SPARQL SOURCE Issue" taken up [from ericP]
14:43:21 [ericP]
ACTION JanneS: review and check what happens when no source support is implemented
14:43:39 [ericP]
14:43:42 [SimonR]
Janne's action continues.
14:44:32 [SimonR]
Janne's preliminary impression is that it may be safe to just omit the binding if SOURCE is not available.
14:44:55 [Zakim]
14:45:01 [ericP]
welcome tom
14:45:05 [TomAdams]
zakim, Tucana is TomAdams
14:45:05 [Zakim]
+TomAdams; got it
14:45:11 [SimonR]
SteveH suggests that either an error should be returned or that we need to permit two sets of test results, one for systems that support SOURCE and another for those that don't.
14:45:19 [SteveH]
SimonR, it was DaveB
14:45:36 [SimonR]
Oops, sorry. Noted!
14:45:36 [DaveB]
those were design alternatives, I chose not to give a hard failure
14:46:32 [afs]
I worry it makes variables in SOURCE ?var different from other vars and problems (?) if ?var used elsewhere
14:46:57 [DaveB]
PatH refers to TAG recommending hard failure for features that aren't supported
14:48:16 [SimonR]
Eric notes that XML is an example of a spec which mandates hard failures and it seems to have helped XML.
14:49:38 [ericP]
ACTION DaveB: update the dawg test repository to record or amend tests to correspond to the proposal
14:50:01 [ericP]
14:50:37 [AlbertoR]
wonders if SOURCE tests I provided might fit DaveB model ?
14:51:01 [AlbertoR]
actually AndyS & DaveB model/proposal
14:51:50 [ericP]
here we get any the scary area of negotiable compliance ala SGML
14:52:42 [ericP]
JosD: we could use Unsupported Option for queries where a source is specified but not supported
14:53:10 [SimonR]
AndyS's main concern is that variables in SOURCE behavior the same way as variables in other positions.
14:53:19 [JosD]
14:53:58 [AlbertoR]
yes basically, system graph might just be qt:data - but in the bNodes graph-name case does not work
14:54:41 [SimonR]
AndyS asks what NAMED is doing? JanneS +1
14:54:53 [ericP]
that was JosD
14:55:05 [SimonR]
Gah. I can't pick anyone by voice. :)
14:55:09 [AlbertoR]
nice DaveB :)
14:56:10 [AlbertoR]
+1 DaveB
14:56:36 [AlbertoR]
one store/database might contain several triples from several different SOURCEs
14:56:46 [howardk]
howardk has joined #dawg
14:56:59 [howardk]
hi all. sorry I'm late.
14:57:09 [ericP]
14:57:28 [howardk]
be along phonewise in a second ...
14:57:32 [SimonR]
Hi Howard. We're at agenda item 4 (SOURCE)
14:57:38 [howardk]
thanks simon
14:57:42 [afs]
Isn't it early, not late? :-)
14:57:53 [AlbertoR]
PatH: I would hash some URI for the SOURCE - to blind "sign" the triples
14:58:31 [Zakim]
14:58:46 [ericP]
zakim, who's here
14:58:46 [Zakim]
ericP, you need to end that query with '?'
14:58:50 [ericP]
zakim, who's here?
14:58:50 [Zakim]
On the phone I see AndyS, EricP, SimonR, DaveB, JanneS, AlbertoR, PatH, HolbrookGibbins, JosD, TomAdams, HowardK
14:59:14 [ericP]
11 people, good.
14:59:40 [AlbertoR]
I was thinking of FROM <somesource.rdf> AS ?src - which implies some dawg:source or dc:source between some resource and <somesource.rdf>
15:01:56 [SteveH]
AlbertoR, thats sugar
15:02:17 [ericP]
SELECT * FROM uri:group:government-documents WHERE ...
15:02:29 [AlbertoR]
sure - sugar - but as tried to motivate we need some sugar to bind a variable to a SOURCE i.e. log:semantics like
15:03:10 [AlbertoR]
EricP but uri:group:government-documents might contain several triples from several different SOURCE - you might want to filter uri:group:government-documents by SOURCE
15:03:34 [ericP]
right, i considered exploring that in the WHERE
15:04:18 [AlbertoR]
ok dc:source - but then you need to distinguish between qt:data and qt:metadata (triples about the qt:data) - which can be harder to "sell"
15:04:43 [ericP]
ala life sciences ids
15:04:45 [SteveH]
AlbertoR, I dont think theres a problem
15:06:19 [AlbertoR] was my attmept to motivate an "idirect reference (name) to a graph"
15:07:06 [SimonR]
ACTION TomAdams: forward internal Tucana email about permanent model naming
15:07:15 [afs]
Dave - will you be able to answer the questions in email (0256, 0260) - if so, I will wait for that.
15:07:41 [DaveB]
afs: big project deadline, not before next Wed
15:08:28 [DanC-AIM]
DanC-AIM has joined #dawg
15:08:34 [afs]
Next wed = 24th?
15:08:44 [SimonR]
JosD: N3 has the log:semantics property that relates a document and its graph
15:08:46 [AlbertoR]
JosD I like that idea i.e. have a way to bind a ?src to a URI
15:10:08 [ericP]
ACTION DanC: suggest formal definitions for SOURCE
15:10:13 [ericP]
15:10:13 [SimonR]
AndyS will attempt to incorporate DaveB's work into the SPARQL query document, while Dave is busy.
15:10:48 [ericP]
zakim, take up next agendum
15:10:48 [Zakim]
agendum 5. "SPARQL Disjunction issue" taken up [from ericP]
15:10:59 [ericP]
ACTION SimonR: explain how much of disjuction can be done with optionals, nested or otherwise. Point to references, problems with LEFT OUTER JOIN in the literature.
15:13:00 [DanC-AIM]
RRSAgent, make logs world-access
15:19:36 [ericP]
15:20:14 [SimonR]
SteveH: his system does things in OPTIONAL clauses involving back-chained inference which isn't included in usual outer join.
15:21:23 [AlbertoR]
15:21:32 [ericP]
ack AlbertoR
15:24:23 [AlbertoR]
I I underatand it right dc10:title OR dc11:title would become WHERE....( ?s ?p ?title) AND ( ?p == dc10:title or ?p == dc11:title ) - simple enough
15:24:34 [SteveH]
AlbertoR, yes, thats right
15:24:51 [AlbertoR]
what about the case no matches? I guess OPTIONAL in front of the triple-pattern - I like that - nice and simple
15:25:40 [ericP]
ACTION SteveH: own (i.e. propose resolution to) disjunction issue
15:25:45 [afs]
Or ask two queries (i.e.rewrite to canoical OR/AND form)
15:25:54 [ericP]
ACTION SteveH: owns issue 'nested optionals'
15:26:08 [ericP]
zakim, take up next agendum
15:26:08 [Zakim]
agendum 6. "SPARQL update-related issues" taken up [from ericP]
15:26:16 [AlbertoR]
afs: clear - that's always possible I expect
15:26:29 [SimonR]
EricP: Aiming for a release of SPARQL in December.
15:26:31 [ericP]
ACTION EricP: supply definitions for SELECT (vars ordered or not?) ala
15:26:34 [afs]
And optimal in some cases!
15:27:58 [ericP]
15:28:04 [SimonR]
The question is whether the order of the variables in the SELECT clause should specify the order of the variables in the result set.
15:28:33 [DanC-AIM]
Did kendall's protocol doc depend on var order?
15:28:51 [TomAdams]
In don't recall reading that Dan
15:29:01 [ericP]
WITHDRAW -- and look for "Definition: Projection" in
15:32:25 [AlbertoR]
what about RDF/XML table result-set format then? do we need collections? Please no....
15:32:57 [ericP]
we have passed the 1 hour marker. i will soon interupt this interesting discussion to ask if folks want to stay on.
15:34:07 [SteveH]
afs, if your api doesnt allow you to access the nth column then its irrelevent isnt it?
15:34:51 [afs]
No - if the spec says "in order" I feel I have to offer col iterator
15:35:30 [AlbertoR]
I remark that Perl DBi for example has both - fetchrow_arrayref() fetchrow_hashref()
15:35:43 [SteveH]
so does php
15:36:53 [AlbertoR]
15:37:08 [ericP]
zakim, take um agendum 8
15:37:08 [Zakim]
I don't understand 'take um agendum 8', ericP
15:37:16 [ericP]
zakim, take up agendum 8
15:37:16 [Zakim]
I only see 7 items on the agenda
15:37:19 [ericP]
zakim, take up agendum 7
15:37:19 [Zakim]
agendum 7. "Test Status" taken up [from ericP]
15:38:14 [SimonR]
Scribe volunteer for next week? [crickets...]
15:38:17 [DaveB]
DaveB: regrets for next 2 telcons (23, 30th Nov)
15:38:40 [AlbertoR]
me most probably traveling
15:38:41 [howardk]
I'm away for the next two weeks as well.
15:38:42 [SimonR]
TomA: regrets for next week -- off the hook for scribing.
15:39:21 [SimonR]
AndyS volunteers to be the next scribe.
15:39:35 [ericP]
next meeting; 1 week, scribe: AndyS
15:40:16 [SimonR]
15:40:16 [Zakim]
15:40:18 [Zakim]
15:40:19 [Zakim]
15:40:20 [Zakim]
15:40:21 [Zakim]
15:40:22 [Zakim]
15:40:25 [Zakim]
15:40:49 [TomAdams]
Andy, I'm back from my training so will be able to start looking at the comments list now. Thanks for handling those last ones.
15:40:53 [ericP]
rrsagent, pointer?
15:40:53 [RRSAgent]
15:46:43 [afs]
zakim, who is on the phone?
15:46:43 [Zakim]
On the phone I see AndyS, EricP, SimonR, HolbrookGibbins
15:48:22 [Zakim]
15:48:38 [ericP]
that was steve
15:50:32 [Zakim]
15:53:56 [DaveB]
DaveB has joined #dawg
16:01:09 [afs]
zakim, who is on the phone?
16:01:09 [Zakim]
On the phone I see AndyS, EricP
16:12:31 [Zakim]
16:12:32 [Zakim]
16:12:33 [Zakim]
SW_DAWG()9:30AM has ended
16:12:34 [Zakim]
Attendees were EricP, AndyS, SimonR, DaveB, JanneS, AlbertoR, PatH, Holbrook|Gibbins, JosD, [Tucana], TomAdams, HowardK
17:37:21 [AndyS]
zakim, please leave us
17:37:21 [Zakim]
I don't understand 'please leave us', AndyS
17:38:18 [AndyS]
zakim, excuse us
17:38:18 [Zakim]
Zakim has left #dawg
18:49:52 [DaveB]
DaveB has joined #dawg