See also: IRC log
<ericP> who's scribe?
<DaveB> kendall
<ericP> win go 20
<ericP> agenday request: grammar
ericp: is there "another" grammar from which the grammar in the spec is generated?
i was looking for some text to cut and paste w/out all the production numbers and such :>
<DaveB> I saw "Regrets from JanneS, Yoshio, patH tentative regrets: DanC, SteveH"
<ericP> kendall, http://www.w3.org/2005/01/yacker/uploads/sparqlTest/bnf?lang=perl
ah, excellent. thx eric.
<ericP> kendall, but it's not generatd from that
(we seem to have a quorom :>)
<AndyS> KC: It's produced by a perl (!) script from javacc. Can take and tailor the perl script
<AndyS> Could produce text if you need
hmm, okay. no biggie. i'm using antlr to genereate a python parser (:>)
<DaveB> minutes was http://www.w3.org/2005/06/07-dawg-irc
+HiroyukiS
<DaveB> for last week
erp: >
2 requests for "more human readable" minutes from last week's meeting
(er, at least 2 requests)
so, formally, minutes from last week's meeting not accepted today
<scribe> ACTION: DanC To produce a more humanly readable version of last week's meeting minutes [recorded in http://www.w3.org/2005/06/14-dawg-minutes.html#action01]
CONTINUE other actions from agenda
<AndyS> Can fly Newark <-> Bristol nowadays
Nice: >
Jeen on holiday from 4 August to 8 August
EricP on holiday during August
<AndyS> AndyS:Not week of August 15
KendallC also on holiday in Houston for some 7 week period during August
<ericP> my vacation is roughly the same dates as Jeen
(Append attendees: +Jos)
AndyS notes that cancelling a scheduled f2f is easier than slipping one in at the last minute)
<AndyS> VLDB: Aug 30 - Sept 2
Souri also attending
(sorry, I missed you too)
er, he's not on IRC :>
+Souri
<DaveB> re http://lists.w3.org/Archives/Public/public-rdf-dawg/2005AprJun/0338.html
<DaveB> and http://lists.w3.org/Archives/Public/public-rdf-dawg/2005AprJun/0341.html
Request to put brackets around FILTER (is that right?)
re: some ambiguities in the grammar
this change affects tests we've already approved
er, this change *would* affect them
DaveB in favor of the change
<AndyS> Example: "{ FILTER q:name() :a :b }"
+q to ask about "compliance" (sorry!)
DaveB would rather have "brackets everywhere", doesn't like the special cases
<ericP> http://www.w3.org/2005/01/yacker/uploads/sparqlTest?lang=perl&markup=html#prod-sparqlTest-Query
AndyS thinks regex will be commonly used
<ericP> the above link is how i solved it
<ericP> (in the sparql grammar)
<DaveB> discussion of regex in python, seems it has a perl5 compat regex library
(eek, sorry, badly scribing AND misleading the discussion!)
Jeen: doesn't like the special cases either
All the builtin functions are allowed w/out the outer brackets
Proposal: All the builtin functions are allowed w/out the outer brackets
<DaveB> with FILTER
Proposal: All the builtin functions are allowed w/out the outer brackets with FILTER
Where all the builtin functions is...
<AndyS> 'STR' '(' Expression ')'
<AndyS> | 'LANG' '(' Expression ')'
<AndyS> | 'DATATYPE' '(' Expression ')'
<AndyS> | RegexExpression
<AndyS> | 'BOUND' '(' Var ')'
<ericP> http://www.w3.org/2005/01/yacker/uploads/sparqlTest?lang=perl&markup=html#prod-sparqlTest-Query
<AndyS> | 'isURI' '(' Expression ')'
<AndyS> | 'isBLANK' '(' Expression ')'
<AndyS> | 'isLITERAL' '(' Expression ')'
(thanks andy!)
<AndyS> "{ FILTER q:name() :a :b }"
<ericP> { FILTER (q:name()) :a :b }
AndyS: can also use an explicit syntax for function calls
<AndyS> &q:name()
I hate it. :>
<DaveB> it makes q:name and () - empty list, different from &q:name() function call
<AndyS> Separate function call and q:name as constant value
Proposal2: To adopt an explict syntax for function calls (which makes q:name and () empty list separate from &q:name() function call)
Jeen doesn't like it
Straw poll: proposal 1, proposal 2, neither
(we rely too much on people being on irc. :>)
JosD: don't really care
Souri: don't really care
Kendall: proposal1+, proposal2-
Souri: proposal1+, proposal2-neutral
Jeen: prop1+, prop2-
Hiroyuki: prop2-mild pref
EricP: mild for p2
Howard: prop2-
AndyS: p1+, p2-
DaveB: p2 mild pref, but doesn't care much
Editors are encouraged to take the straw poll as advice from the WG
<howardk> i'll talk louder jeen!
<jeen> :)
yacker grammar, even :>
<ericP> http://www.w3.org/2005/01/yacker/uploads/sparqlTest?lang=perl&markup=html#prod-sparqlTest-Query
(/me not scribing this very closely...)
<DaveB> eric epxlaining what yacker does
<DaveB> had trouble using javacc grammar
<DaveB> machine generated something that can validate sparql
Issue seems to be whether there is a link in the spec to more than one grammar
Editorial disagreement put to the WG (yes?)
<ericP> http://www.w3.org/2005/01/yacker/uploads/sparqlTest/bnf?lang=perl
AndyS prefers to reccomend one, rather than >1 grammars
Jeen: can we add an informative link to other grammar, but only one normative grammar
?
Very confusing, Eric.
<jeen> EricP wants to add both grammars to WD, let users 'vote' on which should be in Rec.
I'm confused: is the point to help yacc users, to have 2 grammars, or both?
these are v. diff!
Souri: do these 2 grammars describe exactly the same language?
Andy: No.
Kendall: I want you guys as editors to pick one and recommend it. Period.
Souri: Doesn't mind 2 grammars, minds 2 languages
other grammar proposal w/drawn
<DaveB> andy's action done re rq23 unionqu decision
Eric: is caught up or will be soon
Andy has a comment from DanC and one from teh comments list pending
There are some markup issues
<DaveB> andys said - need to sync with xmlres namespace when it gets a datespace one
Most of the pending stuff seems a matter of small edits
<DaveB> reviewer SH regrets for today
<DaveB> KC - nothing hold up for last call
<DaveB> JB - mostly editorial, seem addressed
JB: comments mostly editorial, all of which seem to have been addressed. Pretty much happy.
SteveH: possible issue re: precision of ints & decimals...
er, this is via EricP's recollection
We don't know for certain Steve's review of the doc's status
<ericP> last of the @@s removed from section 11
Discussion of creating a namespace URI for the results bindings format
rq23 mentions this namespace, so it needs to be an acceptable one
Gathering advice about LC timing...
Andy: wants there to be a period during which all 3 docs are LC simultaneously
DaveB: seems okay w/ that
<DaveB> kc - would like overlap to be a month
<DaveB> and protocol to have a longer lc
<DaveB> kc - lc period for query should be subsantial, 2-3 mths
<DaveB> and think protocol can go to lc in the next month
<DaveB> ... queyr is clean, good but complex. seen buyin from rdql users, impl but not from other parts of community
I'd like to see a 10 to 12 week LC period for the query doc, with at least 4 weeks of simultaneous LC for all 3 docs.
<DaveB> Revision 1.395 2005/06/14 15:19:37 eric
<DaveB> got rid of last @@s in section 11
Proposed: to publish rq23 (1.395)
as a LC doc
... to publish rq23 (1.395), addressing FILTER and minor
editorial changes, as a LC doc
KC, JosD volunteer to review (after the 17th for KC)
Proposed: to publish rq23 (1.395), addressing FILTER and 3 editorial changes, with KC & JosD reviewing these changes post 1.395, as a LC doc
publish: UMD, Bristol, HP, Agfa, NT&T, Howard the K, JeenB, W3C, (Souri in spirit!)
opposed: ()
abstains: ()
RESOLVED, go to last call for sparql ql
yay
<scribe> ACTION: AndyS to revise rq23 per fromUnionQuery decision [FINISHED] [recorded in http://www.w3.org/2005/06/14-dawg-minutes.html#action02]
<scribe> ACTION: ericp arrange publication [recorded in http://www.w3.org/2005/06/14-dawg-minutes.html#action03]
Jeen to scribe, DanC to chair meeting next Tuesday
<scribe> ACTION: ericp work w/ danc on SOTD and LC [recorded in http://www.w3.org/2005/06/14-dawg-minutes.html#action04]
<scribe> ACTION: danc message to chairs (?) [recorded in http://www.w3.org/2005/06/14-dawg-minutes.html#action05]
<scribe> ACTION: AndyS editorial changes to rq23 [recorded in http://www.w3.org/2005/06/14-dawg-minutes.html#action06]
<scribe> ACTION: KendallC to review post 1.395 doc [recorded in http://www.w3.org/2005/06/14-dawg-minutes.html#action07]
<scribe> ACTION: JosD to review post 1.395 doc [recorded in http://www.w3.org/2005/06/14-dawg-minutes.html#action08]
did I get them all?
(er, some of those may want to be combined...)
<DanC_mtg> I'm not available next tuesday. I said as much in email.
<DaveB> meeting over
<DaveB> DanC_mtg: oops...
<DanC_mtg> oh well... if there's energy for meeting next week, I can find a chair again
<DaveB> ericP can you drive Zakim or RRSagent to print actions and make irc logs public?
This is scribe.perl Revision: 1.126 of Date: 2005/05/16 16:49:48 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/agendas/actions/ No ScribeNick specified. Guessing ScribeNick: kendall Inferring Scribes: kendall WARNING: No "Present: ... " found! Possibly Present: Andy AndyS DanC_mtg DaveB Eric Example Hiroyuki HiroyukiS Howard JB JosD KC Kendall Kendall_Clark P29 P30 P31 P33 P37 Proposal Proposal2 Proposed Souri SteveH VLDB abstains ericp howardk inserted jeen opposed publish You can indicate people for the Present list like this: <dbooth> Present: dbooth jonathan mary <dbooth> Present+ amy Got date from IRC log name: 14 Jun 2005 Guessing minutes URL: http://www.w3.org/2005/06/14-dawg-minutes.html People with action items: andys changes danc editorial ericp josd kendallc message WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option.[End of scribe.perl diagnostic output]