Chatlog 2009-08-25

From SPARQL Working Group
Jump to: navigation, search

See original RRSAgent log and preview nicely formatted version.

Please justify/explain all edits to this page, in your "edit summary" text.

<LeeF> Present: LeeF, AxelPolleres, AndyS, Prateek, kasei, SimonS, ivanh, KjetilK, EricP, AlexPassant
13:53:20 <trackbot> Meeting: SPARQL Working Group Teleconference
13:53:20 <trackbot>  Date: 25 August 2009
13:53:22 <LeeF> zakim, this will be SPARQL
13:53:22 <Zakim> ok, LeeF; I see SW_(SPARQL)10:00AM scheduled to start in 7 minutes
13:53:26 <LeeF> Chair: Axel
14:00:39 <AxelPolleres> AxelPolleres has changed the topic to:
14:02:22 <LeeF> Regrets: AndyS, SteveH, pgearon, LukeWM, Chimezie
14:03:07 <AxelPolleres>
14:03:48 <ericP> scribenick: ericP
14:04:20 <ericP> topic: Admin
14:04:43 <ericP> AxelPolleres: want to see if we can get a pub by september
14:04:52 <AxelPolleres> PROPOSED: Approve minutes at
14:05:08 <AxelPolleres> RESOLVED: Approve minutes at
14:05:32 <AlexPassant> ok
14:05:33 <ericP> next meeting: 1 Sept (note this corrects the agenda)
14:05:41 <ericP> next scribe: AlexPassant 
14:05:41 <ivanh> Regrets for next meeting
14:05:51 <ericP> topic: Liason reports
14:06:14 <ericP> nada
14:06:21 <ericP> topic: Action Tracker
14:06:53 <ericP> -> pending actions
14:07:17 <LeeF> My action is still valid & pending.
14:07:19 <ericP> AxelPolleres: action 16 on Lee appears stale
14:07:31 <LeeF> Not stale.
14:07:35 <ericP> AxelPolleres: action 19 on iv_an_ru appears stale
14:07:38 <LeeF> Except in time :)
14:08:19 <Prateek> Prateek has joined #sparql
14:08:26 <LeeF> Half done
14:08:48 <Zakim> +Prateek
14:08:49 <LeeF> Just leave it, I think
14:09:14 <AndyS> q+
14:09:38 <LeeF> I have nothing to report on 77
14:10:01 <ericP> topic: Status report FWPDs end-of-September
14:10:52 <ericP> AndyS: do you mind adding special forms (e.g. coalese) to the functions and operators mission?
14:11:25 <ericP> ... the only special functions we have are &&, || and BOUND
14:11:56 <LeeF> Glitter has other functions besides those that are in this category (well, at least one)
14:12:22 <ericP> ... COALESCE also needs to see the variables (instead of its values)
14:13:50 <ericP> AxelPolleres: AndyS, please send mail reminding the group that you are extending action 75
14:14:05 <AxelPolleres> topic: SPARQL/Query
14:14:21 <AxelPolleres> 
14:14:25 <ericP> AndyS: first one is subselect
14:14:36 <ericP> ... steve and i discussed how we'd edit the doc
14:15:03 <ericP> ... since it's a refinment of the current doc, we'll work in HTML in CVS
14:15:20 <ericP> ... we'll work on major sections in the wiki
14:16:04 <AxelPolleres> ivanh: same space can be used as DAWG.
14:16:42 <AxelPolleres> ericP: let's take this offline, will check.
14:16:51 <ericP> ACTION: eric to figure out where to find SPARQL2 CVS edit space
14:16:51 <trackbot> Created ACTION-88 - Figure out where to find SPARQL2 CVS edit space [on Eric Prud'hommeaux - due 2009-09-02].
14:17:16 <ericP> AndyS: subselect looks pretty good
14:17:20 <ericP> ... need to sort out syntax
14:17:44 <ericP> ... basically, subselect has no prolog or dataclauses
14:17:58 <ericP> ... steve has tbe mappings from syntax to algebra
14:18:10 <LeeF> I thought we already decided against it?
14:18:22 <ericP> AxelPolleres: what about FROM in subselects?
14:18:49 <ericP> AndyS: if you don't have vars in datasetclauses, there's not much motivation
14:18:53 <SimonS> How about different default graphs?
14:18:56 <SimonS> q+
14:19:00 <AndyS> q-
14:19:02 <ivanh> q+
14:19:09 <ericP> LeeF: i thing we've already ruled them out
14:19:28 <ericP> SimonS: second motivation could be different default graph in the subquery
14:19:38 <ericP> q+ to mention query composability
14:19:45 <ivanh> ack SimonS
14:20:05 <ericP> AndyS: would like to motivated with missing usability of the lang
14:20:26 <ivanh> ack ivan
14:20:33 <ericP> ivanh: we can mark these as issues in the WG to solicit feedback from the public
14:20:43 <LeeF> q+ to talk about re-opening decided issues
14:22:25 <LeeF> q-
14:22:35 <LeeF> I note that is still open
14:22:38 <LeeF> Contrary to my memory.
14:23:17 <ericP> ericP: i assumed a major motivation for subselects was to have query composability, so scripts could be composed with concatonation
14:23:38 <ericP> LeeF: consensus in earlier discussion was that there was no reason to re-define the dataset
14:24:15 <ericP> ... we can just write it down in the issues and move on
14:24:22 <ericP> ... we'll have to address it before last call
14:25:19 <ericP> AxelPolleres: there is only one proposal for variable scoping
14:25:57 <ericP> AndyS: [re aggregates] have a list of issues
14:26:09 <ericP> ... multisets v.s bags (choosing multisets)
14:26:24 <ericP> ... syntax around project expression (ericP's prob)
14:26:58 <ericP> ... have some minor issues mapping from abstract syntax to algebra issues
14:27:47 <ericP> ... we're looking at things that syntactically look like functions, but we're safe because uris are unique [and cheap - ED]
14:27:58 <AxelPolleres>
14:29:08 <ericP> ericP: created and tested a grammer, using AS var
14:29:26 <ericP> LeeF: project_Expressions should own the syntax
14:29:36 <Zakim> -KjetilK
14:30:05 <AxelPolleres> issue there if binindg to AS variable already exists
14:30:06 <ericP> AndyS: only algebra implication is that they create var bindings, so what if they already exist
14:30:34 <ericP> (fascist) LeeF: propose: requires an AS, requires a var, and error if the var is used
14:30:54 <LeeF> ISSUE: What happens when expressions are projected with the same name as an existing variable?
14:30:54 <trackbot> Created ISSUE-36 - What happens when expressions are projected with the same name as an existing variable? ; please complete additional details at .
14:31:07 <LeeF> ACTION: Axel to start thread on mailing list re: ISSUE-36
14:31:07 <trackbot> Created ACTION-82 - Start thread on mailing list re: ISSUE-36 [on Axel Polleres - due 2009-09-01].
14:31:07 <ericP> AndyS: Virtuoso uses a slightly different syntax so need to take to mailing list
14:31:44 <AxelPolleres>  Negation:
14:31:48 <LeeF>
14:32:39 <ericP> AndyS: [Re: Negation] last time we straw polled, there was emphasis on the NOT EXISTS and we said we'd mention MINUS
14:32:59 <ericP> ... am working on formalizing the relationship between the two
14:33:43 <ericP> ericP: found diffs between them?
14:34:10 <ericP> AndyS: yes in OPTIONALs and FILTERs cases
14:34:43 <ericP> ... same issues that arrise in vars introduced in OPTIONALs
14:35:04 <ericP> AxelPolleres: add a pointer to that email thread?
14:35:26 <ericP> AndyS: not sure it will make a difference
14:35:46 <ericP> AxelPolleres: propose a link to concrete examples
14:35:49 <ivanh> q+
14:35:56 <ericP> q-
14:36:12 <ericP> AndyS: prefer to leave to editors to see what they can get done
14:36:31 <ericP> ivanh: current negation page is inconsistent with the examples we want to use
14:36:43 <ericP> ... NOT EXISTS, !EXISTS, UNSAID
14:37:11 <LeeF> UNEXISTS
14:37:12 <LeeF> :-D
14:37:27 <ericP> AndyS: pref [NOT] EXISTS
14:37:36 <ericP> ivanh: does NOT show up elsewhere?
14:37:39 <LeeF> +1
14:37:42 <ivanh> 0
14:37:53 <AxelPolleres> strawpoll on [NOT] EXISTS
14:37:55 <AxelPolleres> +1
14:38:05 <SimonS> +1
14:38:10 <kasei> 0
14:38:18 <Prateek> +1
14:39:26 <ericP> ivanh: EXISTS { .. } is usefull
14:39:36 <AxelPolleres> P1 EXISTS P2 != P1 P2
14:39:38 <ericP> ... (not as useful as negation, but still useful)
14:39:57 <AxelPolleres> strawpoll !EXISTS 
14:40:02 <ivanh> -1
14:40:07 <AxelPolleres> 0
14:40:13 <LeeF> -1
14:40:27 <SimonS> -1
14:40:29 <kasei> 0
14:40:53 <AndyS> -1
14:41:01 <AxelPolleres> strawpoll UNSAID
14:41:02 <ericP> ericP: "!" vs. "NOT" seems syntactic but atomic EXISTS is a new feature
14:41:03 <ivanh> +1
14:41:08 <LeeF> 0
14:41:09 <AxelPolleres> 0
14:41:09 <AndyS> -1
14:41:14 <ericP> +1
14:41:18 <kasei> +1
14:41:29 <SimonS> +1
14:41:43 <AlexPassant> +1 on exists / not exists
14:42:29 <AxelPolleres> andyS: have already sent an example on EXISTS
14:42:32 <ericP> ivanh: if we are motivated by EXISTS, then sure, "NOT EXISTS", otherwise two keywords is funny and potentially confusing
14:42:58 <ericP> ericP: we need to not just see it, but decide if we're motivated by it
14:43:15 <AxelPolleres> topic: Errata
14:44:03 <ericP> LeeF: AndyS and i have lists
14:44:20 <ericP> ... we don't need to handle them by FPWD
14:44:37 <AxelPolleres> ACTION: Lee to synchronize list of Errata with AndyS
14:44:37 <trackbot> Created ACTION-83 - Synchronize list of Errata with AndyS [on Lee Feigenbaum - due 2009-09-01].
14:45:00 <ericP> AndyS: would like to sync before FPWD
14:45:16 <ivanh> q+
14:45:17 <ericP> AxelPolleres: when do you think we'll have a 1st version for us to review?
14:45:28 <ericP> AndyS: steve and i haven't disucssed that
14:45:42 <ericP> ... maybe by the end of september
14:46:05 <ericP> ... we can vary any two of { time , resources , content }
14:46:20 <ericP> ... would like text on each mandatory feature
14:46:38 <ericP> ... not necessarily consistent, noted in the SOTD
14:46:38 <AxelPolleres> q?
14:47:11 <ericP> ivanh: FPWD will contain only new features? or will it integrate with SPARQL1?
14:47:36 <ericP> AndyS: timeliness might dictate just the new features. haven't consulted steve
14:48:10 <ericP> ... following structure of SPARQL1 until someone persuades us otherwise
14:48:34 <ericP> ivanh: a delta doc for FPWD is easier, quicker, and potentially better for public feedback
14:48:43 <ericP> ... does it cause us more work later?
14:49:27 <ericP> AndyS: will give progress reports at every conference
14:49:27 <AxelPolleres> ACTION: AndyS and SteveH to decide whether FPWD should be inegrated or only new features
14:49:28 <trackbot> Created ACTION-84 - And SteveH to decide whether FPWD should be inegrated or only new features [on Andy Seaborne - due 2009-09-01].
14:50:43 <ericP> SimonS: we integrate now (to save later integration) and hilight new sections
14:50:51 <AxelPolleres> Topic: SPARQL/Update
14:51:08 <SimonS>
14:51:28 <ericP> SimonS: plan is to extend SPARUL submission
14:51:36 <ericP> ... plan to use CVS from the start
14:52:49 <AxelPolleres> pls use q, thanks!
14:53:02 <ivanh> q+
14:53:05 <SimonS> q+
14:53:09 <ivanh> ack ivan
14:53:10 <AndyS> q+ to ask what's the cost of retro fitting the XSLT process to the existing SPARQL/Query
14:53:28 <ericP> ericP: [mentions xml-spec]
14:53:47 <ericP> ivanh: sharing text between the two motivates using the same approach for both
14:53:55 <AxelPolleres> q?
14:54:26 <AndyS> EricP: Choices: CVS+xml-spec , CVS+HTML, wiki.
14:54:35 <AxelPolleres> ack SimonS
14:54:42 <ericP> M$?
14:55:26 <AxelPolleres> ack AndyS
14:55:26 <Zakim> AndyS, you wanted to ask what's the cost of retro fitting the XSLT process to the existing SPARQL/Query
14:55:26 <ericP> SimonS: didn't know about xml-spec, but prefer to use same authoring tool
14:56:14 <AndyS>
14:58:28 <AxelPolleres> ACTION: Eric to look into xmlspec for SPARQL/Query with AndyS
14:58:28 <trackbot> Created ACTION-85 - Look into xmlspec for SPARQL/Query with AndyS [on Eric Prud'hommeaux - due 2009-09-01].
14:59:04 <KjetilK> Zakim, mute me
14:59:04 <Zakim> KjetilK should now be muted
14:59:09 <ericP> AxelPolleres: what's a reasonable ETA for Update?
14:59:33 <ericP> SimonS: Paul is on vacation for two weeks, but i could create a first version in the mean time
15:00:01 <AxelPolleres> ACTION: SimonS to draft first version of SPARQL/Update by 2009-09-15 
15:00:01 <trackbot> Created ACTION-86 - Draft first version of SPARQL/Update by 2009-09-15  [on Simon Schenk - due 2009-09-01].
15:00:03 <AndyS> EricP:
15:00:22 <AndyS> q+ to ask where SPARQL/Update/Protocol goes for FPWD?
15:01:02 <ericP> AndyS: where will the ReSTful update stuff go?
15:01:28 <ericP> SimonS: also had that question
15:01:43 <ericP> q+ to propose protocol
15:01:53 <AxelPolleres> Can we put this on email? I need to run.
15:01:56 <ericP> SimonS: expected another doc, or protocol
15:02:09 <ericP> ... prefer separate, no strong opinion