13:59:18 <RRSAgent> RRSAgent has joined #ldp
13:59:18 <RRSAgent> logging to http://www.w3.org/2014/08/11-ldp-irc
13:59:20 <trackbot> RRSAgent, make logs public
13:59:20 <Zakim> Zakim has joined #ldp
13:59:22 <trackbot> Zakim, this will be LDP
13:59:22 <Zakim> ok, trackbot; I see SW_LDP()10:00AM scheduled to start in 1 minute
13:59:23 <trackbot> Meeting: Linked Data Platform (LDP) Working Group Teleconference
13:59:23 <trackbot> Date: 11 August 2014
13:59:50 <Zakim> SW_LDP()10:00AM has now started
13:59:58 <Zakim> +Arnaud
14:00:28 <Zakim> +Matt
14:00:37 <deiu> Zakim, Matt is me
14:00:37 <Zakim> +deiu; got it
14:00:38 <Zakim> +EricP
14:00:55 <Zakim> +Ashok_Malhotra
14:01:16 <deiu> Zakim, mute me s'il te plait
14:01:16 <Zakim> I don't understand 'mute me s'il te plait', deiu
14:01:26 <deiu> bah, learn some French then
14:01:30 <deiu> Zakim, mute me please
14:01:30 <Zakim> deiu should now be muted
14:01:48 <Zakim> +[IBM]
14:01:58 <Zakim> +Alexandre
14:01:59 <SteveS> Zakim, [IBM] is me
14:02:00 <Zakim> +SteveS; got it
14:03:02 <Zakim> +Sandro
14:05:19 <ericP> scribenick: ericP
14:05:19 <Zakim> +OpenLink_Software
14:05:25 <TallTed> Zakim, OpenLink_Software is temporarily me
14:05:25 <Zakim> +TallTed; got it
14:05:30 <TallTed> Zakim, mute me
14:05:30 <Zakim> TallTed should now be muted
14:06:28 <nmihindu> nmihindu has joined #ldp
14:06:55 <ericP> RESOLVED: approve the minutes of 4 Aug
14:06:58 <Arnaud1> http://www.w3.org/2013/meeting/ldp/2014-08-04
14:07:41 <bblfish> bblfish has joined #ldp
14:07:42 <SteveS> I’m not going to SemTech, can make this meeting
14:07:48 <ericP> next meeting: 18 Aug
14:08:09 <ericP> topic: Open Actions
14:08:46 <ericP> sandro: [re: action 145 -- figure out where JSON-LD context goes]
14:09:16 <Zakim> +??P11
14:09:18 <ericP> ... i need to send mail to the list saying what we're going to do
14:09:24 <Zakim> +[IPcaller]
14:09:39 <ericP> ... anyone know who will provide the @context
14:09:42 <nmihindu> Zakim, ??P11 is me
14:09:42 <Zakim> +nmihindu; got it
14:09:55 <nmihindu> Zakim, mute me
14:09:55 <Zakim> nmihindu should now be muted
14:09:57 <ericP> Arnaud: iirc, nandana will provide the file and update the primer accordingly
14:10:17 <nmihindu> Sandro, I can provide the context file.
14:10:57 <ericP> ericP: [re: action 147]
14:11:02 <Arnaud> action-147
14:11:03 <trackbot> action-147 -- Eric Prud'hommeaux to Follow up on yves's way to turn 2nn into a number without waiting for the draft to go to rfc -- due 2014-08-04 -- OPEN
14:11:03 <trackbot> http://www.w3.org/2012/ldp/track/actions/147
14:11:12 <ericP> ... anyone ahve contacts within the IETF Applications WG
14:11:32 <codyburleson> codyburleson has joined #ldp
14:12:46 <ericP> action: ericP to ask John Arwe (on a Wednesday) how to switch to IETF applications WG
14:12:46 <trackbot> Created ACTION-149 - Ask john arwe (on a wednesday) how to switch to ietf applications wg [on Eric Prud'hommeaux - due 2014-08-18].
14:13:28 <Zakim> +[IPcaller.a]
14:13:37 <codyburleson> zakim, IPcaller.a is me
14:13:37 <Zakim> +codyburleson; got it
14:13:55 <ericP> topic: Paging Spec
14:14:14 <ericP> Arnaud: we were almost ready to publish last week.
14:14:40 <ericP> ... but John pointed out that we forget to set and end date for the review period
14:15:28 <ericP> ... there's one outstanding comment on the paging size: commenter would rather control the number of resources rather than triples
14:15:51 <ericP> Ashok: agree with commenter
14:16:01 <ericP> ... what if it's a non-RDF resource?
14:17:36 <TallTed> +1 limit really *should* be based on bytes, not triples, not resources
14:17:51 <TallTed> Zakim, unmute me
14:17:51 <Zakim> TallTed should no longer be muted
14:17:53 <ericP> sandro: i hadn't thought about Ashok's point, but I think that the original comment is not that big a deal because the number of triples and number of resoruces in a basic container is either 1:1 or 2:1
14:18:23 <ericP> ... but i strongly believe we should use bytes 'cause that's what machine's have to allocate
14:18:41 <SteveS> I wonder why we are opening this up, as commentor isn’t suggesting we change to bytes
14:18:49 <ericP> Arnaud: the mechanism allows us to use different units
14:19:25 <ericP> sandro: my arg is that the client only knows the amount of available memory.
14:20:00 <sandro> sandro: my arg is that paging is offered because of limits on bandwidth and storage, both of which are allocated in bytes
14:20:03 <ericP> ... counter arg is that the server has an easier time calculating the number of triples
14:20:59 <codyburleson> Speaking is Miguel.
14:21:00 <ericP> codyburleson: client can set limits on both
14:21:39 <ericP> sandro: we don't say in the spec what happens if you provide multiple limits. logical choice is the stop at the first limit the server hits
14:22:07 <ericP> Arnaud: these things can be fine-tuned in another version of the spec, but if we want it in this version, now's the time
14:22:18 <ericP> ... we can also add it as "at risk"
14:23:58 <MiguelAraCo> MiguelAraCo has joined #ldp
14:24:16 <bblfish> makes sense to me that one should be able to use bytes for client
14:24:27 <ericP> PROPOSED: client and send both triple and byte-count limits. server advised to stop at first one it hits.
14:24:42 <TallTed> s/client and send/client can send/
14:24:55 <ericP> PROPOSED: marked at risk: client can send both triple and byte-count limits. server advised to stop at first one it hits.
14:26:19 <ericP> PROPOSED: marked at risk: client can send both triple and byte-count limits. server advised to stop at first one it hits. At-risk fallback is to have only triple limits.
14:27:13 <sandro> sandro: and we'll ask the commenter if this is what they meant / really want
14:28:11 <ericP> +1
14:28:12 <sandro> +1
14:28:21 <deiu> +0
14:28:22 <MiguelAraCo> +1
14:28:26 <nmihindu> +0
14:28:28 <betehess> +0.0
14:28:30 <SteveS> +0
14:28:36 <TallTed> +1
14:28:41 <bblfish> +0.7
14:29:18 <ericP> Ashok: can we add resouce-count limits?
14:29:29 <ericP> Arnaud: we can invent all sorts of units
14:29:32 <ericP> APPROVED: marked at risk: client can send both triple and byte-count limits. server advised to stop at first one it hits. At-risk fallback is to have only triple limits.
14:29:32 <bblfish> What's the point?
14:31:06 <ericP> ericP: can the client calculate the number of triples for any member limit?
14:31:13 <bblfish> q+
14:31:24 <ericP> Arnaud: no, on the first page, you might get lots of non-member triples
14:32:25 <ericP> sandro: do we have a standard terminology for this?
14:32:29 <sandro> PROPOSED: Also include at risk, item-count (which is member or contained resource)
14:32:30 <ericP> ... propose "item count"
14:33:11 <ericP> SteveS: we're adding two features for one comment. concearned about feature creep.
14:33:31 <ericP> Arnaud: true, but we are responding to other concearns within the WG.
14:33:45 <sandro> PROPOSED: Also include at risk, item-count (which is members / contained resources, only defined for containers)
14:33:46 <ericP> ... i'd rather have more stuff At-Risk than do another last call
14:34:27 <ericP> bblfish: the word that you're looking for is the rows in SPARQL result sets.
14:34:38 <sandro> Or member-count, knowing what we mean.
14:34:56 <ericP> ... and the next version can [require SPARQL and] include special requests on graph parts
14:35:19 <Ashok> +1 to member count
14:35:32 <sandro> PROPOSED: Also include at risk, member-count (which is members / contained resources, only defined for containers)
14:35:46 <ericP> +1
14:35:46 <codyburleson> +1 to member count
14:36:01 <sandro> ... Prefer: return=representation; page-size="500 rdf-triples"
14:36:04 <SteveS> -0 Thinking rdf-triples is a close enough approx, bytes helps for other needs, adding yet another doesn’t fix too much in my mind
14:36:12 <sandro> ... Prefer: return=representation; page-size="500 members"
14:36:18 <sandro> ... Prefer: return=representation; page-size="500000 bytes"
14:36:36 <sandro> kB
14:36:39 <sandro> kbytes
14:36:44 <bblfish> -0.4
14:36:49 <TallTed> +0.5
14:36:50 <betehess> +0
14:36:56 <nmihindu> +0
14:37:11 <Ashok> +1
14:37:13 <sandro> +1 members
14:38:24 <sandro> RESOLVED: We'll add text saying that if muliple page-size arguments are present, the server is advised to take the first reached.   We also add AT RISK two new unites "members" and "kbytes".
14:38:29 <deiu> +0
14:39:06 <Arnaud> Proposal: Publish 2nd Last Call with end date for the review period on 15 September 2014
14:40:11 <sandro> +1
14:40:13 <ericP> +1
14:40:17 <deiu> +1
14:40:20 <nmihindu> +1
14:40:56 <betehess> +0 ship it!
14:41:00 <SteveS> +1
14:41:09 <TallTed> +1
14:41:26 <Arnaud> Resolved: Publish 2nd Last Call with end date for the review period on 15 September 2014
14:41:49 <ericP> topic: Access Control
14:42:05 <ericP> Ashok: disagreement on one issue: ACLs on triples
14:42:12 <betehess> I think the requirement (on triple) is a valid requirement
14:42:24 <ericP> ... i feel this has to be included, possibly as OPTIONAL
14:42:32 <sandro> teach the controversy   :-)
14:42:39 <ericP> ... after that, we're ready to go
14:43:28 <ericP> Arnaud: pushback has been that in HTTP-land, we use resource-level ACLs
14:43:57 <ericP> sandro: since this is a Note, let's "teach the controversy", i.e. say that the WG doesn't agree on this and here are the arguments.
14:44:19 <ericP> ... specifically adding that some people feel that it's a bad practice.
14:45:13 <betehess> I completely agree with TallTed
14:45:39 <ericP> TallTed: we only say "triple-level" once. after that it's a hand-wave. we don't need to specify this.
14:46:37 <bblfish> q+
14:47:26 <betehess> since when gathering valid use-cases became something controversial???
14:47:39 <Arnaud> ack bblfish
14:47:42 <ericP> ericP: we're disagreeing about whether to mention the arguments on either side
14:47:59 <ericP> Ashok: i'd rather not as this is a high-level, abstract document
14:48:19 <ericP> bblfish: there are many ways to read "fine-graind", e.g. per-triple graphs
14:49:23 <TallTed> q+
14:49:35 <Arnaud> ack TallTed
14:49:45 <ericP> Arnaud: if you look into the document, there are lots of things where we might want to spell out technical points
14:49:54 <ericP> TallTed: this feels like implementation details.
14:50:51 <ericP> sandro: [we should say more because] this is the first time we're having a substantive conversation about ACLs.
14:51:25 <ericP> bblfish: we have three impls at the resource level
14:51:31 <ericP> ... doing further could be a lot more work
14:51:51 <ericP> Ashok: it could be more work but if it's worthwhile...
14:52:00 <betehess> doing nothing works as well, but it doesn't solve any use-case
14:54:25 <Arnaud> PROPOSED: Publish Access Control draft as FPWD, as is
14:54:48 <ericP> +0.8
14:54:50 <betehess> +1
14:54:56 <deiu> +1
14:54:58 <Ashok> +1
14:54:59 <nmihindu> +1
14:55:04 <TallTed> +1
14:55:06 <codyburleson> +0
14:55:10 <SteveS> +1
14:55:25 <Arnaud> RESOLVED: Publish Access Control draft as FPWD, as is
14:55:37 <TallTed> different people seeing different things falls right into Open World -- you have to always expect that you're not seeing everything there is to know about Entity:X
14:55:58 <ericP> topic: admin
14:56:20 <ericP> Arnaud: BP doc still waiting for approval -- team help please
14:56:38 <ericP> Arnaud: we are stuck until we get implementation reports
14:57:07 <ericP> topic: LDP Patch
14:57:36 <betehess> strawpoll?
14:59:33 <ericP> sandro: can we poll more widely?
14:59:38 <betehess> this approach was focused on the requirements we agreed on. I would be ok with the other solutions if the requirements were different
14:59:50 <ericP> ... who do we want to ask who would understand the poll with less than 10 mins study?
15:00:07 <ericP> ... sparql vendors would be natural to ask but they would be biased
15:00:32 <ericP> Arnaud: LDPatch representes the status quo so the burden is on other proposals
15:01:11 <ericP> sandro: we didn't give a blank check, more conditional based on it being simple
15:01:42 <ericP> Arnaud: paging is out of the way. next week we'll focus on LDP Patch.
15:02:53 <ericP> sandro: i would like see how complicated use cases are handled in LDPatch
15:03:12 <bblfish> Sounds good, there could be a number of these, with results in the different formats
15:03:15 <ericP> Arnaud: can you post challenges to the LDP Patch folks and we can use those to decide?
15:03:19 <Zakim> -SteveS
15:03:20 <Zakim> -nmihindu
15:03:20 <Zakim> -Ashok_Malhotra
15:03:21 <Zakim> -Alexandre
15:03:21 <Zakim> -Sandro
15:03:24 <Zakim> -TallTed
15:03:25 <Zakim> -[IPcaller]
15:03:27 <Zakim> -deiu
15:03:28 <codyburleson> codyburleson has left #ldp
15:03:29 <Zakim> -Arnaud
15:03:51 <MiguelAraCo> MiguelAraCo has joined #ldp
15:03:56 <Zakim> -EricP
15:04:01 <Zakim> -codyburleson
15:04:01 <Zakim> SW_LDP()10:00AM has ended
15:04:01 <Zakim> Attendees were Arnaud, deiu, EricP, Ashok_Malhotra, Alexandre, SteveS, Sandro, TallTed, [IPcaller], nmihindu, codyburleson
15:20:16 <betehess> sandro, any estimate when you'll send the examples or challenge how "simple" LD Path for users and implementers?
15:20:30 <sandro> In about five minutes
15:20:33 <betehess> cool
15:20:37 <betehess> I appreciate
15:46:54 <SteveS> SteveS has joined #ldp
17:03:10 <Zakim> Zakim has left #ldp
18:54:04 <deiu> deiu has joined #ldp