13:58:44 <RRSAgent> RRSAgent has joined #ldp
13:58:44 <RRSAgent> logging to http://www.w3.org/2014/09/15-ldp-irc
13:58:46 <trackbot> RRSAgent, make logs public
13:58:48 <trackbot> Zakim, this will be LDP
13:58:48 <Zakim> ok, trackbot, I see SW_LDP()10:00AM already started
13:58:49 <trackbot> Meeting: Linked Data Platform (LDP) Working Group Teleconference
13:58:49 <trackbot> Date: 15 September 2014
13:59:28 <Zakim> +Matt
13:59:34 <deiu> Zakim, Matt is me
13:59:34 <Zakim> +deiu; got it
14:00:15 <JohnArwe> JohnArwe has joined #ldp
14:00:23 <Zakim> + +1.857.928.aabb
14:00:29 <betehess> Zakim, aabb is Alexandre
14:00:29 <Zakim> +Alexandre; got it
14:00:39 <Zakim> +JohnArwe
14:00:46 <Zakim> +Arnaud
14:00:52 <MiguelAraCo> MiguelAraCo has joined #ldp
14:01:15 <Zakim> +[IPcaller]
14:01:23 <codyburleson> Zakim, IPcaller is me.
14:01:23 <Zakim> +codyburleson; got it
14:01:44 <Ashok> Ashok has joined #ldp
14:01:46 <sergio> sergio has joined #ldp
14:01:56 <Zakim> +[IBM]
14:02:05 <SteveS> Zakim, [IBM] is SteveS
14:02:05 <Zakim> +SteveS; got it
14:02:31 <Zakim> +Ashok_Malhotra
14:02:32 <Zakim> + +43.660.274.aacc
14:02:45 <sergio> Zakim, +43.660.274.aacc is me
14:02:45 <Zakim> +sergio; got it
14:02:57 <nmihindu> nmihindu has joined #ldp
14:03:26 <Arnaud> zakim, who's on the phone?
14:03:26 <Zakim> On the phone I see +1.505.603.aaaa, Sandro, deiu (muted), Alexandre, JohnArwe, Arnaud, codyburleson, SteveS, Ashok_Malhotra, sergio
14:03:28 <Zakim> +OpenLink_Software
14:03:40 <TallTed> Zakim, OpenLink_Software is temporarily me
14:03:40 <Zakim> +TallTed; got it
14:03:42 <TallTed> Zakim, mute me
14:03:42 <Zakim> TallTed should now be muted
14:04:20 <Arnaud> scribe: codyburleson
14:04:28 <sandro> like this s/old/new/
14:05:00 <deiu> 3 dots actually
14:05:04 <Arnaud> chair: Arnaud
14:05:07 <sandro> 2 or 3
14:05:16 <SteveS> …Arnaud continues on
14:05:19 <Arnaud> agenda: https://www.w3.org/2012/ldp/wiki/Meetings:Telecon2014.09.15
14:05:32 <codyburleson> TOPIC: Approval of Minutes from Sep 8 meeting.
14:05:37 <azaroth> zakim, aaaa is me
14:05:37 <Zakim> +azaroth; got it
14:05:59 <codyburleson> Arnaud: RESOLVED: Minutes of Sep 8 approved. (no objection heard)
14:06:28 <Zakim> +??P2
14:06:32 <codyburleson> TOPIC: Actions and Issues
14:06:43 <nmihindu> Zakim, ??P2 is me
14:06:43 <Zakim> +nmihindu; got it
14:06:50 <nmihindu> Zakim, mute me
14:06:50 <Zakim> nmihindu should now be muted
14:06:53 <codyburleson> TOPIC: Paging
14:07:19 <codyburleson> Arnaud: Paging is now in last call and we have received a couple of comments; one, in fact, which I forgot to put on the agenda to discuss.
14:09:01 <codyburleson> … We have to prove to W3C mgmt that we have kept track of all the comments we have received. There is a tracker tool for this, but I haven't found it very helpful. We are free to use whatever tool we want. In my opinion, it would be better just to use a wiki page and list the comments (links) and outcome for each. Thoughts?
14:09:10 <SteveS> I’m fine with using wiki for this, unfortunate the tool isn’t that helpful
14:09:34 <codyburleson> Ashok: The wiki ought to be OK.
14:09:58 <codyburleson> Arnaud: OK. We'll use the wiki.
14:10:11 <codyburleson> … OK. First comment on Paging.
14:10:36 <SteveS> http://lists.w3.org/Archives/Public/public-ldp-comments/2014Aug/0002.html
14:11:09 <codyburleson> … The comment is about - Why do you guys don't choose something closer to HTTP where you use HTTP Range.
14:11:45 <codyburleson> … We need to agree on how we respond to this and then somebody needs to send the response. So, how do we want to respond?
14:12:30 <azaroth> q+
14:12:34 <sandro> q+
14:12:45 <roger> roger has joined #ldp
14:12:55 <Zakim> +Roger
14:13:22 <codyburleson> JohnArwe: I don't think it's practical for RDF. It goes against the grain of RDF. Sets are unordered and every graph is a set of triples. Also, HTTP Range, from every thing I've heard, is not well implemented. That is to say, not a lot of servers implement it at all. I think this is why we threw it out a couple of years ago; we had this discussion long ago.
14:13:29 <Arnaud> ack azaroth
14:14:00 <codyburleson> Roger: I agree. RDF not being supported seems like a problem for paging in general.
14:14:10 <Arnaud> ack sandro
14:14:35 <azaroth> s/Roger/azaroth/
14:14:48 <TallTed> range changed while we were working...  http://svn.tools.ietf.org/svn/wg/httpbis/specs/rfc7233.html
14:15:04 <TallTed> particularly http://svn.tools.ietf.org/svn/wg/httpbis/specs/rfc7233.html#range.units.other
14:15:53 <codyburleson> Sandro: I'm having a hard time actually seeing a problem with range. 2 possible problems. I don't know how to do server initiated paging with range. The other is that we lose the ability to have metadata on pages, but I haven't been able to come up with a compelling use case.
14:16:10 <azaroth> s/in general./in general. So order isn't a good response for the comment./
14:17:08 <TallTed> Zakim, unmute me
14:17:08 <Zakim> TallTed should no longer be muted
14:17:39 <codyburleson> Sandro: One of the problems with Paging is that it doesn't solve: you can't traverse a resource while it's changing in any safe way.
14:17:48 <sandro> sandro: doesn't allow server initiation
14:18:09 <sandro> sandro: doesn't allow getting the date safely while things are being added/removed
14:18:22 <codyburleson> TallTed: It's based on something, which is static to some degree. We're doing this hybrid sort of thing. Decisions we've already made make using range more difficult.
14:19:32 <codyburleson> Sandro: RDF graph is changed often by adding or deleting a few triples.
14:20:16 <codyburleson> Arnaud: I think the difference is what you guys are touching on. With Paged resources, like what we have in our model, there is an indirection, which you lose if you go down to the level of triples as you likely would with HTTP Range.
14:20:45 <codyburleson> JohnArwe: I don't think Jena gives me any guarantee I'm gonna get the same amount of triples in the same order.
14:22:00 <betehess> you need more than a Jena Graph/Model to do paging
14:22:01 <codyburleson> Sandro: In the general case, with a Jena based server, how would you do paging? First 10,000 triples, from that first batch, how would you continue?
14:22:09 <azaroth> +q to ask about max-triple-count param
14:22:57 <codyburleson> JohnArwe: What I would probably do is a sort of imperfect compromise. It gets ugly, I admit. But you could iterate through…. use predicate names themselves to divide the graoph into thin slices, and just keep track of the predicates you've already gone through.
14:23:08 <codyburleson> Sandro: What if it's pretty much all the same predicates.
14:23:28 <Ashok> q+
14:24:20 <Arnaud> ack azaroth
14:24:20 <Zakim> azaroth, you wanted to ask about max-triple-count param
14:24:29 <sandro> sandro: another problem with range: our data structures don't generally support integer indexing
14:24:32 <codyburleson> Sandro: I guess an answer here is that the data structures typically used by this sort of thing are not easy to implement. There is an implementation burden trying to add an extra index.
14:25:24 <JohnArwe> I agree w/ Rob
14:26:04 <Arnaud> ack Ashok
14:26:58 <codyburleson> Ashok: If you do not want to order triples, then you get some kind of default order. And if you are storing this stuff in any tables, there is always an order. You can take advantage of that.
14:27:34 <codyburleson> Sandro: Ashok, is it reasonable for any table to ask for the arbitrary Nth row? In your technology is that something you can easily do?
14:27:38 <codyburleson> Ashok: Yes.
14:27:58 <codyburleson> TallTed: There is no assurance the order is going to be set unless you put an ORDER BY.
14:28:12 <codyburleson> Ashok: That's what the SQL standard "says", yes.
14:28:45 <codyburleson> Arnaud: So…Do we have an answer for the commenter?
14:28:57 <betehess> afaik, SQL doesn't expose the row number to the user
14:29:09 <betehess> not even the row id
14:29:30 <codyburleson> … the level of indirection we're getting from our model makes it easier to implement than if we were going to use HTTP Range with triples.
14:29:40 <Ashok> Yes, Alexandre, but many implementations support this
14:29:51 <sandro> proposed answer:    our data structures don't expose integer indexes
14:30:08 <sandro> + server initiation
14:30:20 <sandro> + safe changes during paging
14:30:28 <codyburleson> … I'm suggesting that's our answer to the commenter. My first reaction was "Yeah! Why don't we use this?" But I am convinced now that there are advantages to the model we have adopted so far. Some might think it is overkill but...
14:31:00 <codyburleson> Sandro: 1) Our data structures … indexes 2) Don't know how to do server initiation with paging 3) Don't know how to do safe changes while paging
14:31:18 <codyburleson> … I'm not sure that's a comment back to the commenter or a question back to him or her.
14:31:22 <sandro> "We don't know how to do these things"
14:32:08 <codyburleson> Arnaud: We can make it both ways. We state that this is WHY we have what we have and leave it open for the commenter to continue the thread.
14:32:28 <JohnArwe> SQL idiom included in this sample: SELECT * FROM MyTable OFFSET 50 LIMIT 25
14:32:40 <codyburleson> Arnaud: Sandro, can you answer the commenter?
14:33:14 <codyburleson> Sandro: I think I should bring this up with the HTTP working group first.
14:33:19 <sandro> ericP, you around
14:33:24 <codyburleson> JohnArwe: How long will that take?
14:33:45 <codyburleson> Sandro: It's completely unpredictable; I don't know.'
14:33:58 <codyburleson> JohnArwe: We have a 3 week last call period.
14:34:28 <codyburleson> Arnaud: OK. We will go with HTTP Working Group.
14:34:42 <codyburleson> … Another comment on Paging…
14:35:03 <SteveS> http://lists.w3.org/Archives/Public/public-ldp-comments/2014Sep/0003.html
14:35:50 <SteveS> sounds like we need a paging best practices doc
14:36:21 <azaroth>  /me was thinking the same thing
14:36:24 <codyburleson> Arnaud: The commenter is asking for us to document a comparison of options and why we made the choices we made.
14:36:42 <codyburleson> JohnArwe: Is that asking for A W3C Note?
14:36:50 <codyburleson> Arnaud: I think an email will do.
14:37:55 <codyburleson> JohnArwe: We're taking a conservative approach that is least likely to break anything.
14:38:15 <codyburleson> … and we have two levels of ETag. On the page and the container.
14:38:24 <codyburleson> Arnaud: Can you reply to this email, John?
14:38:45 <codyburleson> JohnArwe: I can put something out by Wed this week.
14:40:22 <codyburleson> Arnaud: I suggest John responds to this email unless anyone has any differing opinions that what John has expressed.
14:42:05 <Zakim> -Roger
14:43:04 <JohnArwe> s/that what/than what/
14:43:22 <codyburleson> TOPIC? 2NN?
14:44:05 <codyburleson> Sandro: On that particular problem, the RC says that any client that does not recognize a 2NN code for any 2NN has to treat it as a 200.
14:44:18 <JohnArwe> Topic: 2nn
14:44:22 <codyburleson> Arnaud: What is the status of the 2NN draft?
14:44:31 <JohnArwe> it's an ietf draft
14:45:08 <codyburleson> Sandro: …and the working group is kind of engaging an email on it. It's hard to tell.
14:45:50 <JohnArwe> ietf draft statuses are more opaque from the outside than w3c, imo
14:46:52 <codyburleson> Sandro: I am willing to engage more with the HTTP group on it. But, I am not personally too motivated for 2NN on paging.
14:47:22 <codyburleson> TOPIC: TEst Suite
14:47:37 <codyburleson> s/TEst Suite/Test Suite
14:47:56 <codyburleson> Arnaud: I think there are like 14 tests that are waiting for approval?
14:47:57 <SteveS> http://w3c.github.io/ldp-testsuite/manifest/#tobeapproved
14:48:39 <codyburleson> (There are 19 ready for approval on the list.)
14:49:14 <codyburleson> Arnaud: I'd like to approve those tests. Anybody object?
14:49:15 <SteveS> https://dvcs.w3.org/hg/ldpwg/raw-file/default/tests/reports/ldp.html
14:49:30 <codyburleson> SteveS: There are a number of implementations reporting to pass those tests already.
14:49:56 <azaroth> +1 to approve
14:50:03 <codyburleson> Arnaud: Sounds like we can approve those tests then. Any concerns? Are we rushing it?
14:50:24 <Arnaud> PROPOSAL: approve new tests listed at http://w3c.github.io/ldp-testsuite/manifest/#tobeapproved
14:50:36 <azaroth> +1
14:50:36 <sergio> +1
14:50:37 <SteveS> +1 approving the 17 tests there
14:50:38 <deiu> +1 (most seemed fine to me)
14:50:42 <TallTed> +1
14:50:44 <betehess> +1
14:50:44 <nmihindu> +1
14:50:46 <JohnArwe> +1
14:50:52 <codyburleson> +1
14:51:00 <Arnaud> RESOLVED: : approve new tests listed at http://w3c.github.io/ldp-testsuite/manifest/#tobeapproved
14:51:00 <MiguelAraCo> +1
14:51:08 <sergio> cool
14:51:23 <JohnArwe> SS trying to get word in
14:51:24 <SteveS> thanks all, sergio I can make the change
14:51:28 <codyburleson> TOPIC: Status of various documents
14:51:44 <SteveS> I was just going to say that I fixed the link in the agenda
14:51:48 <sergio> perfect, SteveS
14:51:50 <codyburleson> Arnaud… LDP Spec is about to be published to LAST CALL.
14:52:04 <codyburleson> … Same is true for Access Control Use Cases Requirements Note
14:52:07 <JohnArwe> SS, did you make the pub request for main-LDP for tomorrow?
14:52:29 <codyburleson> … LDP PATCH format: We got the request on getting the short name approved
14:52:29 <SteveS> JA, no..I will now
14:52:53 <codyburleson> … all of our documents have been published now or in the middle of process now.
14:52:59 <SteveS> if not Tues, I wills shoot for Thurs, etc
14:53:24 <SteveS> s/wills/will/
14:53:42 <codyburleson> Arnaud: LDP Spec - we are waiting for implementation reports. We are republishing it as Last Call with indirect containers marked as At Risk.
14:54:19 <codyburleson> Arnaud: We had some problems on Primer with JSON-LD examples.
14:54:29 <JohnArwe> nandana's on IRC...
14:54:32 <nmihindu> Arnaud, the JSON-LD examples are fixed
14:54:36 <codyburleson> … it would be nice to publish an update on the Primer with those updates fixed.
14:54:57 <codyburleson> SteveS: I had some minor side comments that Nandana and Roger were going to work in.
14:55:04 <codyburleson> Arnaud: Nandana responded.
14:55:06 <nmihindu> I am on the call, I can hear you
14:55:16 <nmihindu> yes, we can publish an updated version.
14:55:33 <codyburleson> Arnaud: We should publish the updated draft of the Primer.
14:55:37 <deiu> q+
14:55:43 <deiu> Zakim, ack me
14:55:43 <Zakim> unmuting deiu
14:55:44 <Zakim> I see no one on the speaker queue
14:55:45 <Arnaud> ack deiu
14:56:00 <JohnArwe> I'm off; see folks in a month
14:56:01 <nmihindu> Roger is making some updates to the introductions section, as soon as it is done, we can publish another version.
14:56:04 <Zakim> -JohnArwe
14:56:10 <codyburleson> deiu: The date for LDP Patch document should be October 18th?
14:56:23 <betehess> what if we're ready for tomorrow?
14:56:25 <codyburleson> Arnaud: It should be Sept 18th. That's a mistake.
14:57:12 <betehess> yes, we can probably still do that for tomorrow
14:57:19 <codyburleson> …change the date to Sept 16th.
14:57:36 <codyburleson> Arnaud: I think we're done for the day.
14:57:45 <Zakim> -Ashok_Malhotra
14:57:50 <codyburleson> … I'll put together the wiki page for the Last Call comments.
14:58:04 <Zakim> -TallTed
14:58:08 <Zakim> -deiu
14:58:09 <Zakim> -Sandro
14:58:09 <azaroth> Thanks all :)
14:58:09 <Zakim> -codyburleson
14:58:12 <Zakim> -sergio
14:58:13 <Zakim> -SteveS
14:58:14 <Zakim> -Arnaud
14:58:15 <Zakim> -Alexandre
14:58:21 <azaroth> azaroth has left #ldp
14:58:31 <Zakim> -azaroth
15:20:50 <bblfish> bblfish has joined #ldp
15:35:00 <Zakim> disconnecting the lone participant, nmihindu, in SW_LDP()10:00AM
15:35:01 <Zakim> SW_LDP()10:00AM has ended
15:35:01 <Zakim> Attendees were +1.505.603.aaaa, Sandro, deiu, +1.857.928.aabb, Alexandre, JohnArwe, Arnaud, codyburleson, SteveS, Ashok_Malhotra, sergio, TallTed, azaroth, nmihindu, Roger
15:54:18 <bblfish> bblfish has joined #ldp
16:02:01 <bblfish_> bblfish_ has joined #ldp
16:17:08 <bblfish> bblfish has joined #ldp
16:35:55 <JohnArwe> JohnArwe has joined #ldp
16:52:22 <SteveS> SteveS has joined #ldp
17:05:22 <Zakim> Zakim has left #ldp
17:05:33 <Arnaud> trackbot, make logs public
17:05:33 <trackbot> Sorry, Arnaud, I don't understand 'trackbot, make logs public'. Please refer to <http://www.w3.org/2005/06/tracker/irc> for help.
17:05:51 <Arnaud> rssagent, make log public
17:06:07 <Arnaud>  rrsagent, make log public
17:38:09 <jmvanel> jmvanel has joined #ldp
18:23:02 <bblfish> bblfish has joined #ldp
18:25:03 <SteveS> SteveS has joined #ldp
19:39:28 <bblfish> bblfish has joined #ldp
20:39:33 <bblfish> bblfish has joined #ldp
21:24:44 <deiu> deiu has joined #ldp
21:41:19 <bblfish> bblfish has joined #ldp
22:39:29 <bblfish> bblfish has joined #ldp
23:39:28 <bblfish> bblfish has joined #ldp