13:59:11 RRSAgent has joined #ldp 13:59:11 logging to http://www.w3.org/2014/09/22-ldp-irc 13:59:13 RRSAgent, make logs public 13:59:13 Zakim has joined #ldp 13:59:15 Zakim, this will be LDP 13:59:15 ok, trackbot, I see SW_LDP()10:00AM already started 13:59:16 Meeting: Linked Data Platform (LDP) Working Group Teleconference 13:59:16 Date: 22 September 2014 14:00:16 azaroth has joined #ldp 14:00:43 +Arnaud 14:01:25 Ashok has joined #ldp 14:01:51 zakim, who is here? 14:01:51 On the phone I see Sandro, ericP, azaroth, Arnaud 14:01:53 On IRC I see Ashok, azaroth, Zakim, RRSAgent, pchampin, betehess, deiu, TallTed, bblfish, SteveS, jmvanel, Arnaud, tommorris, taaz, sandro, Yves, ericP, trackbot 14:01:56 +deiu 14:01:57 +[IBM] 14:01:58 Zakim, [IBM] is me 14:01:58 +SteveS; got it 14:02:09 + +1.617.838.aaaa 14:02:15 +Alexandre 14:02:29 Zakim, aaaa is me 14:02:30 +TallTed; got it 14:02:33 Zakim, mute me 14:02:34 TallTed should now be muted 14:03:20 +Ashok_Malhotra 14:05:50 Or forward public-ldpatch list to public-ldp? 14:07:02 zakim, who's on the phone? 14:07:05 On the phone I see Sandro, ericP, azaroth, Arnaud, deiu, SteveS, TallTed (muted), Alexandre, Ashok_Malhotra 14:07:35 la crème de la crème 14:08:26 +[IPcaller] 14:08:40 ok 14:08:45 scribenick: deiu 14:09:12 nmihindu has joined #ldp 14:09:19 Arnaud: I haven't had a chance to look at everything re. ld-patch, as the emails were sent at night 14:09:28 ... I woke up to a pile of emails 14:09:55 sandro: [...planet is round...] 14:10:04 didn't see anything wrong there 14:10:09 Arnaud: objections to minutes from last week? 14:10:19 ...minutes are approved 14:10:31 ...we're meeting next week, Sept. 29th 14:10:46 ...we can skip the Actions and Issues 14:10:51 Topic: LDP Spec 14:11:20 Arnaud: the LC document was published and we need to keep track of the disposition of comments 14:11:32 ...I suggested we do not use the LC tracker and people agreed 14:11:38 ...we'll just use the wiki page 14:11:48 ...I created the first wiki page (with some pain re. styling) 14:12:07 ...I reproduced the document we get as the outcome of our LC tracker 14:12:33 sandro: have you seen how other WGs did that on their wikis? 14:12:44 Arnaud: not really, I'd love to get a pointer if you have one 14:12:54 +??P14 14:13:00 ...the RDF WG had some nice ones 14:13:04 Zakim, ??P14 is me 14:13:04 +nmihindu; got it 14:13:10 Zakim, mute me 14:13:10 nmihindu should now be muted 14:13:13 Reto’s comment be considered a LC3 comment? http://lists.w3.org/Archives/Public/public-ldp/2014Sep/0006.html 14:13:26 Arnaud: basically styling sucks 14:13:57 ...we will keep track of the compliance report; as of now we have enough to go to PR without indirect container 14:14:13 ...we'll see whether we get an implementation by the time we get to PR 14:14:40 -> http://www.w3.org/2011/rdf-wg/wiki/Turtle_Candidate_Recommendation_Comments behold the *fabulous* Turtle CR Comments 14:14:53 SteveS: there was a comment (from Reto) and I don't know if we should treat that as a LC issue 14:14:57 Arnaud: I remember, yes 14:15:20 SteveS: I think the we could provide more clarity there (re. interaction model) 14:15:31 Arnaud: I agree we can treat it as a LC comment 14:16:05 ...I would like that you (SteveS) respond to it and see what the spec needs, so we can improve the text 14:16:08 Topic: Paging 14:16:29 Arnaud: the document is not accurate yet, but I suppose you guys saw that John did reply and it generated quite a thread 14:16:38 ...the two threads are kind of merged now 14:16:54 q+ 14:17:07 ...it seems that in the end they're not saying we got it completely wrong...but we may need to ask if they're ok with this moving forward or if we need to clarify it more 14:17:15 Ashok: I added a couple of comments 14:17:55 Arnaud: should we discuss those comments now? 14:18:05 ...unfortunately John is away for 3 weeks 14:18:38 Arnaud: we have to pick up the slack and deal with the comments while he's not here 14:19:28 Arnaud: we should still answer Ashok's questions but we don't have to treat them as LC comments since Ashok is a WG member 14:19:34 Arnaud: let's discuss them 14:20:03 Ashok: there was one which was asking whether we want to allow the client to specify the ordering (e.g. ORDER BY in SQL) 14:20:14 Arnaud: we don't do that, so that's the answer 14:20:30 Ashok: the question that I could not remember was whether we had spoken about it 14:20:36 Arnaud: yes we had 14:20:55 ...it would seem logical but it is server driven for now 14:21:09 Ashok: I'm a bit concerned about it 14:21:32 ...is the order set by the server arbitrary? 14:21:38 Arnaud: it depends on the server 14:21:51 ...there might be an application specific order 14:22:24 SteveS: we need to take a holistic view of how the client can request the contents of a container to see what's in it but also to filter the response 14:22:50 ...I thought we agreed to discuss that once we finish working on the "core" of paging 14:22:54 Zakim, unmute me 14:22:54 TallTed should no longer be muted 14:23:07 Arnaud: you can come up with a proposal 14:23:51 Zakim, mute me 14:23:51 TallTed should now be muted 14:23:51 TallTed: does it make any sense to at put a disclaimer there, since this issue will keep coming up in the future? 14:23:58 Arnaud: agreed 14:24:23 Ashok: can the client find the type of ordering set by the server? 14:24:40 Arnaud: yes, for containers -- you cannot set the types but you get get it 14:25:26 ...we should still take an action to add a note though 14:25:43 ...could you please do that, SteveS? 14:25:49 SteveS: yes, I'll do it 14:25:54 q? 14:25:58 ack Ashok 14:26:07 Ashok: one more question: Steve, you're going to add filtering too, right? 14:26:31 SteveS: yes, that's correct, and the filtering part could go into the LDP spec, and the sorting can go into paging 14:26:39 Zakim, unmute me 14:26:39 TallTed should no longer be muted 14:26:58 ...is TallTed's point a general one or was it just for the paging spec? 14:27:12 SteveS: we can put it into the non-normative intro for containers 14:27:16 [everyone agrees] 14:27:17 Zakim, mute me 14:27:17 TallTed should now be muted 14:27:27 Arnaud: shall we talk 2NN now or ld-patch? 14:27:41 ericP: I can be quick about 2nn 14:28:27 ericP: the gatekeepers for http-2 are Roy Fielding and Mark N. 14:29:01 ...for them is perfectly fine re. fragment ids (@@ was that right?) 14:29:26 ...they said they for paging we should be using Range (byte or anything else on the planet -- i.e. unicode) 14:29:45 ...we can invent a scheme with an ETag and a range 14:30:07 ...the effect is that we keep doing GETs with different ranges and we don't have to change the URL 14:30:33 ...we asked about a question re. what is metadata applied to, but didn't really get a reply yet 14:30:55 sandro: I started a new thread about range, maybe you've missed it? 14:31:18 ...after our last meeting on the 15th, I sent an email 14:31:28 ...most people have refuted the reasons I gave 14:31:54 ...we may have to create some 4xx codes 14:32:25 ...we don't know how to do safe changes with range and I don't think anyone had any refutations about it 14:33:02 ...you delete an item and you say there is no "triple 7" anymore and you don't shift things back 14:33:06 ...you keep adding 14:33:20 ...with insert you don't really lose data, you get duplicates 14:33:44 ...third item was that our systems don't do integer retrieval -- no refutations there 14:34:27 ...also, the Web infrastructure support non-byte ranges -- caching will probably not produce errors 14:35:02 sandro: defining a new range unit requires the same process: write an RFC and start a WG 14:35:12 Arnaud: so what's the conclusion? 14:35:35 sandro: my emotional answer is "I'm done with this" 14:36:39 Arnaud: we can just mark it as a feature as risk for now and just leave the redirect there 14:37:01 ericP: ok, so maybe 2nn is not for paging but people might still want 2nn 14:37:10 +1 to not tying 2NN to Paging 14:37:25 ... so the text has to change from paging use-cases to non-IR use-cases 14:37:47 q? 14:38:06 sandro: we just need to bring in fresh blood who gets to argue about it outside our use-cases 14:38:37 ericP: so the use-cases are going to be around non-IRs so it's possible that the best way to make that happen is through civil disobedience 14:39:00 ...here's a little spec that tells you how to do it, so go argue with these guys (IETF I assume?) 14:39:10 Arnaud: I could see Tim doing that 14:40:11 sandro: Roy's email also said that people will probably go ahead and still implement 2nn, which doesn't sound like a formal objection 14:40:27 Arnaud: wasn't there another use-case about packaging? 14:40:53 ericP: yes, you get some Web app and when you get f00, you get back f00 + a bunch of other resources 14:41:16 ...the right answer to that is probably SPDY 14:41:34 sandro: we've waited 25 years already, we can wait a bit more 14:41:59 ericP: what's the best interim solution? 2nn which may break caching or something else? 14:42:17 Arnaud: I would not suggest we do it anyway, we can just mark it as "as risk" 14:42:24 s/as risk/at risk 14:43:21 sandro: I don't understand how IETF gets consensus because they don't have meetings in that WG, it seems the decision belongs to whoever gets the last say on the mailing list 14:43:45 ericP: the answer is that there is no answer 14:44:03 sandro: I suggest you talk to Mark and ask him to give us the code and see how it works 14:44:21 ericP: I suspect sandro has a bit more authority to reply to that thread 14:44:23 sandro: ok 14:45:15 Arnaud: ok, let's move on 14:45:54 Arnaud: I added to links to the agenda 14:46:07 ...we're getting some reaction 14:46:21 ...people are sending suggestions on how to do it 14:46:35 ...what do we need to talk about, betehess? 14:46:50 betehess: I don't think it's worth speaking about Kingsley's email 14:47:10 ...the email from David Booth was expected, since we've been discussing that issue in the group 14:47:12 http://bertails.org/2014/09/20/why-ldpatch 14:47:16 ...I wrote an article about that 14:47:25 ...for me it all boils down to the list of requirements 14:47:28 http://bertails.org/2014/09/20/why-ldpatch#requirements 14:48:01 ...I agree that if the group or someone else can make the point that the requirements are not valid, we can stop the work on ld-paging 14:48:17 s/ld-paging/ld-patch/ 14:49:05 Arnaud: David doesn't really add something that makes a difference at this point, we've been talking about SPARQL 14:49:12 MacTed has joined #ldp 14:49:47 q? 14:50:06 Arnaud: Henry is not here? 14:50:23 bblfish: I've been wondering about another way of doing things -- standardizing bnodes 14:50:43 ...we know that patching triples is simple if we have persistent bnodes 14:50:56 ...I don't get the skolemization in RDF 1.1 14:51:28 ...I just sent a few mails out asking what the limitations are, since I believe it hasn't been explored enough 14:51:30 do we want to go again through the skolemization discussion today? 14:52:03 betehess: we've discussed skolemization several times and we agreed (as a group) that we won't go that way 14:52:46 bblfish: I don't think skolemization is such a big problem (in terms of complexity) 14:53:09 sandro: that only works if the graph is exactly the same, and some of our use-cases involve graphs that have changed 14:53:51 bblfish: imagine you have a patch with a WHERE clause that identifies a path to a bnode 14:54:50 ...[not making too much sense] 14:55:57 Arnaud: to sum it up, bblfish has this idea about skolemization and we also got the comments from David Booth 14:56:22 ...and then we got betehess who reminds us that most of those comments have already been discussed in the group 14:57:19 Ashok: do we actually have to specify that you have to use ld-patch? what if you want to use a SPARQL Update subset? 14:57:32 Arnaud: indeed, we don't say anything about it 14:57:49 ...conneg allows you to specify different formats 14:58:27 ...in the current spec there are no MUSTs associated to a specific format, and it doesn't stop anyone from using other formats 14:58:28 -SteveS 14:58:54 Arnaud: we're about done with today's call; I encourage people to keep answering comments on the list 14:59:11 ...anything else? 14:59:27 ...some good news: we published the piece on ACL 14:59:39 ...and also, all our documents have now been published! yey 14:59:42 -Ashok_Malhotra 14:59:43 -nmihindu 14:59:45 -[IPcaller] 14:59:47 -Alexandre 14:59:47 Thanks! :) 14:59:48 [meeting adjourned] 14:59:50 -Sandro 14:59:52 -ericP 14:59:54 -TallTed 14:59:55 -Arnaud 14:59:57 -deiu 15:00:03 -azaroth 15:00:04 SW_LDP()10:00AM has ended 15:00:04 Attendees were Sandro, ericP, azaroth, Arnaud, deiu, SteveS, +1.617.838.aaaa, Alexandre, TallTed, Ashok_Malhotra, [IPcaller], nmihindu 15:01:03 azaroth has left #ldp 15:32:26 SteveS has joined #ldp 16:58:14 Zakim has left #ldp 20:03:28 SteveS has joined #ldp 20:35:24 SteveS has joined #ldp 20:36:26 SteveS has joined #ldp 21:25:12 bblfish_ has joined #ldp 21:53:56 SteveS has joined #ldp