13:54:46 RRSAgent has joined #ldp 13:54:46 logging to http://www.w3.org/2013/03/25-ldp-irc 13:54:48 RRSAgent, make logs public 13:54:48 Zakim has joined #ldp 13:54:50 Zakim, this will be LDP 13:54:51 Meeting: Linked Data Platform (LDP) Working Group Teleconference 13:54:51 Date: 25 March 2013 13:54:51 ok, trackbot; I see SW_LDP()10:00AM scheduled to start in 6 minutes 13:55:47 AndyS1 has joined #ldp 13:57:27 SW_LDP()10:00AM has now started 13:57:34 +[IPcaller] 13:57:50 IPCaller is Cody 13:57:56 +??P2 13:58:36 zakim, IPCaller is really Cody 13:58:36 +Cody; got it 13:58:54 Zakim, ??P2 is nmihindu 13:58:54 +nmihindu; got it 13:59:31 +SteveBattle 13:59:52 +??P7 14:00:02 BartvanLeeuwen has joined #ldp 14:00:46 +??P11 14:00:57 +Arnaud 14:01:05 +bblfish 14:01:18 +??P14 14:01:29 Zakim, ??P14 is me 14:01:29 +BartvanLeeuwen; got it 14:01:29 hi 14:01:30 Ashok has joined #ldp 14:01:31 +OpenLink_Software 14:01:34 +[IBM] 14:01:35 Zakim, OpenLink_Software is temporarily me 14:01:35 +TallTed; got it 14:01:37 hi all 14:01:37 Zakim, mute me 14:01:37 TallTed should now be muted 14:01:42 zakim, code? 14:01:42 the conference code is 53794 (tel:+1.617.761.6200 sip:zakim@voip.w3.org), Ashok 14:01:48 zakim, [IBM] is me 14:01:48 +SteveS; got it 14:01:48 JohnArwe has joined #ldp 14:01:55 +JohnArwe 14:02:09 zakim, who is here? 14:02:09 On the phone I see Cody, nmihindu, SteveBattle, ??P7, ??P11, Arnaud, bblfish, BartvanLeeuwen, TallTed (muted), SteveS, JohnArwe 14:02:12 On IRC I see JohnArwe, Ashok, BartvanLeeuwen, Zakim, RRSAgent, cody, SteveS, TallTed, nmihindu, betehess, stevebattle, bblfish, bhyland, jmvanel, davidwood, Arnaud, thschee, 14:02:12 ... sandro, ericP, Yves, trackbot 14:02:16 dret has joined #LDP 14:02:27 +Ashok_Malhotra 14:02:29 hi 14:02:32 +[IPcaller] 14:02:33 Kalpa has joined #ldp 14:03:05 zakim, IP{Caller is me 14:03:05 sorry, dret, I do not recognize a party named 'IP{Caller' 14:03:16 zakim, IPCaller is me 14:03:16 +dret; got it 14:04:33 I can scribe, but I need to figure out how to tell who is talking at any given moment. 14:04:53 who is talking? 14:05:03 zakim, who is talking? 14:05:13 cody, listening for 10 seconds I heard sound from the following: Arnaud (24%), BartvanLeeuwen (9%), Ashok_Malhotra (38%) 14:05:21 scribe: cody 14:05:29 cody, it was BartvanLeeuwen 14:05:29 cody - you can vocally interrupt at any time, to ask who's speaking, clarification, etc. 14:05:51 Cody, the IRC queue should help you figure out who is speaking. 14:05:59 Arnaud, starting with the Minutes of March 11; propose we accept them. 14:06:02 dret_ has joined #LDP 14:06:05 Steve : + 1 14:06:22 Arnaud, hearing no objections, we hereby approve those minutes of March 11. 14:06:44 +1 to F2F minutes, look good 14:06:54 Arnaud, and the minutes from the Face to Face 2 14:07:07 Yes - they look OK 14:07:08 dret_ has joined #LDP 14:07:35 Arnaud, so I hereby approve the minutes of the F2F2, hearing no objections. 14:07:47 http://www.w3.org/2012/ldp/meeting/2013-03-13#Primer 14:07:54 dret has joined #LDP 14:07:59 +[GVoice] 14:08:07 Zakim, [GVoice] is me 14:08:07 +ericP; got it 14:08:21 pchampin has joined #ldp 14:08:46 +??P30 14:08:49 Arnaud: I will look and if I find it (regarding Primer), I will let you know. 14:09:48 Arnaud: As results from F2F2, we gave several actions. Want to check progress... 14:10:00 q+ 14:10:02 Arnaud: 2 pending actions: 1) propose use case for issue 33 14:10:08 ack stevebattle 14:10:33 Steve Battle: I am reviewing the draft and getting it into a form that is consitent with the use cases. 14:10:34 looking at actions http://www.w3.org/2012/ldp/track/actions/pendingreview 14:10:41 Arnaud, Very good. 14:10:55 s/consitent/consistent/ 14:10:57 Action-40 14:10:57 ACTION-40 -- Ashok Malhotra to propose how to modify the text of 4.1.4 to go with closing ISSUE-49 (with no material modification to spec) -- due 2013-03-18 -- PENDINGREVIEW 14:10:57 http://www.w3.org/2012/ldp/track/actions/40 14:11:35 Issue-49 14:11:35 ISSUE-49 -- Canonical URL - how to communicate its value to clients -- closed 14:11:35 http://www.w3.org/2012/ldp/track/issues/49 14:12:05 Ashok and Arnaud you were talking over each other 14:12:30 Ashok: I have recommended that we take action 414 out 14:13:11 Arnaud: Just remove the section or replace with other text (Section 4.1.4) 14:13:23 Ashok: Just take it out 14:13:27 I remember us talking about it 14:13:54 http://www.w3.org/2012/ldp/meeting/2013-03-15 says Close ISSUE-49 saying that LDP will not further restrict HTTP in this area. Remove section 4.1.4 from the spec and consider giving some guidance in the deployment guide. link 14:14:05 Arnaud: I put the resolution at the end of 49. Resolution says that we remove section 4.1.4 in the spec and consider giving some guidance in deployment guide 14:14:28 zakim, who is talking? 14:14:39 cody, listening for 10 seconds I heard sound from the following: Cody (58%), Arnaud (64%), SteveS (9%) 14:14:42 that was steve speicher 14:15:01 Arnaud: I am closing action 40. 14:15:10 action-43 14:15:10 ACTION-43 -- Steve Speicher to draft a use case for container ordering -- due 2013-03-20 -- PENDINGREVIEW 14:15:10 http://www.w3.org/2012/ldp/track/actions/43 14:15:37 Arnaud: Alright, so we can close Action 43 then. 14:15:59 Arnaud: Let me ask if there are any other open actions for which anyone wants to claim victory... 14:16:21 Arnaud: Hearing none. We have a few issues that were raised…starting with ISSUE-57 14:16:25 Issue-57 14:16:25 ISSUE-57 -- How can a client determine that it is in communication with an LDP server? -- raised 14:16:25 http://www.w3.org/2012/ldp/track/issues/57 14:16:46 oberger has joined #ldp 14:16:54 regrets: sandro, andys, roger 14:17:07 Arnaud: There was discussion about that in the F2F. Question was, "(How) Can I figure out if I am talking to an LDP Server?" 14:17:10 +1 open 14:17:11 We should open it. 14:17:19 Arnaud: Do we close it or open it? Any opinions? 14:17:30 -??P11 14:17:31 q+ 14:17:49 ack bart 14:18:08 "discovery" is the term we used during F2F 14:18:11 q? 14:18:15 BartvanLeeuwen: I was wondering if this was completely new. To me it sounds logical to add this. 14:18:20 +1 to open 14:18:36 +1 to open and discuss it 14:18:48 Arnaud: Do we need a separate issue? 14:19:02 Zakim, unumute me 14:19:02 I don't understand 'unumute me', TallTed 14:19:11 Zakim, unmute me 14:19:11 TallTed should no longer be muted 14:19:25 +1 open 14:19:31 Zakim, mute me 14:19:31 TallTed should now be muted 14:19:35 TallTed: Please open it. 14:19:48 -0 (I think it is covered by issue 32) 14:19:51 trackbot, reopen issue-57 14:19:52 Re-opened ISSUE-57 How can a client determine that it is in communication with an LDP server?. 14:19:52 Arnaud: Hearing no objections we hereby open issue 57. 14:20:03 issue-59 14:20:03 ISSUE-59 -- Reconsider usage of Aggregate/Composite construct to get predictable container delete behavior -- raised 14:20:03 http://www.w3.org/2012/ldp/track/issues/59 14:20:23 -??P7 14:20:38 dret has joined #LDP 14:21:06 No objection 14:21:11 Arnaud: Anyone want to object to opening this issue? 14:21:16 trackbot, reopen issue-59 14:21:16 Re-opened ISSUE-59 Reconsider usage of Aggregate/Composite construct to get predictable container delete behavior. 14:21:25 Issue-60 14:21:25 ISSUE-60 -- The specification does not allow GETting empty containers -- raised 14:21:25 http://www.w3.org/2012/ldp/track/issues/60 14:21:50 q+ 14:21:58 well, a set can be an empty set :-) 14:22:09 ack steve 14:22:16 Arnaud: This one is interesting. His point is, the text seems to indicate that a container cannot be empty. But I don't think there is any intention to prohibit empty containers. 14:22:18 "sets can be empty" was exactly cygri's response 14:22:48 looks like fixing the text is good enough, then? 14:22:51 sounds editorial to me 14:23:18 editorial to me as well 14:23:18 Arnaud: I think we agree that containers CAN be empty. If you don't agree, please speak up. 14:23:32 +1 to the change in concept. not sure the suggestion works, as written. 14:23:36 Arnaud: I don't think Raul's suggestion is the right one, by the way. 14:24:02 agree. better just add that the set can be empty 14:24:22 so +1 open issue with action for editor to come up with revision 14:24:27 Maybe add the text "possibly empty set" 14:24:28 +1 for clarrification 14:24:31 Arnaud: I suggest we close it with an action to the editor to come up with uh…. I think in respect to Raul's issue…I would like the editor to consider clarifying. 14:24:46 +! 14:24:49 +1 14:24:51 +1 14:24:52 Arnaud: I say we close it; with recommendation to editor to fix the spec to clarify intent 14:25:00 +1 14:25:01 +1 14:25:04 +1 14:25:05 +1 14:25:05 +1 14:25:06 +1 14:25:21 +1 14:25:27 +1 14:25:33 Arnaud: So, very good. I hereby declare ISSUE-60 closed. 14:25:55 Arnaud: I think that takles care of all the admin stuff. Moving on... 14:26:30 Which e-mails? 14:27:03 Arnaud: We got an email asking for the LDP working group to look at (two drafts ?) that may be of interest to us 14:27:03 http://lists.w3.org/Archives/Public/public-ldp/2013Mar/0019.html 14:27:35 "This document includes sections on service discovery and provenance "ping-back" 14:27:35 which may be related to mechanisms you are working to specify." 14:27:48 I can volunteer to take a look 14:27:50 I am interested in learning more about this, doing it for LDP WG might be a good way to force me to do it 14:27:50 Arnaud: At the very least, we need to reply. Just ignoring this email is not good form. 14:27:59 q+ 14:28:33 ack dret 14:28:46 Arnaud: Very good. nmihindu can look at it. 14:29:03 yes, we can't hear anything 14:29:05 dret, losing your voice in the bkground hiss 14:29:55 dret: I think it will be an interesting one to look at. 14:30:18 Arnaud: We'll give nmihindu and fret the action item to look at it. 14:30:30 q+ 14:30:33 Issue-13 14:30:33 ISSUE-13 -- Include clarifications about BPC representations that include member triples -- open 14:30:33 http://www.w3.org/2012/ldp/track/issues/13 14:30:38 Arnaud: Now, let's move on to trying to making progress on Open Issues. Starting with ISSUE-13 14:30:39 s/fret/dret 14:31:27 http://www.w3.org/TR/ldp/#general-1 14:33:20 Roger Menday: My objection was purely around ensuring the wording wasn 14:33:31 Roger: wasn't so ambiguous 14:33:51 speaker is steve battle 14:34:14 thx 14:34:17 q+ 14:34:18 Roger will not be with us for 3 weeks (via private email) 14:35:26 stevebattle: does anybody deny that there is two different clauses within that…? 14:35:36 ack steve 14:35:42 ack bblfish 14:37:14 http://lists.w3.org/Archives/Public/public-ldp-wg/2013Mar/0145.html 14:37:29 stevebattle: I posted an example to the email log. This is based on the Net Worth Example. 14:37:52 dret has joined #LDP 14:38:47 stevebattle: This is really about server managed properties 14:39:56 wondering: when you "sign" content, shouldn't you only be able to do that for content, and not for the member? i am saying that because a member is always under control of the server anyway, which might add metadata as it sees fit. it thus would be impossible for a client to do anything that depends on full control over that resource. 14:40:50 q? 14:40:50 stevebattle: A1 is the key member of this Net Worth container. That met data about A1 is not actually within A1. Now if you do a GET on A1, you can see it has a value of 100. If you do a GET on the container, you get all of those triples combined. You get A1 and you can see the values. I can't now do a PATCH on the container. I've actually got to do the patch on A1 itself. 14:41:07 I think SB said "NOT" about server managed props. Provenance being a good example, since LDP says nothing about Prov he assumes it is "user"-managed. 14:41:26 stevebattle: Maybe there needs to be an option when you ddl a GET on the container to prevent inlining. 14:41:29 AndyS has joined #ldp 14:42:23 Arnaud: I understand your example. I just don't know about the distinction your making by using "about" 14:42:26 q+ 14:42:41 q+ 14:42:46 stevebattle: The inline view is kind of the "constructed" view. It's not a "real" view. 14:42:53 ack pchampin 14:43:19 I'm fine with the suggested revision. servers MAY refuse such ... which implies that they MAY accept such, but it's not *required* that they accept (nor refuse). application specific might handle a PATCH against LDPC that actually targets inlined LDPR triples -- but compliance doesn't require it. 14:43:25 JohnArwe, i'd say that prov is server-managed. a particular server will have policies about what parts of the graph it provides 14:43:58 ack steves 14:44:06 ericP: Do we really want containers to have such a complex behavior? I think the current assumption in the spec is that you cannot add arbitrary triples to the container. I see the use-case of talking about the members somewhere. I'm wondering if containers is the right place to do that? 14:44:07 @ericp, I was not making any assertion other than the existence of a difference between what I heard @steveb say and what the scribe minutes. 14:44:14 s/minutes/minuted/ 14:44:33 s/ericP:/pchampin:/ 14:44:33 very good point: the data that containers accept (and thus can represent) should be controlled/limited. 14:44:59 cody, i don't think we have any choice. I had this issue with Annotea, where the dc:author property was tied to the http auth user who created an annotation 14:45:01 SteveS: My thought was… seems like we don't have…. would this be more about augmenting the PATCH section to what it means to PATCH containers? 14:45:02 That also makes sense, to add information about patching containers 14:45:08 Arnaud: You lost me, sorry. 14:45:35 SteveS: I'm wondering why there isn't a proposal for something in the PATCH section to add the clarification. 14:45:56 s/cody, i don't think/@JohnArwe, i don't think/ 14:46:14 This seems no different (some properties(triples) can be updated, others not) through a given URL than other cases like the generic "subject to access control" disclaimer. 14:46:31 i buy that 14:46:32 What was the proposal? 14:46:41 original: “Close the remainder of ISSUE-13 by saying that servers may refuse to update inlined members through PUT/PATCH to a container.” 14:46:52 amended: “Close the remainder of ISSUE-13 by saying that servers may refuse to update the content of an inlined LDPR through PUT/PATCH to a container.” 14:47:08 Arnaud: This was the original we tried to agree to. Then Steve sent an ammended version. 14:47:27 Arnaud: At this point, I don't think we have introduced the term "inlined resources" in the spec. 14:48:22 q+ 14:48:32 ack bblfish 14:48:32 stevebattle: If we go with the original proposal, we'd also need to change 5.2.6; which is describing exactly what you can save in a container 14:48:41 Arnaud: Henry? 14:48:58 "5.2.6 The representation of a LDPC MAY include an arbitrary number of additional triples whose subjects are the members of the container, or that are from the representations of the members (if they have RDF representations). This allows a LDPC server to provide clients with information about the members without the client having to do a GET on each member individually. See section 5.1.1 Container Member Information for additional 14:48:58 details." 14:48:58 so 14:48:58 “Close the remainder of ISSUE-13 by saying that servers may refuse to update such included triples through PUT/PATCH to the container.” 14:49:45 a particular app can close the world. for instance, if the data is backed by an RDB 14:50:01 very good point, bblfish. 14:50:19 bblfish: explained a concern (scribe didn 14:50:24 q+ 14:50:35 ack steves 14:50:39 actually, it is one that pretty much influences every single interaction in a RESTful design based on RDF. 14:50:56 +1 to "it won't search the internet" 14:51:17 bblfish: RDF works under the open world assumption; so I don't think we can state something like "here are *all* the triples about this resource 14:51:23 q+ 14:51:26 Zakim, unmute me 14:51:26 TallTed should no longer be muted 14:52:05 TallTed: What I am seeing as a distraction about "inline triples" is easy to eliminate by not using those terms. 14:52:21 bblfish, 'SELECT ?s ?p ?o { GRPAH { ?s ?p ?o } }' gives me all of the triples in 14:52:26 TallTed: I say changing "inline …whatever" to terminology that is already used. 14:53:00 q+ 14:53:02 my point was that in RDF ther is for a container to say that it has published all the relations that are in the element. So if a container has <> rdf:member , rdf does not make it possible to express that it has all the triples published inA 14:53:11 bblfish, likewise 'SELECT ?p ?o { ?p ?o }' gives me all of the triples in the data store's default graph which as a subject 14:53:27 ack tallted 14:53:31 ack steve 14:53:47 -BartvanLeeuwen 14:53:52 what's the latest proposal? 14:53:53 stevebattle: I'm happy with Ted's "amended amended" proposal. Saying the same thing as what I was saying. 14:54:07 And better worded :) 14:54:21 bblfish, ahh, what about { <> rdf:member ; foo:count 5 } ? 14:54:22 proposal: close issue-13, adding “Close the remainder of ISSUE-13 by saying that servers may refuse to update the content of an inlined LDPR through PUT/PATCH to a container.” to section 5.2.6 14:54:27 Close the remainder of ISSUE-13 by saying that servers may refuse to update such included triples through PUT/PATCH to the container. 14:54:45 TallTed: No that's not right. Hold on. 14:54:53 PROPOSAL: Close the remainder of ISSUE-13 by saying in 5.2.6 that servers may refuse to update such included triples through PUT/PATCH to the container. 14:55:14 stevebattle: yup. That's right. 14:55:19 ericp foo:count 5 , might work if is a document 14:55:35 Arnaud: Right, my copy/paste failed. Thank you. You're right. 14:55:39 +1 14:55:43 bblfish, but not if it's a ...? 14:55:44 would not work with <> rdf:member . 14:55:49 Arnaud: I'd like to hear from people on this proposal. 14:56:07 TallTed: This is where the arbitrary number of included triples comes from. 14:56:13 bblfish, why the distinction? 14:56:28 @bblfish: it would work! OWL has that. The problem is, it can leads to unexpected conclusions by inference engines. 14:56:32 because a document can contain triples, an object always can have an infinte number of tripels 14:56:36 Arnaud: As always, editors have some room for figuring out how to keep spec consistent (editing somewhere else if necessary to keep spec consistent) 14:56:36 +1 14:56:36 +1 14:56:39 +1 14:56:49 +1 14:56:53 +1 may need to update 5.5 PUT and 5.8 PATCH as well 14:56:54 +0.5 ( just because I have only thought of it today ) 14:56:56 +1 14:56:58 +1 14:57:05 +0 14:57:11 +1 14:57:21 resolved: Close the remainder of ISSUE-13 by saying in 5.2.6 that servers may refuse to update such included triples through PUT/PATCH to the container. 14:57:33 Zakim, mute me 14:57:33 TallTed should now be muted 14:57:38 Thanks Ted :) 14:57:44 pchampin: what does owl say? 14:58:02 Arnaud: I will call it resolved and leave it to the editors to select the resolution in the spec - along the lines of what is proposed. If there is more that needs to be done, then please do so. 14:58:12 owl, can create restrictions on the number of ojbects a relation has, not on the number of relations on an object 14:58:33 it's like owl:FunctionalPrperty 14:58:51 Arnaud: We had some discussion on ISSUE-59 14:58:57 Issue-59 14:58:57 ISSUE-59 -- Reconsider usage of Aggregate/Composite construct to get predictable container delete behavior -- open 14:58:57 http://www.w3.org/2012/ldp/track/issues/59 14:59:02 Zakim - unmute me 14:59:08 Zakim, unmute me 14:59:08 TallTed should no longer be muted 14:59:30 Arnaud: Composite versus aggregate containers. We resolved, but most people don't seem to be happy with the status quo. 15:00:33 need to get on a different meeting; thanks everybody! 15:00:40 -dret 15:00:59 I like the proposal 15:01:02 The question I have is what would be the method for a recursive delete? 15:01:12 Arnaud: If I have thousands of triples in my container, and I delete the container, I want the server to delete them for me…. and this led to discussion of need for aggregate container. Ted, had some disagreements. So the proposal is to forget the separation between composite and aggregation. Please, look at the email thread on this and try to indicate what you agree and don't agree to on this. 15:01:24 Yes, be interested to hear about the part (a) and also the part (b) (recursive delete) 15:01:27 DELETE x?recursive 15:01:31 Arnaud: Please - look at that email. For this, I will close the call for now. 15:01:32 just an idea 15:01:38 bye 15:01:39 -Ashok_Malhotra 15:01:53 -JohnArwe 15:02:03 A cool, a Retry-After 15:02:18 -SteveS 15:02:20 Can you POST info on that in HTTP/1.1 15:02:23 -SteveBattle 15:02:32 Kalpa has left #ldp 15:02:41 http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.37 15:03:43 zakim, who is talking? 15:03:55 cody, listening for 10 seconds I heard sound from the following: Cody (15%), TallTed (48%), ericP (15%) 15:04:27 -bblfish 15:04:30 ericP: is not rude to raise issues without prior discussion? 15:04:35 -ericP 15:04:36 -Arnaud 15:04:36 -TallTed 15:04:45 Arnaud: no. What's rude is to Open it, but not raise it, no. 15:05:12 cody has left #ldp 15:05:58 stevebattle has joined #ldp 15:08:41 bhyland has joined #ldp 15:10:51 -Cody 15:44:08 oberger has joined #ldp 16:06:25 stevebattle2 has joined #ldp 16:28:47 SteveS has joined #ldp 17:40:24 jmv has joined #ldp 17:46:57 gavinc has joined #ldp