13:57:28 <RRSAgent> RRSAgent has joined #ldp
13:57:28 <RRSAgent> logging to http://www.w3.org/2014/05/05-ldp-irc
13:57:33 <trackbot> RRSAgent, make logs public
13:57:33 <Zakim> Zakim has joined #ldp
13:57:35 <trackbot> Zakim, this will be LDP
13:57:35 <Zakim> ok, trackbot; I see SW_LDP()10:00AM scheduled to start in 3 minutes
13:57:36 <trackbot> Meeting: Linked Data Platform (LDP) Working Group Teleconference
13:57:36 <trackbot> Date: 05 May 2014
13:58:39 <Zakim> SW_LDP()10:00AM has now started
13:58:46 <Zakim> +??P2
13:59:26 <Arnaud> agenda: http://www.w3.org/2012/ldp/wiki/Meetings:Telecon2014.05.05
14:00:13 <Zakim> +??P3
14:00:17 <BartvanLeeuwen> Zakim, ??p3 is me
14:00:17 <Zakim> +BartvanLeeuwen; got it
14:00:30 <Zakim> +Arnaud
14:00:43 <Ashok> Ashok has joined #ldp
14:01:49 <Zakim> +Ashok_Malhotra
14:04:07 <Zakim> +??P10
14:04:33 <Zakim> +JohnArwe
14:04:47 <JohnArwe> JohnArwe has joined #ldp
14:04:59 <pchampin> pchampin has joined #ldp
14:05:04 <Zakim> +Sandro
14:05:06 <pchampin> zakim, ??P10 is me
14:05:06 <Zakim> +pchampin; got it
14:05:59 <Zakim> +??P14
14:06:06 <deiu> Zakim, ??P14 is me
14:06:06 <Zakim> +deiu; got it
14:06:14 <deiu> Zakim, mute me please
14:06:14 <Zakim> deiu should now be muted
14:06:39 <Zakim> +bblfish
14:06:42 <Arnaud> zakim, who's on the phone?
14:06:42 <Zakim> On the phone I see ??P2, BartvanLeeuwen, Arnaud, Ashok_Malhotra, pchampin, JohnArwe, Sandro, deiu (muted), bblfish
14:06:45 <sandro> scribe: sandro
14:06:55 <Zakim> +OpenLink_Software
14:07:16 <TallTed> Zakim, OpenLink_Software is temporarily me
14:07:16 <Zakim> +TallTed; got it
14:07:17 <TallTed> Zakim, mute me
14:07:17 <Zakim> TallTed should now be muted
14:07:54 <Arnaud> Proposal: Approve the minutes of the 28 April teleconf: http://www.w3.org/2013/meeting/ldp/2014-04-28
14:08:13 <sandro> RESOLVED: Approve the minutes of the 28 April teleconf: http://www.w3.org/2013/meeting/ldp/2014-04-28
14:08:14 <bblfish> was not around for last month or so, due to moving house, computer breaking down, and other major headaches...
14:08:25 <sandro> Arnaud: meeting as usual next week
14:08:46 <sandro> issue-98?
14:08:46 <trackbot> issue-98 -- HTTP status code for application specific errors -- raised
14:08:46 <trackbot> http://www.w3.org/2012/ldp/track/issues/98
14:08:49 <JohnArwe> regrets: alexandre, sergio, steve speicher
14:09:01 <Zakim> +??P17
14:09:13 <nmihindu> Zakim, ??P17 is me
14:09:13 <Zakim> +nmihindu; got it
14:09:23 <nmihindu> Zakim, mute me
14:09:23 <Zakim> nmihindu should now be muted
14:09:46 <Zakim> -??P2
14:09:57 <sandro> Arnaud: Any victory on open actions?
14:10:19 <sandro> Arnaud: Cody?   ACTION-234?
14:10:24 <sandro> Arnaud: Cody?   ACTION-134?
14:10:53 <sandro> topic: ISSUE-98
14:11:32 <sandro> Arnaud: I proposed we close it along the lines of option 1 -- as a best practice.   The server should respond with "400" with rel=describedby
14:11:41 <sandro> sandro: 4xx or 400?
14:11:59 <sandro> Arnaud: The spec says 4xx, then the best practice says 400 is typically besty
14:12:38 <sandro> JohnArwe: The real trick is if you decide you want to constrain the servers behavior, how to do avoid unnatural acts.
14:12:58 <sandro> JohnArwe: usually you want the constraint to only be active when you're rejecting a message for this particular reason ONLY
14:13:16 <sandro> JohnArwe: What if the message has three different problems?     Let the server pick one.
14:13:58 <sandro> sandro: So best practice -- not a constraint -- but "here's a way to think about this problem"
14:14:02 <Arnaud> PROPOSED: close issue-98 adding to the BP&G doc option 1
14:14:21 <sandro> issue-98?
14:14:21 <trackbot> issue-98 -- HTTP status code for application specific errors -- raised
14:14:21 <trackbot> http://www.w3.org/2012/ldp/track/issues/98
14:14:21 <sandro> ISSUE-98?
14:14:21 <trackbot> ISSUE-98 -- HTTP status code for application specific errors -- raised
14:14:22 <trackbot> http://www.w3.org/2012/ldp/track/issues/98
14:14:30 <Arnaud> PROPOSED: close issue-98 adding to the BP&G doc option 1, as described in issue-98
14:14:44 <JohnArwe> i.e. 400 + rel=describedby
14:14:52 <sandro> " 1. 400 + rel=describedby"
14:14:57 <Arnaud> PROPOSED: close issue-98 adding to the BP&G doc option 1, as described in issue-98 ( i.e. 400 + rel=describedby)
14:15:10 <sandro> +1
14:15:11 <TallTed> +1
14:15:12 <pchampin> +1
14:15:13 <deiu> +1
14:15:14 <BartvanLeeuwen> +1
14:15:15 <nmihindu> +1
14:15:25 <Ashok> +1
14:15:26 <bblfish> +0 Did not really follow this issue
14:15:33 <JohnArwe> +1
14:15:38 <Arnaud> RESOLVED: close issue-98 adding to the BP&G doc option 1, as described in issue-98 ( i.e. 400 + rel=describedby)
14:15:46 <sandro> close ISSUE-98
14:15:46 <trackbot> Closed ISSUE-98.
14:16:03 <sandro> topic: Status of spec
14:16:21 <JohnArwe> I'm going to change the "product" field on issue-98 to match the resolution (from LDP spec to BP&G)
14:16:31 <sandro> Arnaud: commenter was happy about removal of named graphs, of course
14:17:02 <sandro> .. on relative URIs for posting, he's not happy says we didn't given bnode approach proper consideration
14:17:33 <sandro> Arnaud: So I think we'd best make sure we have that topic properly discussed.
14:17:47 <Arnaud> http://lists.w3.org/Archives/Public/public-ldp-wg/2013Mar/0077.html
14:18:01 <sandro> Arnaud: That's the proposal from way back when
14:18:47 <sandro> q+
14:19:23 <Arnaud> ack sandro
14:19:55 <sandro> sandro: How does this work?
14:20:29 <JohnArwe> ... sandro asks: if the input has 2 blank nodes, how does server know which one should be used?
14:20:45 <sandro> pchampin: The blank node that's the placeholder is highlighted by a triple.   It works pretty well.   You can create several resources at the same time
14:20:57 <sandro> .. I however accept the group deciding otherwise.
14:21:37 <sandro> pchampin: So I'm fine with the relative URI solution
14:21:45 <JohnArwe> How does the server know to create >1 resource?  (fuzzy on this still)  would it be possible for the server to create an HTTP resource (assign a URL) that contains all of the posted bnodes?
14:22:34 <JohnArwe> sandro: this is about ensuring your proposal was understood
14:22:39 <sandro> pchampin: I felt like my proposal was understood, but considered inelegant
14:22:59 <sandro> pchampin: ... in part because some people don't like blank nodes
14:23:08 <TallTed> Zakim, unmute me
14:23:08 <Zakim> TallTed should no longer be muted
14:23:39 <sandro> Arnaud: procedurally, lets clearly minute that we've discussed it now and in the past, so Reto knows we didn't miss it.
14:24:50 <sandro> q+
14:25:27 <sandro> JohnArwe: How does the server know it should create multiple resources?   And which triples goes into each resource?
14:25:45 <sandro> JohnArwe: Why would the server create three instead of one with three blank nodes
14:26:03 <TallTed> Zakim, mute me
14:26:03 <Zakim> TallTed should now be muted
14:26:34 <sandro> JohnArwe: So far, all our notions of POST is that it creates one resource.
14:27:11 <sandro> q+
14:27:19 <Arnaud> ack sandro
14:27:48 <sandro> something:    <container> ldp:contains [ ... ]
14:28:01 <bblfish> The discussion is going on because pchampin said that the bnode proposal he put forward allowed a POST to create more than one resource...
14:28:27 <pchampin> q+
14:29:08 <Arnaud> ack pchampin
14:29:36 <sandro> something:    <container> ldp:contains [ ... ], [ ... ].   <a> <b> <c>.                 BUT which one would the <a> <b> <c> triple go into?
14:30:11 <sandro> pchampin: I can still have my own kind of ldp-ish container that accepts posts like this.    So I don't mind going a separate way on this.
14:30:19 <sandro> pchampin: I don't want to push this back again
14:30:53 <bblfish> q+
14:31:07 <sandro> pchampin: I would like us to be clear that using the null relative URI is not technically an RDF graph, so it requires some fiddling in some frameworks.
14:31:19 <Arnaud> ack bblfish
14:31:20 <sandro> Arnaud: Yes, the BP is to include instructions for how to do this in various tools.
14:31:43 <sandro> s/something:/sandro:/
14:32:26 <sandro> bblfish: It depends how the specs are worded.     Is it a "document" that's being passed or a "graph"?   It's not a graph, yes.     So one needs to look closely at the spec.
14:33:00 <sandro> bblfish: maybe pchampin or Reto could find the actual bit in the spec that's wrong
14:33:12 <sandro> Arnaud: That was analyzed at the f2f.
14:33:27 <sandro> Arnaud: I don't hear anyone saying they prefer the bnode approach
14:33:48 <sandro> Arnaud: It seems like we have given this proper consideration.    Everyeone agree?
14:34:33 <sandro> ( also, I don't think we want the <container> ldp:contains [ ... ]  to have to be in the posted content. )
14:34:56 <sandro> topic: Going to CR
14:35:07 <deiu> deiu has joined #ldp
14:35:12 <sandro> Arnaud: We have a spec ready....   Editors?
14:35:21 <sandro> JohnArwe: It was ready last week
14:35:33 <sandro> Arnaud: We have a test suite now, too.
14:35:40 <Zakim> -deiu
14:35:42 <sandro> Arnaud: We've disposed of comments
14:35:56 <sandro> Arnaud: I think we're set to go to CR.
14:36:55 <sandro> sandro: Does the test suite show approved-ness.
14:37:03 <pchampin> @bblfish: I'm not arguing that a precise sentence is wrong in the spec; my point is that the handling of relative URIs in POST is preventing you to work *purely* at the abstract syntax level (which does *not* allow relative URIs), and this should be stated explicitly. Indeed, to some readers (including me and, apparently, Reto), talking about RDF (vs. RDF/XML or Turtle) amounts to talking about the *abstract* syntax.
14:37:29 <Zakim> +??P2
14:37:33 <deiu> Zakim, ??P2 is me
14:37:33 <Zakim> +deiu; got it
14:37:37 <deiu> Zakim, mute me please
14:37:37 <Zakim> deiu should now be muted
14:37:40 <sandro> sandro: It's a bit iffy to issue a call for implementation until we have a set of APPROVED tests.
14:39:08 <sandro> PROPOSAL: Request publication of "ldp" (main spec) as CR
14:39:20 <JohnArwe> +1
14:39:25 <BartvanLeeuwen> +1
14:39:33 <JohnArwe> +1 (for steve speicher, per his email)
14:39:33 <TallTed> +1
14:39:44 <deiu> +1
14:39:50 <sandro> +1  (but maybe we can get test approved before it actually goes out?)
14:40:54 <nmihindu> +1
14:40:55 <Ashok> +1
14:41:03 <sandro> RESOLVED: Request publication of "ldp" (main spec) as CR
14:41:10 <JohnArwe> note that steve is out this week, and I've never done the "capture spec for WD" editor dance.  He was intending to do that when he returns next monday.
14:41:26 <sandro> topic: Test Suite
14:41:53 <sandro> Arnaud: we have the github repo, we got IBM approval to contribute the IBM tests, Sergio has ball adding Marmotta tests
14:42:20 <sandro> Arnaud: So it would be good to clarity about official test suite
14:43:16 <sandro> sandro: I'd suggest a simple machine-readable file listing each approved test (or each test a flag whether it's approved) and a link to where in the minutes it was approved.
14:43:30 <sandro> Arnaud: Maybe we can have that for test week
14:44:48 <sandro> sandro: We could do it manually or automatically
14:45:26 <sandro> Arnaud: So let's have a PROPOSED batch to add, run them, and ACCEPT them as a batch.
14:45:51 <sandro> Arnaud: it becomes everyone's responsibility to run the test, and complain if there's a problem.
14:46:12 <sandro> sandro: That's fine if we have enough people paying attention, willing to run the tests
14:46:40 <sandro> Arnaud: Steve was working to make sure all of Raoul's tests were included
14:46:45 <JohnArwe> Thinking of the test suite as a document, with editors and readers (in this case, machine-based readers) to catch and raise issues sounds like a good form of process re-use.
14:46:55 <Arnaud> q?
14:47:30 <sandro> topic: Other Documents
14:47:43 <sandro> Arnaud: Are the other documents ready for review...???
14:48:02 <sandro> Arnaud: They were supposed to be ready last week!
14:48:11 <nmihindu> Primer is ready to be reviewed by the WG (not perfect) and we will send the email to the group today or tomorrow
14:48:25 <sandro> Arnaud: Access Control Requirements, Best Practice, Primer
14:48:54 <JohnArwe> based on commits from roger, he worked on it a week ago and then again on friday
14:48:54 <bblfish> I am looking at the thread on Access Control, I think it needs some discussion
14:48:55 <nmihindu> About the primer, Roger and I will do a small review ourselves and will send the email to the group asap.
14:49:00 <bblfish> q+
14:49:14 <sandro> Arnaud: Ashok and Henry agreed to review the Primer
14:49:17 <Arnaud> ack bblfish
14:49:25 <sandro> bblfish: Okay
14:49:50 <sandro> bblfish: On access control, perhaps we should get together (Sandro?  Ashok?)  and go through that and be done?
14:50:30 <sandro> Arnaud: Primer sounds like it's ready for review.    Henry, Ashok, can you do that and tell us before Monday if it's okay to go?    Everyone else is welcome to do that too.
14:50:37 <sandro> s/welcome/encouraged/
14:51:12 <sandro> Arnaud: NEXT MONDAY i'm expecting us to resolve to publish Primer, unless someone raises a problem before then.
14:51:41 <sandro> Arnaud: Henry is suggesting side meeting on ACR
14:52:01 <sandro> Ashok: I'm fine with answering email or side meeting
14:52:29 <bblfish> bblfish skype
14:53:00 <sandro> Arnaud: Ashok do you have more work you're planning to do, or are you ready for reviews?
14:53:07 <sandro> Ashok: I want to answer Henry's email
14:53:19 <sandro> Ashok: I *think* we agree, except on one issue
14:53:37 <sandro> Ashok: should just be a couple of paragraphs
14:53:43 <TallTed> Zakim, unmute me
14:53:43 <Zakim> TallTed should no longer be muted
14:53:54 <sandro> Arnaud: Can Tedd review it for next week
14:54:02 <sandro> s/Tedd/Ted/
14:54:27 <Arnaud> http://www.w3.org/2012/ldp/wiki/AccessControlTake2
14:55:28 <sandro> sandro: convert to respec?
14:55:38 <sandro> arnaud: soon-ish, not yet
14:55:43 <sandro> Arnaud: BP?
14:56:13 <sandro> sandro: I pointed out problems with relative uri
14:56:24 <sandro> topic: Paging?
14:57:04 <sandro> Ashok: The problem with paging is that we have a number of ideas, but we haven't got one answer that's perfect.
14:57:18 <sandro> Ashok: That's how it is.
14:57:28 <sandro> Ashok: I don't know how to do better than that.
14:57:39 <sandro> Arnaud: I think that's okay and understood.  Is what we have useful enough>
14:58:18 <sandro> Arnaud: Maybe it'll be easier when we have an updated draft.
14:58:51 <sandro> Arnaud: my take is we can boil the ocean, or just focus on the solution in the spec and see if it's useful.
14:59:49 <sandro> Arnaud: Trying to find sweet spot of something that can be implemented at reasonable cost.
15:01:00 <sandro> Ashok: When we spoke about this, we said "Okay, you're going to be paging, and ...    you can find out if it has changed, and if so, you can start again"   Turns out to be hard to find out of collection has changed.  etags on collection are kind of difficult.
15:01:24 <JohnArwe> ashok, what is difficult about etags on collections?  is that an implementation issue or something general?
15:01:29 <sandro> Arnaud: Interesting.   We're out of time, more later.
15:01:45 <Zakim> -Ashok_Malhotra
15:01:50 <bblfish> bye
15:01:50 <Zakim> -Sandro
15:01:51 <Zakim> -BartvanLeeuwen
15:01:53 <BartvanLeeuwen> bye
15:01:53 <Zakim> -JohnArwe
15:01:55 <Zakim> -Arnaud
15:01:56 <Zakim> -deiu
15:01:57 <Zakim> -bblfish
15:01:58 <Zakim> -nmihindu
15:01:59 <Zakim> -TallTed
15:02:09 <Zakim> -pchampin
15:02:10 <Zakim> SW_LDP()10:00AM has ended
15:02:10 <Zakim> Attendees were BartvanLeeuwen, Arnaud, Ashok_Malhotra, JohnArwe, Sandro, pchampin, deiu, bblfish, TallTed, nmihindu
15:02:16 <JohnArwe> ...I can think of some specific cases where a perfect etag would be hard, but a good-enough(?) might not be.
15:03:46 <ericP> JohnArwe, I think Apache does etags over directories
15:04:05 <ericP> Yves, you spec'd that, didn't you?
15:05:00 <TallTed> I'm thinking some of the eTag difficulty may be implementation-specific... so such might not be required for simple-compliance feature, but I *think* we already said it wouldn't be, anyway
15:05:02 <sandro> conceptually I cant see the problem
15:05:21 <TallTed> s/simple-compliance feature/simple-compliance/
15:23:56 <deiu> deiu has joined #ldp
15:27:57 <stevebattle1118> stevebattle1118 has joined #ldp
15:45:58 <jmvanel> jmvanel has joined #ldp
15:49:21 <bblfish> bblfish has joined #ldp
16:54:38 <jmvanel> jmvanel has joined #ldp
17:22:57 <Zakim> Zakim has left #ldp