IRC log of ldp on 2013-11-04

Timestamps are in UTC.

14:59:02 [RRSAgent]
RRSAgent has joined #ldp
14:59:02 [RRSAgent]
logging to http://www.w3.org/2013/11/04-ldp-irc
14:59:04 [trackbot]
RRSAgent, make logs public
14:59:04 [Zakim]
Zakim has joined #ldp
14:59:06 [trackbot]
Zakim, this will be LDP
14:59:07 [trackbot]
Meeting: Linked Data Platform (LDP) Working Group Teleconference
14:59:07 [trackbot]
Date: 04 November 2013
14:59:08 [Zakim]
ok, trackbot; I see SW_LDP()10:00AM scheduled to start in 1 minute
14:59:59 [Zakim]
SW_LDP()10:00AM has now started
15:00:06 [Ashok]
Ashok has joined #ldp
15:00:06 [Zakim]
+Arnaud
15:00:33 [Zakim]
+??P1
15:00:41 [Zakim]
+Roger
15:00:49 [stevebattle7]
zakim, ??P1 is me
15:00:49 [Zakim]
+stevebattle7; got it
15:01:00 [Zakim]
+Ashok_Malhotra
15:01:20 [Zakim]
+[IBM]
15:01:28 [roger]
roger has joined #ldp
15:01:34 [SteveS]
zakim, [IBM] is SteveS
15:01:34 [Zakim]
+SteveS; got it
15:01:51 [Zakim]
+OpenLink_Software
15:01:59 [TallTed]
Zakim, OpenLink_Software is temporarily me
15:02:01 [Zakim]
+TallTed; got it
15:02:04 [JohnArwe]
JohnArwe has joined #ldp
15:02:11 [Zakim]
+JohnArwe
15:02:25 [TallTed]
Zakim, mute me
15:02:25 [Zakim]
TallTed should now be muted
15:05:07 [TallTed]
Zakim, unmute me
15:05:07 [Zakim]
TallTed should no longer be muted
15:05:13 [TallTed]
Zakim, mute me
15:05:13 [Zakim]
TallTed should now be muted
15:05:28 [JohnArwe]
scribe: JohnArwe
15:05:38 [JohnArwe]
regrets: Bart
15:06:03 [JohnArwe]
chair: Where have all the people gone?
15:06:18 [JohnArwe]
chair: arnaud
15:06:36 [JohnArwe]
topic: last meeting's minutes
15:06:47 [JohnArwe]
ashok read them
15:06:58 [JohnArwe]
resolution: minutes of oct 28 approved
15:07:20 [JohnArwe]
next meeting: Mon Nov 11
15:08:48 [JohnArwe]
F2F penciled in for Jan 14-16, NOT confirmed. were going to see when LC2 published and schedule for after the comment period closes. possible to shift 1 week if you have conflict, ashok. ashok says following week works for him.
15:09:02 [SteveS]
Nov 11th is Veteran's Day US and Remembrance Day CA, though usually these are still work days
15:09:05 [JohnArwe]
Probably want to shift that 1 week later.
15:09:28 [JohnArwe]
topic: issues/actions
15:09:40 [JohnArwe]
none pending review
15:10:30 [JohnArwe]
action-83?
15:10:30 [trackbot]
action-83 -- Roger Menday to Ensure ISSUE-62 is addressed in Primer or Best Practices & Guidelines doc -- due 2013-06-27 -- OPEN
15:10:30 [trackbot]
http://www.w3.org/2012/ldp/track/actions/83
15:11:05 [JohnArwe]
action-104?
15:11:05 [trackbot]
action-104 -- Roger Menday to Review the Use Cases section of the document -- due 2013-03-21 -- OPEN
15:11:05 [trackbot]
http://www.w3.org/2012/ldp/track/actions/104
15:11:29 [TallTed]
Zakim, unmute me
15:11:29 [Zakim]
TallTed should no longer be muted
15:11:30 [JohnArwe]
action-77?
15:11:30 [trackbot]
action-77 -- Ted Thibodeau to Review and comment the WG Access Control draft at http://www.w3.org/2012/ldp/wiki/AccessControl -- due 2013-07-18 -- OPEN
15:11:30 [trackbot]
http://www.w3.org/2012/ldp/track/actions/77
15:11:53 [stevebattle7]
Steve reminds Roger about action 104 :)
15:12:33 [TallTed]
Zakim, mute me
15:12:33 [Zakim]
TallTed should now be muted
15:13:00 [JohnArwe]
topic: Proposals regarding Paging & 209 vs 200
15:14:54 [JohnArwe]
Arnaud: TimBL comment - did not like 303 for server-initiated paging, avoid extra round trip, create new status code for that (we ended up on 209). David Wood polled Mark Baker, who suggested using 200. Arwe discussed with Erik Wilde, who said same thing. While it seemed unnatural to some of us at first, the REST/HTTP expert feedback is that 200 is fine.
15:15:43 [JohnArwe]
...TimBL interested in doing this for all, not just LDP. LDP could take advantage of it. Defining 209 now would make LDP dependent on getting an RFC done.
15:16:14 [JohnArwe]
Proposal: Eliminate 303 and indicate that a client can learn it received a Page based on the existence of Link rel="next/prev" headers.
15:16:47 [JohnArwe]
Arwe: FWIW a client would also see a link header with type=ldp:Page (this is already in LDP, has been for months)
15:18:06 [JohnArwe]
Arnaud: take advantage of the new code in LDP once it's defined, but for now use 200 and pursue the new code separately.
15:18:23 [SteveS]
+1
15:18:27 [roger]
+1
15:18:40 [Ashok]
+1
15:18:42 [JohnArwe]
+1
15:18:44 [stevebattle7]
+1
15:19:01 [TallTed]
+1
15:19:10 [TallTed]
Zakim, who's here?
15:19:10 [Zakim]
On the phone I see Arnaud, stevebattle7, Roger, Ashok_Malhotra, SteveS, TallTed (muted), JohnArwe
15:19:12 [Zakim]
On IRC I see JohnArwe, roger, Ashok, Zakim, RRSAgent, stevebattle7, SteveS, TallTed, bhyland, jmvanel, Arnaud, davidwood, trackbot, Yves, thschee, sandro, ericP
15:19:23 [Arnaud]
Resolved: Eliminate 303 and indicate that a client can learn it received a Page based on the existence of Link rel="next/prev" headers.
15:19:28 [JohnArwe]
Resolved: Eliminate 303 and indicate that a client can learn it received a Page based on the existence of Link rel="next/prev" headers.
15:19:31 [sandro]
+1 (based on what I see on IRC)
15:19:52 [JohnArwe]
proposal: Launch an effort to define 209 as a separate IETF RFC that applies in general to 303s and that we can use in LDPnext
15:20:12 [SteveS]
q+
15:20:31 [Arnaud]
ack steveS
15:21:15 [Arnaud]
Proposed: Launch an effort to define 209 as a separate IETF RFC that applies in general to 303s and that we can use in LDPnext
15:21:47 [stevebattle7]
0
15:22:19 [TallTed]
+0.1
15:22:24 [SteveS]
+1
15:22:29 [Zakim]
+EricP
15:22:29 [JohnArwe]
+0
15:22:48 [JohnArwe]
(good for the community, but not strictly needed for LDP)
15:23:15 [Zakim]
-EricP
15:23:21 [Zakim]
+EricP
15:26:27 [TallTed]
Zakim, unmute me
15:26:27 [Zakim]
TallTed should no longer be muted
15:27:17 [roger_]
roger_ has joined #ldp
15:27:49 [JohnArwe]
EricP: why not make 209 at risk, and then remove at CR/PR transition if no one has agreed to do it?
15:28:02 [roger_]
roger_ has joined #ldp
15:28:29 [JohnArwe]
...discussion, several expressing preferences against it.
15:28:48 [JohnArwe]
Ted asks EricP for concrete proposal.
15:29:43 [JohnArwe]
EricP: it's 2 chars, 200 vs 209. could also say in spec that server can use either, and mark just the 209 at risk
15:29:56 [JohnArwe]
... trying to avoid future objections late in the process
15:30:25 [JohnArwe]
Arnaud: want to set expectations realistically
15:31:06 [JohnArwe]
Insufficient consensus to call 10:21 proposal resolved.
15:32:01 [JohnArwe]
Ashok: what harm in starting effort to add 209? Arnaud: down-side is the effort.
15:32:39 [JohnArwe]
Arnaud: if the state is that everybody's happy to support it if Someone Else defines it, it won't happen.
15:34:08 [JohnArwe]
Ashok: would Erik Wilde do it? Arwe: he told me he could draft 303-replacement text *if the wg wanted him to*, but he could not draft 200-replacement text on his own.
15:34:14 [JohnArwe]
Net: unresolved
15:34:20 [JohnArwe]
topic: put create
15:36:27 [JohnArwe]
Arnaud: TimBL wanted a MUST, then upon further discussion he ack'd it would not work in general and that snowballed into how to advertise the URI space where clients can expect PUT-create to work. We don't have enough experience to solve this problem; in ideal world we'd take the time, figure it out, and spec it, but the WG has a timeline in the charter and we cannot take the time in this version. I'm saying
15:36:27 [JohnArwe]
let's move on and we can improve it in next version.
15:36:42 [JohnArwe]
proposed: Leave spec unchanged - "servers MAY choose to allow the creation of new resources using HTTP PUT" - and defer how servers can advertise this to post LDP 1.0 until we get more feedback on best practices.
15:36:48 [stevebattle7]
+1
15:36:50 [SteveS]
+1
15:36:51 [roger_]
+1
15:36:55 [ericP]
+0
15:37:00 [JohnArwe]
+1
15:37:10 [TallTed]
+1
15:37:33 [JohnArwe]
resolved: Leave spec unchanged - "servers MAY choose to allow the creation of new resources using HTTP PUT" - and defer how servers can advertise this to post LDP 1.0 until we get more feedback on best practices.
15:37:49 [JohnArwe]
topic: Proposal regarding ISSUE-81 Part I: ldp:container
15:38:53 [JohnArwe]
Arnaud: last week we noted that proposal uses both ldp:container and ldp:Container => confusing. we could not get coherent proposals on the call, so I created one.
15:39:12 [JohnArwe]
proposed: Change ldp:container to ldp:containingResource.
15:39:38 [JohnArwe]
... or propose alternative now if you have better one. This seems good enough to me.
15:40:07 [ericP]
+1
15:40:08 [stevebattle7]
q+
15:40:12 [SteveS]
+1
15:40:14 [Arnaud]
ack stevebattle7
15:40:19 [JohnArwe]
ericp: usually when this case occurs, the predicate's range is the like-named class. that is not true here, so will be more confusing than normal.
15:40:44 [JohnArwe]
steveb: like where you're going; how about ...erResource instead of ...ing
15:41:05 [JohnArwe]
for me, either is fine
15:41:07 [roger_]
+1 to ...er
15:41:20 [SteveS]
I prefer ..ing
15:41:22 [JohnArwe]
proposed: Change ldp:container to ldp:containerResource.
15:41:35 [ericP]
+1
15:42:15 [JohnArwe]
SteveS: "ing" reads better in a sentence. SteveB: reads like "resource it contains" too.
15:42:24 [stevebattle7]
+1
15:42:33 [JohnArwe]
+1
15:42:43 [Ashok]
+0.5
15:43:06 [roger_]
+0.5
15:43:16 [TallTed]
+0.5 containerResource because confusion persists
15:43:24 [SteveS]
+0.4
15:43:57 [JohnArwe]
Resolved: Change ldp:container to ldp:containerResource.
15:44:17 [JohnArwe]
Still open to future proposals if someone is struck on the head by an apple.
15:44:32 [JohnArwe]
topic: Proposal regarding ISSUE-81 Part II: Keeping the simple case simple
15:44:49 [JohnArwe]
Arnaud: wishing Henry was here, he has expressed strong opinions on this in past.
15:46:03 [JohnArwe]
...SteveS suggested we address this after renaming settled. With all the new predicates, the simple case now looks "not so simple". We made things mandatory rather than introducing non-monotonic behavior, but it's pretty ugly.
15:46:16 [JohnArwe]
Proposed: Make ldp:insertContentRelation optional, default is ldp:MembershipSubject
15:48:11 [JohnArwe]
SteveS: this affects how membership triples are defined when members are added. a client would need to fetch it from non-member properties of the container in order to use it; no monotonicity issue there. If admin changes things, expectation is that server would do the right thing and keep the representations coherent.
15:49:06 [JohnArwe]
...don't think it making it optional introduces any new issues. In my usage for example, I never need it.
15:49:30 [stevebattle7]
q+
15:49:30 [JohnArwe]
Amended proposal (fix pred name): Make ldp:insertedContentRelation optional, default is ldp:MembershipSubject
15:50:29 [Arnaud]
ack stevebattle
15:51:00 [SteveS]
Note the name is ldp:MemberSubject and don't think we are suggesting a new one here
15:51:03 [JohnArwe]
Ericp: if A sends to B this predicate, and intermediary strips it out, does B do the right thing? that's the easy monotonicity test.
15:51:49 [JohnArwe]
SteveB: go whole hog and fix it; cannot change it. get rid of it altogether.
15:52:47 [JohnArwe]
Arnaud: zaza the cat example wants the member URI to be zaza's, not the uri of the document describing her. that's where it came from.
15:53:12 [JohnArwe]
...Roger was a big motivator for this case.
15:56:04 [JohnArwe]
EricP: working through his A/B case, decides this is not an issue.
15:57:30 [JohnArwe]
Arnaud: let's vote; if need more time, feel free to -1 and ask for a week
15:57:36 [Arnaud]
Proposed: Make ldp:insertedContentRelation optional, default is ldp:MembershipSubject
15:57:50 [JohnArwe]
Proposed: Make ldp:insertContentRelation optional, default is ldp:MemberSubject
15:58:06 [Arnaud]
Proposed: Make ldp:insertedContentRelation optional, default is ldp:MemberSubject
15:58:24 [SteveS]
+1
15:58:32 [JohnArwe]
third proposal above is the one with all the corrections
15:58:41 [JohnArwe]
+1
15:58:45 [ericP]
+0
15:58:57 [TallTed]
+0
15:59:03 [Ashok]
+0
15:59:07 [stevebattle7]
-1 : I'm not comfortable (I would still ditch the option completely)
15:59:26 [roger_]
+1 I believe all the monotonicity experts
16:00:08 [stevebattle7]
I'll try for next week then
16:00:21 [JohnArwe]
Arnaud: Steve B, please explain to people your position
16:01:02 [JohnArwe]
Will hold ISSUE-81 Part I bis: ldp:membershipRule until next time
16:01:28 [JohnArwe]
Arnaud: encourage editors to contact commenters with proposed resolutions to be sure they're ok with the changes
16:01:28 [Zakim]
-Ashok_Malhotra
16:01:41 [JohnArwe]
run away!
16:02:10 [JohnArwe]
ericp: how many people expect to submit implementations for CR exit?
16:02:21 [JohnArwe]
min: 2 impls supporting every feature.
16:02:36 [JohnArwe]
more convincing: large # of commercial and academic impls
16:04:57 [JohnArwe]
adjourned
16:05:07 [Zakim]
-stevebattle7
16:05:23 [Zakim]
-SteveS
16:08:51 [stevebattle7]
stevebattle7 has joined #ldp
16:09:57 [SteveS]
Maybe we put on agenda next week to get an update on http://www.w3.org/wiki/LDP_Implementations
16:12:10 [Zakim]
-JohnArwe
16:40:54 [Zakim]
-Roger
16:40:55 [Zakim]
-TallTed
16:40:55 [Zakim]
-Arnaud
16:40:57 [Zakim]
-EricP
16:40:57 [Zakim]
SW_LDP()10:00AM has ended
16:40:57 [Zakim]
Attendees were Arnaud, Roger, stevebattle7, Ashok_Malhotra, SteveS, TallTed, JohnArwe, EricP
18:03:30 [bhyland]
bhyland has joined #ldp
18:04:34 [stevebattle7]
stevebattle7 has joined #ldp
18:20:46 [Zakim]
Zakim has left #ldp
18:58:12 [deiu]
deiu has joined #ldp
20:07:10 [SteveS_]
SteveS_ has joined #ldp
20:47:20 [bhyland]
bhyland has joined #ldp
20:53:33 [TallTed]
TallTed has joined #ldp
21:00:42 [SteveS]
SteveS has joined #ldp