edit

Linked Data Platform (LDP) Working Group Teleconference

Minutes of 16 December 2013

Agenda
http://www.w3.org/2012/ldp/wiki/Meetings:Telecon2013.12.16
Seen
Alexandre Bertails, Arnaud Le Hors, Ashok Malhotra, Cody Burleson, Eric Prud'hommeaux, Henry Story, John Arwe, Nandana Mihindukulasooriya, Roger Menday, Steve Speicher
Chair
Arnaud Le Hors
Scribe
Cody Burleson
IRC Log
Original
Resolutions
  1. Minutes of Dec 9 approved. link
  2. close ISSUE-91, by adding that for an LDPC the link header is: Link: <http://www.w3.org/ns/ldp#Container>; rel="type" link
  3. define hierarchy of resources as proposed by Alexandre Issue-90 Proposal 1 link
  4. add that an LDPG/LDPC is a Named Graph as proposed by Alexandre Issue-90 Proposal 2 amended to cover PATCH-create and remove hashless link
Topics
14:57:05 <RRSAgent> logging to http://www.w3.org/2013/12/16-ldp-irc

RRSAgent IRC Bot: logging to http://www.w3.org/2013/12/16-ldp-irc

14:57:07 <trackbot> RRSAgent, make logs public

Trackbot IRC Bot: RRSAgent, make logs public

14:57:09 <trackbot> Zakim, this will be LDP

Trackbot IRC Bot: Zakim, this will be LDP

14:57:09 <Zakim> ok, trackbot; I see SW_LDP()10:00AM scheduled to start in 3 minutes

Zakim IRC Bot: ok, trackbot; I see SW_LDP()10:00AM scheduled to start in 3 minutes

14:57:10 <trackbot> Meeting: Linked Data Platform (LDP) Working Group Teleconference
14:57:10 <trackbot> Date: 16 December 2013
15:00:35 <Zakim> SW_LDP()10:00AM has now started

Zakim IRC Bot: SW_LDP()10:00AM has now started

15:00:43 <Zakim> +Arnaud

Zakim IRC Bot: +Arnaud

15:01:32 <Zakim> +[IBM]

Zakim IRC Bot: +[IBM]

15:01:45 <SteveS> Zakim, [IBM] is me

Steve Speicher: Zakim, [IBM] is me

15:01:45 <Zakim> +SteveS; got it

Zakim IRC Bot: +SteveS; got it

15:01:59 <Zakim> +Roger

Zakim IRC Bot: +Roger

15:02:02 <Zakim> +Alexandre

Zakim IRC Bot: +Alexandre

15:02:14 <Zakim> +Ashok_Malhotra

Zakim IRC Bot: +Ashok_Malhotra

15:03:31 <Zakim> +bblfish

Zakim IRC Bot: +bblfish

15:03:49 <Zakim> +JohnArwe

Zakim IRC Bot: +JohnArwe

15:03:49 <bblfish_> hi

Henry Story: hi

15:05:02 <Arnaud> zakim. who's on the phone?

Arnaud Le Hors: zakim. who's on the phone?

15:05:14 <Arnaud> zakim, who's on the phone?

Arnaud Le Hors: zakim, who's on the phone?

15:05:14 <Zakim> On the phone I see Arnaud, SteveS, Roger, Alexandre, Ashok_Malhotra, bblfish, JohnArwe

Zakim IRC Bot: On the phone I see Arnaud, SteveS, Roger, Alexandre, Ashok_Malhotra, bblfish, JohnArwe

15:05:43 <Zakim> +[IPcaller]

Zakim IRC Bot: +[IPcaller]

15:06:02 <codyburleson> Zakim, IPCaller is me.

Cody Burleson: Zakim, IPCaller is me.

15:06:02 <Zakim> +codyburleson; got it

Zakim IRC Bot: +codyburleson; got it

15:09:18 <codyburleson> i can scribe, but give me a bit

Cody Burleson: i can scribe, but give me a bit

15:09:27 <codyburleson> trouble with skype

Cody Burleson: trouble with skype

15:09:59 <Zakim> -codyburleson

Zakim IRC Bot: -codyburleson

15:10:44 <Zakim> +[IPcaller]

Zakim IRC Bot: +[IPcaller]

15:11:29 <JohnArwe> Zakim, IPCaller is cody

John Arwe: Zakim, IPCaller is cody

15:11:29 <Zakim> +cody; got it

Zakim IRC Bot: +cody; got it

<codyburleson> scribe: codyburleson

(Scribe set to Cody Burleson)

<codyburleson> chair: Arnaud
<codyburleson> agenda: http://www.w3.org/2012/ldp/wiki/Meetings:Telecon2013.12.16
<codyburleson> topic: Admin

1. Admin

<codyburleson> subtopic: Approval of minutes of December 9

1.1. Approval of minutes of December 9

<codyburleson> http://www.w3.org/2013/meeting/ldp/2013-12-09

http://www.w3.org/2013/meeting/ldp/2013-12-09

15:12:08 <betehess> no objection

Alexandre Bertails: no objection

15:12:35 <codyburleson> RESOLVED: Minutes of Dec 9 approved.

RESOLVED: Minutes of Dec 9 approved.

15:12:51 <codyburleson> subTopic: Next meeting

1.2. Next meeting

15:13:13 <SteveS> I'm unavailable on Dec 23rd

Steve Speicher: I'm unavailable on Dec 23rd

15:13:36 <Zakim> +??P0

Zakim IRC Bot: +??P0

15:14:08 <roger> * i can't make it next week

Roger Menday: * i can't make it next week

15:14:26 <nmihindu> Zakim, ??P0 is me

Nandana Mihindukulasooriya: Zakim, ??P0 is me

15:14:26 <Zakim> +nmihindu; got it

Zakim IRC Bot: +nmihindu; got it

15:14:33 <nmihindu> Zakim, mute me

Nandana Mihindukulasooriya: Zakim, mute me

15:14:33 <Zakim> nmihindu should now be muted

Zakim IRC Bot: nmihindu should now be muted

15:15:00 <nmihindu> I don't think I will make it

Nandana Mihindukulasooriya: I don't think I will make it

15:15:07 <bblfish> I can be here

Henry Story: I can be here

15:16:29 <betehess> JohnArwe, the second one was already LDPG/LDPC, which is the intent

Alexandre Bertails: JohnArwe, the second one was already LDPG/LDPC, which is the intent

15:17:19 <codyburleson> For now, next meeting on Jan 6th, but let's bring it up again at end of call.

For now, next meeting on Jan 6th, but let's bring it up again at end of call.

15:17:30 <codyburleson> Topic: Tracking of Actions

2. Tracking of Actions

15:17:56 <codyburleson> Arnaud: Action-95 and Action-96 can be closed.

Arnaud Le Hors: ACTION-95 and ACTION-96 can be closed.

15:18:09 <Zakim> +EricP

Zakim IRC Bot: +EricP

15:19:12 <codyburleson> Arnaud: We close Action-115 and Action-116 as well.

Arnaud Le Hors: We close ACTION-115 and ACTION-116 as well.

15:19:59 <JohnArwe> ericp talking in staccato

John Arwe: ericp talking in staccato

15:20:09 <codyburleson> ... Action-117 can be closed, right Eric?

... ACTION-117 can be closed, right Eric?

15:21:19 <codyburleson> ... The only thing about action-117 is that we need to follow up with Tim. I sent an email, but got no response yet. So, I need you to keep bugging him. Should we keep 117 open?

... The only thing about ACTION-117 is that we need to follow up with Tim. I sent an email, but got no response yet. So, I need you to keep bugging him. Should we keep 117 open?

15:22:05 <codyburleson> ... We'll leave it open and move on to next topic, which this is about.

... We'll leave it open and move on to next topic, which this is about.

15:22:14 <codyburleson> TOPIC: Paging

3. Paging

15:24:20 <codyburleson> Arnaud: Tim suggesting a new code (like 209) to combine two requests to save client trouble of having to do an extra get. We discussed the possibility, but it did not seem feasible in our schedule. Others suggested sending first page with the initial request and a 200. Essentially, we're losing the capability to have different addresses for the first page and the rest of the collection.

Arnaud Le Hors: Tim suggesting a new code (like 209) to combine two requests to save client trouble of having to do an extra get. We discussed the possibility, but it did not seem feasible in our schedule. Others suggested sending first page with the initial request and a 200. Essentially, we're losing the capability to have different addresses for the first page and the rest of the collection.

15:25:46 <codyburleson> + Does this mean what we have today is unacceptable, or is it only that it is pending that it actually works? I don't think we can give him (Tim) the 209. We either stick with the resolution we have and work on the remaining questions and consequent issues. Or we revert back to what we had and look at 209 for LDP Next.

+ Does this mean what we have today is unacceptable, or is it only that it is pending that it actually works? I don't think we can give him (Tim) the 209. We either stick with the resolution we have and work on the remaining questions and consequent issues. Or we revert back to what we had and look at 209 for LDP Next.

15:26:39 <codyburleson> + I would like to know what Tim would prefer: the redirect for now sow he can really have what he wants in the end?

+ I would like to know what Tim would prefer: the redirect for now sow he can really have what he wants in the end?

15:27:44 <codyburleson> EricP: One other option is carry on current approach, but add another header. We could invent a header that makes a 200 a 209. It's harder to invent new status codes than it is new headers.

Eric Prud'hommeaux: One other option is carry on current approach, but add another header. We could invent a header that makes a 200 a 209. It's harder to invent new status codes than it is new headers.

15:28:55 <betehess> looks like this approach will change the semantics of HTTP 200 significantly

Alexandre Bertails: looks like this approach would change the semantics of HTTP 200 significantly

15:29:11 <betehess> s/will change/would change/
15:29:20 <bblfish> q+

Henry Story: q+

15:30:01 <Arnaud> ack bblfish

Arnaud Le Hors: ack bblfish

15:30:32 <codyburleson> Henry: My guess is 303 would be more helpful (JohnArwe just brought this up).

Henry Story: My guess is 303 would be more helpful (JohnArwe just brought this up).

15:31:12 <SteveS> q+

Steve Speicher: q+

15:31:29 <Arnaud> ack SteveS

Arnaud Le Hors: ack SteveS

15:31:38 <codyburleson> EricP: My suspicion is that the 200 and a header has an easier migration path to 209.

Eric Prud'hommeaux: My suspicion is that the 200 and a header has an easier migration path to 209.

15:32:13 <JohnArwe> @ericp: playing off the new header notion, what if we use link type=collection href=LPDC#allpages (since the collection URI != the page URI, you know they're different even if you cannot directly retrieve the collection as a whole)

John Arwe: @ericp: playing off the new header notion, what if we use link type=collection href=LPDC#allpages (since the collection URI != the page URI, you know they're different even if you cannot directly retrieve the collection as a whole)

15:32:44 <Ashok> q+

Ashok Malhotra: q+

15:32:45 <codyburleson> SteveS: I think we should stick with / revert to the 303, perhaps.

Steve Speicher: I think we should stick with / revert to the 303, perhaps.

15:33:42 <Arnaud> ack Ashok

Arnaud Le Hors: ack Ashok

15:33:47 <SteveS> saying that if we do some stopgap 209 solution, it will only be for LDP server-initiated paging...so 303 is well-known, we can "fix" the general 303 problem with the 209 approach

Steve Speicher: saying that if we do some stopgap 209 solution, it will only be for LDP server-initiated paging...so 303 is well-known, we can "fix" the general 303 problem with the 209 approach

15:33:55 <codyburleson> Ashok: I am wondering - what does ATOM do?

Ashok Malhotra: I am wondering - what does ATOM do?

15:34:02 <codyburleson> ... ATOM Pub

... ATOM Pub

15:34:33 <codyburleson> JohnArwe: I'm pretty sure it doesn't have this concept at all. 5005 is Paging and Archiving

John Arwe: I'm pretty sure it doesn't have this concept at all. 5005 is Paging and Archiving

15:35:51 <JohnArwe> http://tools.ietf.org/html/rfc5023#section-5.2 says To list the Members of a Collection, the client sends a GET request

John Arwe: http://tools.ietf.org/html/rfc5023#section-5.2 says To list the Members of a Collection, the client sends a GET request

15:35:51 <JohnArwe>    to the URI of a Collection.  An Atom Feed Document is returned whose

John Arwe: to the URI of a Collection. An Atom Feed Document is returned whose

15:35:51 <JohnArwe>    Entries contain the IRIs of Member Resources.  The returned Feed may

John Arwe: Entries contain the IRIs of Member Resources. The returned Feed may

15:35:51 <JohnArwe>    describe all, or only a partial list, of the Members in a Collection

John Arwe: describe all, or only a partial list, of the Members in a Collection

15:36:03 <codyburleson> Arnaud: We've been trying to support a model where it is server-initiated paging.

Arnaud Le Hors: We've been trying to support a model where it is server-initiated paging.

15:36:03 <bblfish> q+

Henry Story: q+

15:36:17 <Arnaud> ack bblfish

Arnaud Le Hors: ack bblfish

15:36:17 <JohnArwe> 200 response

John Arwe: 200 response

15:36:23 <codyburleson> ... How does the server communicate "Here is the first page."

... How does the server communicate "Here is the first page."

15:37:24 <ericP> q+ to ask if any existing 303 machinery knows about link rel="next"

Eric Prud'hommeaux: q+ to ask if any existing 303 machinery knows about link rel="next"

15:38:27 <codyburleson> Henry: There is (used to be) something around that 200 where you used to get back the content-location. In HTTP 2, I think they reversed that because people weren't using that correctly in browsers. I think it's kind of close to that. And then there was something in old HTTP that would tell you the base of the page. Perhaps, then there is something not so silly about this. It's just the browser people couldn't make sense of it, it seems.

Henry Story: There is (used to be) something around that 200 where you used to get back the content-location. In HTTP 2, I think they reversed that because people weren't using that correctly in browsers. I think it's kind of close to that. And then there was something in old HTTP that would tell you the base of the page. Perhaps, then there is something not so silly about this. It's just the browser people couldn't make sense of it, it seems.

15:39:01 <SteveS> q+

Steve Speicher: q+

15:39:16 <codyburleson> ... There WAS a content-location and there was a change between HTTP 1.1 and 1.2.

... There WAS a content-location and there was a change between HTTP 1.1 and 1.2.

15:39:24 <Arnaud> ack ericp

Arnaud Le Hors: ack ericp

15:39:24 <Zakim> ericP, you wanted to ask if any existing 303 machinery knows about link rel="next"

Zakim IRC Bot: ericP, you wanted to ask if any existing 303 machinery knows about link rel="next"

15:40:09 <bblfish> Content-Location http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.14

Henry Story: Content-Location http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.14

15:40:14 <codyburleson> EricP: SteveS, you talked about existing machinery that understands 303s. I'm wondering if any of that would know how to deal with the 303, plus the link-rep next?

Eric Prud'hommeaux: SteveS, you talked about existing machinery that understands 303s. I'm wondering if any of that would know how to deal with the 303, plus the link-rep next?

15:40:32 <codyburleson> SteveS: I am aware if it does.

Steve Speicher: I am not aware if it does.

15:40:53 <SteveS> s/I am aware/I am not aware/
15:41:13 <Arnaud> ack SteveS

Arnaud Le Hors: ack SteveS

15:42:10 <codyburleson> SteveS: Didn't we have this whole conversation at one of the F2Fs? We looked at the HTTP specs, and went off and talked to Mark Baker. We came back and said it was unclear, and we'd have to clarify. Maybe we can go back through the minutes.

Steve Speicher: Didn't we have this whole conversation at one of the F2Fs? We looked at the HTTP specs, and went off and talked to Mark Baker. We came back and said it was unclear, and we'd have to clarify. Maybe we can go back through the minutes.

15:42:22 <bblfish> http://tools.ietf.org/html/draft-ietf-httpbis-p2-semantics-25#section-3.1.4.2

Henry Story: http://tools.ietf.org/html/draft-ietf-httpbis-p2-semantics-25#section-3.1.4.2

15:42:44 <codyburleson> Arnaud: You're right.

Arnaud Le Hors: You're right.

15:43:48 <codyburleson> Arnaud: Group would like to work on 303 and save 209 as an optimization for the future?

Arnaud Le Hors: Group would like to work on 303 and save 209 as an optimization for the future?

15:44:15 <codyburleson> EricP: How many would prefer 303 with 200, plus a header?

Eric Prud'hommeaux: How many would prefer 303 with 200, plus a header?

15:44:26 <bblfish> It depends

Henry Story: It depends

15:44:28 <betehess> I definitely prefer 303-now and 209-later over any other solution

Alexandre Bertails: I definitely prefer 303-now and 209-later over any other solution

15:44:46 <bblfish> If the header is really good then it would be cool

Henry Story: If the header is really good then it would be cool

15:44:58 <SteveS> It depends, if 200 is still a bit fuzzy...then I like 303-now and 209-later

Steve Speicher: It depends, if 200 is still a bit fuzzy...then I like 303-now and 209-later

15:45:17 <bblfish> http://tools.ietf.org/html/draft-ietf-httpbis-p2-semantics-25#appendix-B
   The definition of Content-Location has been changed to no longer

Henry Story: http://tools.ietf.org/html/draft-ietf-httpbis-p2-semantics-25#appendix-B
 The definition of Content-Location has been changed to no longer

15:45:17 <bblfish>    affect the base URI for resolving relative URI references, due to

Henry Story: affect the base URI for resolving relative URI references, due to

15:45:17 <bblfish>    poor implementation support and the undesirable effect of potentially

Henry Story: poor implementation support and the undesirable effect of potentially

15:45:18 <codyburleson> EricP: When I talk to Tim, should I suggest that 200, plus a header still on the table?

Eric Prud'hommeaux: When I talk to Tim, should I suggest that 200, plus a header still on the table?

15:45:19 <bblfish>    breaking relative links in content-negotiated resources.

Henry Story: breaking relative links in content-negotiated resources.

15:45:21 <bblfish>    (Section 3.1.4.2)

Henry Story: (Section 3.1.4.2)

15:45:35 <codyburleson> ... Today, we don't have a header to extend the semantics of that 200.

... Today, we don't have a header to extend the semantics of that 200.

15:45:39 <bblfish> q+

Henry Story: q+

15:45:56 <codyburleson> JohnArwe: Well, we have the Next. Which is what we're claiming today extends that 200.

John Arwe: Well, we have the Next. Which is what we're claiming today extends that 200.

15:46:07 <Arnaud> ack bblfish

Arnaud Le Hors: ack bblfish

15:46:50 <SteveS> 303 has clear usage in LD world, using 200+ existing header not so clear...that is why 209 for LD would be very clear

Steve Speicher: 303 has clear usage in LD world, using 200+ existing header not so clear...that is why 209 for LD would be very clear

15:47:08 <codyburleson> Henry: Arnaud: Henry, if you can send the list what you think we can do with Content-Location, that would be helful.

Henry Story: Arnaud: Henry, if you can send the list what you think we can do with Content-Location, that would be helful.

15:48:17 <betehess> I understand how we got there, but I prefer the clean and clear semantics of 303 for now, and take the time to do things right later with a nice 209

Alexandre Bertails: I understand how we got there, but I prefer the clean and clear semantics of 303 for now, and take the time to do things right later with a nice 209

15:49:06 <JohnArwe> I'm struggling to re-find content-location in bis, but what I remember is that bis clarified the usage of content-location (quite a bit IIRC) to essentially say it was used to point to the resource "after" content negotiation. i.e. you could GET R, and if the us-english version of R *also* had a URI (R.en-us), then the response to GET R could contain content-location R.en-us

John Arwe: I'm struggling to re-find content-location in bis, but what I remember is that bis clarified the usage of content-location (quite a bit IIRC) to essentially say it was used to point to the resource "after" content negotiation. i.e. you could GET R, and if the us-english version of R *also* had a URI (R.en-us), then the response to GET R could contain content-location R.en-us

15:49:19 <bblfish> 303 is the easiest to go back to. But presumably the 201 Location-Header shows that there was something that people thought could be done with 201

Henry Story: 303 is the easiest to go back to. But presumably the 201 Location-Header shows that there was something that people thought could be done with 201

15:50:20 <ericP> ACTION: ericP to report to TimBL: some pref for reverting to 303, 200+Header still on the table, Henry considering 200+Location

ACTION: ericP to report to TimBL: some pref for reverting to 303, 200+Header still on the table, Henry considering 200+Location

15:50:20 <trackbot> Created ACTION-118 - Report to timbl: some pref for reverting to 303, 200+header still on the table, henry considering 200+location [on Eric Prud'hommeaux - due 2013-12-23].

Trackbot IRC Bot: Created ACTION-118 - Report to timbl: some pref for reverting to 303, 200+header still on the table, henry considering 200+location [on Eric Prud'hommeaux - due 2013-12-23].

15:50:47 <codyburleson> Arnaud: That's cool. Then close action 117.

Arnaud Le Hors: That's cool. Then close ACTION-117.

15:52:08 <codyburleson> TOPIC: Issue 91

4. ISSUE-91

15:53:22 <JohnArwe> q+

John Arwe: q+

15:54:13 <bblfish> Issue-91?

Henry Story: ISSUE-91?

15:54:13 <trackbot> Issue-91 -- The LDP (REST) interactions must be driven by the rel='type' Link header -- open

Trackbot IRC Bot: ISSUE-91 -- The LDP (REST) interactions must be driven by the rel='type' Link header -- open

15:54:13 <trackbot> http://www.w3.org/2012/ldp/track/issues/91

Trackbot IRC Bot: http://www.w3.org/2012/ldp/track/issues/91

15:54:32 <codyburleson> Arnaud: We added to the spec a header that says 'This is not just any RDF; it's RDF served by an LDP server), so you know what interaction model. And we read the RDF to distinguish between LDPC and LDPR. Alex(?) was saying this is no good. Wants to specify which interaction model in a header.

Arnaud Le Hors: We added to the spec a header that says 'This is not just any RDF; it's RDF served by an LDP server), so you know what interaction model. And we read the RDF to distinguish between LDPC and LDPR. Alex(?) was saying this is no good. Wants to specify which interaction model in a header.

15:54:38 <Arnaud> ack JohnArwe

Arnaud Le Hors: ack JohnArwe

15:54:49 <ericP> q+ to ask if we currently have any semantics around LDPC DELETE

Eric Prud'hommeaux: q+ to ask if we currently have any semantics around LDPC DELETE

15:56:07 <bblfish> q+

Henry Story: q+

15:56:13 <Arnaud> ack ericP

Arnaud Le Hors: ack ericP

15:56:13 <Zakim> ericP, you wanted to ask if we currently have any semantics around LDPC DELETE

Zakim IRC Bot: ericP, you wanted to ask if we currently have any semantics around LDPC DELETE

15:56:15 <codyburleson> JohnArwe: I'm not aware of this kind of specification being supported by the Link registry. It doesn't say anything about interaction model. It's not clear to me that we're using type as appropriate. Maybe we could define our won or re-open the discussion about defining media-type.

John Arwe: I'm not aware of this kind of specification being supported by the Link registry. It doesn't say anything about interaction model. It's not clear to me that we're using type as appropriate. Maybe we could define our won or re-open the discussion about defining media-type.

15:56:31 <ericP> ack me

Eric Prud'hommeaux: ack me

15:56:54 <Arnaud> zakim, mute ashok

Arnaud Le Hors: zakim, mute ashok

15:56:54 <Zakim> Ashok_Malhotra should now be muted

Zakim IRC Bot: Ashok_Malhotra should now be muted

15:57:22 <betehess> JohnArwe, I wrote this proposal in the context of the group deciding in the past not to define a new media-type. I was myself in favor of using a proper media-type some time ago (not true anymore)

Alexandre Bertails: JohnArwe, I wrote this proposal in the context of the group deciding in the past not to define a new media-type. I was myself in favor of using a proper media-type some time ago (not true anymore)

15:57:27 <codyburleson> EricP: Did we have any semantics around deleting containers deleting containees?

Eric Prud'hommeaux: Did we have any semantics around deleting containers deleting containees?

15:58:07 <Arnaud> ack bblfish

Arnaud Le Hors: ack bblfish

15:58:37 <codyburleson> Henry: I think the idea is that you can't delete a Container if you haven't deleted all the members.

Henry Story: I think the idea is that you can't delete a Container if you haven't deleted all the members.

15:59:36 <codyburleson> ... You can group things by interaction model or by all sorts of things. The 'type' is really quite general. It just defines that something is an aliment of a set, I think.

... You can group things by interaction model or by all sorts of things. The 'type' is really quite general. It just defines that something is an aliment of a set, I think.

16:00:09 <codyburleson> JohnArwe: We added it because I need to know if it's an A or a B, not to know what operations are aloud on it.

John Arwe: We added it because I need to know if it's an A or a B, not to know what operations are aloud on it.

16:00:11 <betehess> wasn't the Link header thing a proposal from Erik Wilde?

Alexandre Bertails: wasn't the Link header thing a proposal from Erik Wilde?

16:01:35 <bblfish> q+

Henry Story: q+

16:01:49 <Arnaud> ack bblfish

Arnaud Le Hors: ack bblfish

16:01:54 <codyburleson> SteveS: When I would see link-rel="type", I would have thought that when I crack open the contents of what I'm seeing, I'd see a type triple that was equal to the same.

Steve Speicher: When I would see link-rel="type", I would have thought that when I crack open the contents of what I'm seeing, I'd see a type triple that was equal to the same.

16:02:03 <JohnArwe> Henry's argument could equally be applied to rdf:type "it's just a type", suggesting we don't need Link rel=type at all except as a shortcut to avoid parsing payload.

John Arwe: Henry's argument could equally be applied to rdf:type "it's just a type", suggesting we don't need Link rel=type at all except as a shortcut to avoid parsing payload.

16:02:42 <codyburleson> Henry: If a document says something, it's not the same as the server saying it. The server could be lying, but it really should be the server that defines the interaction.

Henry Story: If a document says something, it's not the same as the server saying it. The server could be lying, but it really should be the server that defines the interaction.

16:03:05 <JohnArwe> to be clear: I *like* the idea of naming the interaction model via a header, so the server asserts it rather than content.

John Arwe: to be clear: I *like* the idea of naming the interaction model via a header, so the server asserts it rather than content.

16:03:07 <Arnaud> PROPOSED: close ISSUE-91, by adding that for an LDPC the link header is: Link: <http://www.w3.org/ns/ldp#Container>; rel="type"

PROPOSED: close ISSUE-91, by adding that for an LDPC the link header is: Link: <http://www.w3.org/ns/ldp#Container>; rel="type"

16:03:23 <bblfish> +1 makes sense to me

Henry Story: +1 makes sense to me

16:03:24 <betehess> +1

Alexandre Bertails: +1

16:03:26 <JohnArwe> it's simply that I think type is the wrong rel= value

John Arwe: it's simply that I think type is the wrong rel= value

16:03:48 <JohnArwe> -0.5 I can hold my nose

John Arwe: -0.5 I can hold my nose

16:03:51 <codyburleson> codyburleson: +0

Cody Burleson: +0

16:03:59 <nmihindu> +1

Nandana Mihindukulasooriya: +1

16:04:10 <SteveS> -0 think rel="type" isn't right

Steve Speicher: -0 think rel="type" isn't right

16:04:10 <betehess> the proposal can always be amended with another rel= value

Alexandre Bertails: the proposal can always be amended with another rel= value

16:04:11 <Ashok> +1

Ashok Malhotra: +1

16:04:13 <roger> 0

Roger Menday: 0

16:04:23 <ericP> +0 (haven't done research)

Eric Prud'hommeaux: +0 (haven't done research)

16:05:15 <Arnaud> RESOLVED: close ISSUE-91, by adding that for an LDPC the link header is: Link: <http://www.w3.org/ns/ldp#Container>; rel="type"

RESOLVED: close ISSUE-91, by adding that for an LDPC the link header is: Link: <http://www.w3.org/ns/ldp#Container>; rel="type"

16:05:22 <betehess>  /me is more than ok to start a discussion on another rel= value for the purpose

Alexandre Bertails: /me is more than ok to start a discussion on another rel= value for the purpose

<codyburleson> TOPIC: Issue-90

5. ISSUE-90

16:05:44 <bblfish> Issue-90?

Henry Story: ISSUE-90?

16:05:44 <trackbot> Issue-90 -- An LDPC/LDPR is a Named Graph -- open

Trackbot IRC Bot: ISSUE-90 -- An LDPC/LDPR is a Named Graph -- open

16:05:44 <trackbot> http://www.w3.org/2012/ldp/track/issues/90

Trackbot IRC Bot: http://www.w3.org/2012/ldp/track/issues/90

16:05:55 <JohnArwe> y that's why I wanted to be clear which aspects are good/bad IMO

John Arwe: y that's why I wanted to be clear which aspects are good/bad IMO

16:06:53 <JohnArwe> q+

John Arwe: q+

16:06:59 <betehess> once again, the names can all be discussed and amended in another proposal

Alexandre Bertails: once again, the names can all be discussed and amended in another proposal

16:06:59 <Arnaud> ack JohnArwe

Arnaud Le Hors: ack JohnArwe

16:07:13 <codyburleson> Arnaud: The proposal was about trying to name things more clearly. Nobody like the LDPG (Graph), but I don't know what to call it. There is also LDP Binary (misnomer?). Let's just see if we agree on defining this kind of hierarchy - then we can change the names later if necessary.

Arnaud Le Hors: The proposal was about trying to name things more clearly. Nobody like the LDPG (Graph), but I don't know what to call it. There is also LDP Binary (misnomer?). Let's just see if we agree on defining this kind of hierarchy - then we can change the names later if necessary.

16:07:31 <bblfish> http://www.w3.org/2012/ldp/wiki/Issue%2D90

Henry Story: http://www.w3.org/2012/ldp/wiki/Issue%2D90

16:08:05 <codyburleson> JohnArwe: something about semantics versus syntax.

John Arwe: something about semantics versus syntax.

16:08:20 <Arnaud> PROPOSED: define hierarchy of resources as proposed by Alexandre Issue-90 Proposal 1

PROPOSED: define hierarchy of resources as proposed by Alexandre ISSUE-90 Proposal 1

16:08:28 <JohnArwe> ...would LDPC     # LDP Container: representation is RDF + ldp:Container as rel="type" Link header be accurately restated as

John Arwe: ...would LDPC # LDP Container: representation is RDF + ldp:Container as rel="type" Link header be accurately restated as

16:08:32 <betehess> +1

Alexandre Bertails: +1

16:08:34 <SteveS> q+

Steve Speicher: q+

16:08:41 <JohnArwe> ... LDPC     # LDP Container: representation is RDF + interaction model of ldp:Container

John Arwe: ... LDPC # LDP Container: representation is RDF + interaction model of ldp:Container

16:08:49 <Arnaud> ack SteveS

Arnaud Le Hors: ack SteveS

16:08:52 <JohnArwe> ...and similar for LDPG with the ldp:Resource interation model

John Arwe: ...and similar for LDPG with the ldp:Resource interation model

16:08:59 <bblfish> +1 I am for good names, it makes it so much easier to discuss things in this group.

Henry Story: +1 I am for good names, it makes it so much easier to discuss things in this group.

16:09:10 <codyburleson> SteveS: An LDPR. Does it represent the class of all types of resources one would find on the Web?

Steve Speicher: An LDPR. Does it represent the class of all types of resources one would find on the Web?

16:09:39 <codyburleson> JohnArwe: Those 'other' resources don't have the LDP interaction model.

John Arwe: Those 'other' resources don't have the LDP interaction model.

16:09:40 <bblfish> +1 for  the interaction model clarity

Henry Story: +1 for the interaction model clarity

16:09:40 <Ashok> q+

Ashok Malhotra: q+

16:09:47 <Arnaud> ack Ashok

Arnaud Le Hors: ack Ashok

16:10:15 <codyburleson> Ashok: Question for Alexandre - What is an example of a Web resource that isn't an LDPR.

Ashok Malhotra: Question for Alexandre - What is an example of a Web resource that isn't an LDPR.

16:10:53 <codyburleson> ... There are things on the Web that you cannot add to a container?

... There are things on the Web that you cannot add to a container?

16:11:18 <codyburleson> Alexandre: Not saying that. Just saying they'd be treated as a binary.

Alexandre Bertails: Not saying that. Just saying they'd be treated as a binary.

16:11:25 <betehess> guys, I have to leave, will be off  for 3-5 minutes, then will join again from my mobile phone, won't be in front of a computer

Alexandre Bertails: guys, I have to leave, will be off for 3-5 minutes, then will join again from my mobile phone, won't be in front of a computer

16:11:43 <Zakim> -Alexandre

Zakim IRC Bot: -Alexandre

16:12:27 <bblfish> q+

Henry Story: q+

16:12:40 <Arnaud> ack bblfish

Arnaud Le Hors: ack bblfish

16:12:43 <codyburleson> JohnArwe: I don't think we said you can't add other Web resources to a container. I guess the question is - what changes in the interaction model do they have to support?

John Arwe: I don't think we said you can't add other Web resources to a container. I guess the question is - what changes in the interaction model do they have to support?

16:13:13 <SteveS> +1 would help to clarify the "not all Web resources are LDPR", to include interaction model details/explanation

Steve Speicher: +1 would help to clarify the "not all Web resources are LDPR", to include interaction model details/explanation

16:13:30 <JohnArwe> steve, that's the first bullet under Remarks in the proposal

John Arwe: steve, that's the first bullet under Remarks in the proposal

16:13:58 <SteveS> JohnArwe, yes that it why I added "to include"

Steve Speicher: JohnArwe, yes that it why I added "to include"

16:14:05 <JohnArwe> maybe "random html web page" would serve as a useful example for ashok's thought exercise

John Arwe: maybe "random html web page" would serve as a useful example for ashok's thought exercise

16:14:54 <codyburleson> Arnaud: Today the spec allows you to have stuff in a container that may include things you may not be able to delete. Things, for example, that the server may not be in control of.

Arnaud Le Hors: Today the spec allows you to have stuff in a container that may include things you may not be able to delete. Things, for example, that the server may not be in control of.

16:16:10 <JohnArwe> can I add a web page to a LDPC?  I think so (from a membership perspective - still wrapping my head around containment), there might be different paths for create/add-of-existing

John Arwe: can I add a web page to a LDPC? I think so (from a membership perspective - still wrapping my head around containment), there might be different paths for create/add-of-existing

16:16:31 <codyburleson> Arnaud: Vote on the proposal? Ammend it?

Arnaud Le Hors: Vote on the proposal? Ammend it?

16:16:34 <bblfish> JohnArwe: I think a Web Page is an LDPB

John Arwe: I think a Web Page is an LDPB [ Scribe Assist by Henry Story ]

16:16:44 <codyburleson> SteveS: I did my +1 with a little verbiage

Steve Speicher: I did my +1 with a little verbiage

16:16:49 <Arnaud> PROPOSED: define hierarchy of resources as proposed by Alexandre Issue-90 Proposal 1

PROPOSED: define hierarchy of resources as proposed by Alexandre ISSUE-90 Proposal 1

16:16:51 <bblfish> ... ir you add it to an LDPC

Henry Story: ... if you add it to an LDPC

16:16:56 <bblfish> s/ir/if/
16:17:02 <SteveS> +1 would help to clarify the "not all Web resources are LDPR", to include interaction model details/explanation

Steve Speicher: +1 would help to clarify the "not all Web resources are LDPR", to include interaction model details/explanation

16:17:16 <codyburleson> codyburleson: +1

Cody Burleson: +1

16:17:24 <bblfish> +1 as explained above. Good names make it easier to speak about things, and currently it was a big mouthfull to speak about these resources.

Henry Story: +1 as explained above. Good names make it easier to speak about things, and currently it was a big mouthfull to speak about these resources.

16:17:24 <Ashok> +1

Ashok Malhotra: +1

16:17:24 <JohnArwe> +1 assuming we make the interaction model text changes that we discussed (type=ldp:Container => ldp:Container interaction model and so on)

John Arwe: +1 assuming we make the interaction model text changes that we discussed (type=ldp:Container => ldp:Container interaction model and so on)

16:17:42 <nmihindu> +1

Nandana Mihindukulasooriya: +1

16:17:42 <ericP> +1

Eric Prud'hommeaux: +1

16:17:46 <roger> 0

Roger Menday: 0

16:18:19 <bblfish> Agree LDPNR would be better

Henry Story: Agree LDPNR would be better

16:18:29 <Arnaud> RESOLVED: define hierarchy of resources as proposed by Alexandre Issue-90 Proposal 1

RESOLVED: define hierarchy of resources as proposed by Alexandre ISSUE-90 Proposal 1

16:18:46 <JohnArwe> ...so Henry if it's an LDPB, Ashok's thought exercise then asks... "is it really?  are there any interaction model requirements on LDPBs that a random web page would not satisfy?" ... type=LDPR/etc comes to mind.

John Arwe: ...so Henry if it's an LDPB, Ashok's thought exercise then asks... "is it really? are there any interaction model requirements on LDPBs that a random web page would not satisfy?" ... type=LDPR/etc comes to mind.

16:18:51 <codyburleson> Arnaud: I'd rather use LDPNR and maybe we could find something else for the G. I think people should volunteer new names. But we agree on the need for the hierarchy.

Arnaud Le Hors: I'd rather use LDPNR and maybe we could find something else for the G. I think people should volunteer new names. But we agree on the need for the hierarchy.

16:19:56 <JohnArwe> q+

John Arwe: q+

16:20:14 <codyburleson> Arnaud: An LDPG or LDPC is a named graph. We had discussions about what it meant.

Arnaud Le Hors: An LDPG or LDPC is a named graph. We had discussions about what it meant.

16:20:44 <codyburleson> SteveS: If your server only knew Turtle, you'd never know if this was satisfied.

Steve Speicher: If your server only knew Turtle, you'd never know if this was satisfied.

16:21:50 <JohnArwe> "triples belong to the representation of the hashless-ContainerResource" ... I asked on email why hashless, Alexandre clarified first that any URI is allowed, but then continued on to say: If you want to say to what Named Graph those membership triples belong to, it's only natural to use the hashless version of that subject membership.

John Arwe: "triples belong to the representation of the hashless-ContainerResource" ... I asked on email why hashless, Alexandre clarified first that any URI is allowed, but then continued on to say: If you want to say to what Named Graph those membership triples belong to, it's only natural to use the hashless version of that subject membership.

16:21:51 <codyburleson> Arnaud: The URI of the resource (document) is the graph name of which the triples are in; according to graph store protocol. I don't think there is any proposal to go beyond that.

Arnaud Le Hors: The URI of the resource (document) is the graph name of which the triples are in; according to graph store protocol. I don't think there is any proposal to go beyond that.

16:22:21 <JohnArwe> ...that last I found confusing, but if we can drop hashless and add PATCH-create as a covered case I think I'm fine

John Arwe: ...that last I found confusing, but if we can drop hashless and add PATCH-create as a covered case I think I'm fine

16:22:23 <Arnaud> ack JohnArwe

Arnaud Le Hors: ack JohnArwe

16:22:24 <codyburleson> ... The proposal doesn't rely on something more than Turtle.

... The proposal doesn't rely on something more than Turtle.

16:23:17 <codyburleson> JohnArwe: see my comments I just entered.

John Arwe: see my comments I just entered.

16:23:22 <ericP> q?

Eric Prud'hommeaux: q?

16:23:41 <codyburleson> Arnaud: I agree. The whole "hashless" thing is a bit of a distraction, IMHO.

Arnaud Le Hors: I agree. The whole "hashless" thing is a bit of a distraction, IMHO.

16:23:41 <bblfish> q+

Henry Story: q+

16:23:54 <Arnaud> ack bblfish

Arnaud Le Hors: ack bblfish

16:24:31 <codyburleson> Henry: Is there a way of doing a PUT where you don't overwrite content that already exists?

Henry Story: Is there a way of doing a PUT where you don't overwrite content that already exists?

16:25:08 <codyburleson> JohnArwe: Yes - you can make it a conditional. If the constion is satisfied then don't execute the request. There is an idiom for that in base HTTP>

John Arwe: Yes - you can make it a conditional. If the constion is satisfied then don't execute the request. There is an idiom for that in base HTTP>

16:25:10 <Arnaud> PROPOSED: add that an LDPG/LDPC is a Named Graph as proposed by Alexandre Issue-90 Proposal 2

PROPOSED: add that an LDPG/LDPC is a Named Graph as proposed by Alexandre ISSUE-90 Proposal 2

16:25:34 <JohnArwe> ...amended to cover PATCH-create and remove hashless

John Arwe: ...amended to cover PATCH-create and remove hashless

16:25:52 <SteveS> +1 (with amendments)

Steve Speicher: +1 (with amendments)

16:25:57 <Ashok> +1

Ashok Malhotra: +1

16:25:58 <codyburleson> codyburleson: +1

Cody Burleson: +1

16:26:04 <JohnArwe> +1 with amendments

John Arwe: +1 with amendments

16:26:18 <nmihindu> +0

Nandana Mihindukulasooriya: +0

16:26:22 <roger_> +0.5

Roger Menday: +0.5

16:26:24 <bblfish> +1 (will implement PUT myself for creation with Conditional on existing)

Henry Story: +1 (will implement PUT myself for creation with Conditional on existing)

16:27:13 <JohnArwe> http://www.w3.org/TR/rdf11-concepts/#dfn-named-graph is the citation ericp

John Arwe: http://www.w3.org/TR/rdf11-concepts/#dfn-named-graph is the citation ericp

16:27:14 <codyburleson> EricP: Does it make life harder for someone with a relational db?

Eric Prud'hommeaux: Does it make life harder for someone with a relational db?

16:28:10 <ericP> +1

Eric Prud'hommeaux: +1

16:28:13 <bblfish> I think that is how I implement this mysefl...

Henry Story: I think that is how I implement this mysefl...

16:28:15 <Arnaud> RESOLVED: add that an LDPG/LDPC is a Named Graph as proposed by Alexandre Issue-90 Proposal 2 amended to cover PATCH-create and remove hashless

RESOLVED: add that an LDPG/LDPC is a Named Graph as proposed by Alexandre ISSUE-90 Proposal 2 amended to cover PATCH-create and remove hashless

<codyburleson> Arnaud: with these two resolutions we can close issue-90

Arnaud Le Hors: with these two resolutions we can close ISSUE-90

16:29:02 <bblfish> Fine with me to extend

Henry Story: Fine with me to extend

16:29:27 <Zakim> -bblfish

Zakim IRC Bot: -bblfish

16:29:43 <codyburleson> Topic: Issue-89

6. ISSUE-89

16:29:52 <codyburleson> (for those who can stay on)

(for those who can stay on)

16:30:14 <codyburleson> Arnaud: If we have neither Alexandre or Henry, there is no reason to discuss this.

Arnaud Le Hors: If we have neither Alexandre or Henry, there is no reason to discuss this.

16:30:24 <SteveS> bblfish, rejoining?

Steve Speicher: bblfish, rejoining?

16:30:34 <codyburleson> +q unrelated

+q unrelated

16:30:38 <JohnArwe> bblfish are you dialing back in?

John Arwe: bblfish are you dialing back in?

16:30:40 <bblfish> euh can't rejoin the conf

Henry Story: euh can't rejoin the conf

16:30:55 <Arnaud> ack unrelated

Arnaud Le Hors: ack unrelated

16:30:57 <bblfish> it's restricted now

Henry Story: it's restricted now

16:31:25 <bblfish> That's probably why Alex could not rejoin btw

Henry Story: That's probably why Alex could not rejoin btw

16:31:30 <Arnaud> oh shoot, it's probably because we're officially out of time with the bridge

Arnaud Le Hors: oh shoot, it's probably because we're officially out of time with the bridge

16:32:29 <Zakim> -Ashok_Malhotra

Zakim IRC Bot: -Ashok_Malhotra

16:32:32 <Arnaud> maybe but, Alexandre should have been able to call, it was before the 1:30 limit

Arnaud Le Hors: maybe but, Alexandre should have been able to call, it was before the 1:30 limit

16:32:57 <Arnaud> but we'll stop here for today

Arnaud Le Hors: but we'll stop here for today

16:33:11 <Arnaud> maybe we can have a call next week?

Arnaud Le Hors: maybe we can have a call next week?

16:33:17 <bblfish> +1 for me

Henry Story: +1 for me

16:36:12 <codyburleson> JohnArwe: If anyone is planning to minus any of the outstanding proposals in Issue-89, we request you post on the mailing list what your contention / issue is.

John Arwe: If anyone is planning to minus any of the outstanding proposals in ISSUE-89, we request you post on the mailing list what your contention / issue is.

16:36:34 <codyburleson> ... Since we've run out of time.

... Since we've run out of time.

16:38:02 <codyburleson> Arnaud: I plan to try meeting next week. I will try to test the water before then. I think we need it.

Arnaud Le Hors: I plan to try meeting next week. I will try to test the water before then. I think we need it.

16:38:10 <Zakim> -JohnArwe

Zakim IRC Bot: -JohnArwe

16:38:20 <Zakim> -cody

Zakim IRC Bot: -cody

16:38:22 <Zakim> -Roger

Zakim IRC Bot: -Roger

16:38:24 <Zakim> -SteveS

Zakim IRC Bot: -SteveS

16:38:25 <Zakim> -EricP

Zakim IRC Bot: -EricP

16:38:25 <Zakim> -Arnaud

Zakim IRC Bot: -Arnaud

16:39:01 <Zakim> -nmihindu

Zakim IRC Bot: -nmihindu

16:39:03 <Zakim> SW_LDP()10:00AM has ended

Zakim IRC Bot: SW_LDP()10:00AM has ended

16:39:03 <Zakim> Attendees were Arnaud, SteveS, Roger, Alexandre, Ashok_Malhotra, bblfish, JohnArwe, codyburleson, cody, nmihindu, EricP

Zakim IRC Bot: Attendees were Arnaud, SteveS, Roger, Alexandre, Ashok_Malhotra, bblfish, JohnArwe, codyburleson, cody, nmihindu, EricP



Formatted by CommonScribe