IRC log of ldp on 2014-03-17

Timestamps are in UTC.

14:00:17 [RRSAgent]
RRSAgent has joined #ldp
14:00:17 [RRSAgent]
logging to http://www.w3.org/2014/03/17-ldp-irc
14:00:19 [trackbot]
RRSAgent, make logs public
14:00:19 [Zakim]
Zakim has joined #ldp
14:00:21 [trackbot]
Zakim, this will be LDP
14:00:21 [Zakim]
ok, trackbot; I see SW_LDP()10:00AM scheduled to start now
14:00:22 [trackbot]
Meeting: Linked Data Platform (LDP) Working Group Teleconference
14:00:22 [trackbot]
Date: 17 March 2014
14:00:45 [SteveS]
Zakim, who is here?
14:00:45 [Zakim]
SW_LDP()10:00AM has not yet started, SteveS
14:00:47 [Zakim]
On IRC I see RRSAgent, SteveS, nmihindu, codyburleson, stevebattle17, TallTed, deiu, jmvanel, Arnaud1, bblfish, sandro, betehess, ericP, Yves, trackbot
14:01:18 [Ashok]
Ashok has joined #ldp
14:01:33 [SteveS]
Zakim, who is on the phone?
14:01:34 [Zakim]
SW_LDP()10:00AM has not yet started, SteveS
14:01:35 [Zakim]
On IRC I see Ashok, Zakim, RRSAgent, SteveS, nmihindu, codyburleson, stevebattle17, TallTed, deiu, jmvanel, Arnaud1, bblfish, sandro, betehess, ericP, Yves, trackbot
14:02:04 [JohnArwe]
JohnArwe has joined #ldp
14:02:05 [Arnaud1]
zakim, this is sw_ldp
14:02:06 [Zakim]
Arnaud1, I see SW_LDP()10:00AM in the schedule but not yet started. Perhaps you mean "this will be sw_ldp".
14:02:25 [codyburleson1]
codyburleson1 has joined #ldp
14:02:34 [TallTed]
Zakim, who's here?
14:02:35 [Zakim]
SW_LDP()10:00AM has not yet started, TallTed
14:02:36 [Zakim]
On IRC I see codyburleson1, JohnArwe, Ashok, Zakim, RRSAgent, SteveS, nmihindu, codyburleson, stevebattle17, TallTed, deiu, jmvanel, Arnaud, bblfish, sandro, betehess, ericP, Yves,
14:02:36 [Zakim]
... trackbot
14:02:39 [TallTed]
Zakim, this is ldp
14:02:39 [Zakim]
TallTed, I see SW_LDP()10:00AM in the schedule but not yet started. Perhaps you mean "this will be ldp".
14:03:03 [TallTed]
sysreq needs to be involved...
14:03:05 [TallTed]
Zakim, mute me
14:03:05 [Zakim]
sorry, TallTed, I don't know what conference this is
14:05:11 [codyburleson1]
codyburleson1 has left #ldp
14:05:14 [codyburleson1]
codyburleson1 has joined #ldp
14:06:28 [sandro]
Hi
14:06:29 [Arnaud]
hi
14:06:30 [SteveS]
Hi, on the phone
14:06:32 [Ashok]
Hi
14:06:32 [ericP]
hi
14:06:38 [JohnArwe]
phone+irc
14:06:48 [deiu]
phone+irc too
14:06:49 [ericP]
phone only
14:06:58 [TallTed]
dual hi
14:06:58 [codyburleson1]
phone+irc
14:08:30 [bblfish]
hi
14:08:57 [Ashok]
scribenick: Ashok
14:09:03 [TallTed]
TallTed has changed the topic to: LDP WG: http://www.w3.org/2012/ldp - next agenda: http://www.w3.org/2012/ldp/wiki/Meetings:Telecon2014.03.17 - fyi: Zakim is ill
14:09:07 [Arnaud]
agenda: https://www.w3.org/2012/ldp/wiki/Meetings:Telecon2014.03.17
14:09:18 [JohnArwe]
regrets: steve battle
14:09:30 [Ashok]
Topic: Minutes of meeting March 10
14:09:55 [SteveS]
looks good to me
14:10:00 [Ashok]
No objection. Minutes approved
14:10:01 [codyburleson1]
I clicked on the link and scrolled real fast.
14:10:16 [ericP]
+1
14:10:26 [sandro]
regrets for next week
14:10:31 [Ashok]
RESOLUTION: Minutes of March 10 meeting are approved
14:10:49 [Ashok]
Next meeting is next Monday March 24
14:10:59 [Ashok]
Topic: Actions and Issues
14:11:30 [Ashok]
Eric: Progress on 206
14:11:47 [JohnArwe]
s/206/209/
14:13:01 [Ashok]
Arnaud: First issue is about stable vs. lossy paging
14:13:32 [Ashok]
... I suggest we open issue 94
14:14:21 [Arnaud]
PROPOSED: open ISSUE-94 Stable vs lossy paging
14:14:28 [JohnArwe]
+1
14:14:32 [ericP]
+1
14:14:33 [sandro]
+1
14:14:35 [TallTed]
+1
14:14:36 [codyburleson1]
+1
14:14:37 [deiu]
+1
14:14:46 [sandro]
SteveS: +1
14:14:55 [Ashok]
RESOLVED: Issue-94 is open
14:15:03 [bblfish]
Issue-94
14:15:03 [trackbot]
Issue-94 -- Stable vs lossy paging -- raised
14:15:03 [trackbot]
http://www.w3.org/2012/ldp/track/issues/94
14:15:06 [SteveS_]
SteveS_ has joined #ldp
14:15:07 [Arnaud]
PRISSUE-95
14:15:26 [Ashok]
Arnaud: Propose we open ISSUE-95
14:15:27 [Arnaud]
PROPOSED: open ISSUE-95 Paging of LDPRs without ordering
14:16:24 [ericP]
Arnaud: we could say that it's undefined and the client has to deal with whatever the server issues
14:16:40 [ericP]
... or we could remove it from the spec
14:16:55 [Ashok]
q+
14:17:08 [ericP]
sandro: if we don't have ordering on LDPRs, should we permit paging?
14:17:14 [ericP]
... i think the answer is "yes"
14:17:22 [JohnArwe]
+1 to sandro
14:17:23 [Arnaud]
ack ashok
14:18:08 [ericP]
Ashok: if we have a large, unordered container, are we saying that you can't page it or is it a server-defined order
14:18:20 [ericP]
s/server-defined order/server-defined order?/
14:19:15 [sandro]
q?
14:19:19 [ericP]
q+ say that lots of scenarios work fine with impl-defined ordering
14:19:20 [TallTed]
easiest is to make it server-defined, server-tracked, client just has to cope
14:19:20 [TallTed]
hardest is to make it fully negotiable between client/server
14:19:38 [SteveS]
+1 don't think we need this issue
14:19:38 [ericP]
q+ to say that lots of scenarios work fine with impl-defined ordering
14:20:15 [Arnaud]
ack ericP
14:20:15 [Zakim]
ericP, you wanted to say that lots of scenarios work fine with impl-defined ordering
14:20:45 [Ashok]
Sandro: Servers always have some ordering ... are we saying we cannot page in that situation
14:21:49 [Ashok]
Arnaud; maybe we can close this?
14:21:55 [SteveS]
q+
14:22:01 [Ashok]
s/;/:/
14:22:09 [Arnaud]
ack Steves
14:23:11 [Ashok]
Satve: Ordering is optional anyway ... so we should be able to page without ordering
14:23:26 [Ashok]
s/Satve/Steve/
14:24:53 [Ashok]
Sandro: We should be able to page LDPRs also ... what is difference between collections and RDF sources other than lifecycle
14:25:17 [sandro]
PROPOSED: Close ISSUE-95 stating that even if we don't have standards for communication ordering to clients in every situation, paging still applies to all LDP RDF Sources.
14:25:37 [ericP]
+1
14:25:45 [sandro]
+1
14:25:47 [SteveS]
+1
14:26:05 [sandro]
(and possibly all RDF Sources --- we're not sure what that might mean right now....)
14:26:06 [TallTed]
+1
14:26:07 [deiu]
+1
14:26:09 [Ashok]
Sandro: Propose paging applies to all rdf sources even is we cannot communicate to client
14:26:16 [bblfish]
+0 don't have an implementation yet
14:26:17 [JohnArwe]
+1 (blushes)
14:26:18 [Ashok]
s/is/if/
14:26:44 [codyburleson1]
+0 I'm not so sure
14:27:08 [deiu]
paging is implemented in rww.io too btw
14:27:10 [Arnaud]
RESOLVED: Close ISSUE-95 stating that even if we don't have standards for communication ordering to clients in every situation, paging still applies to all LDP RDF Sources
14:27:22 [Ashok]
Steve: Olievier had announced he had implemented paging as defined in spec
14:27:58 [Ashok]
s/Olievier/Olivier/
14:28:12 [SteveS]
Olivier Berger's note on his support for paging http://lists.w3.org/Archives/Public/public-ldp/2013Aug/0007.html
14:28:24 [Arnaud]
PROPOSED: Open ISSUE-96 Can we define ordering for LDPRs?
14:29:05 [Ashok]
Sandro: Argues we should open issue
14:29:16 [sandro]
sandro: let's let ourselves come up with a couple decent ways to communicate ordering
14:29:17 [sandro]
+1
14:29:20 [codyburleson1]
+1
14:29:20 [TallTed]
+1
14:29:23 [deiu]
+1
14:29:25 [SteveS]
+1
14:29:38 [Ashok]
RESOLVED: ISSUE-96 is open
14:29:46 [SteveS]
though I don't have a use case for it at the moment
14:29:51 [deiu]
q+
14:30:04 [Arnaud]
ack deiu
14:30:39 [Ashok]
Deiu: No way to filter on type of resource
14:31:19 [JohnArwe]
didn't we put filtering on the wish list, consciously?
14:31:26 [SteveS]
http://www.w3.org/2012/ldp/wiki/LDPNext#Filtering_collections
14:31:30 [Ashok]
Sandro: This is a different issue ... should we open?
14:33:07 [Ashok]
Arnaud: We have no mechanism for client to control paging
14:33:23 [Ashok]
... maybe put on wishlist
14:34:12 [Ashok]
Topic: Publictaions
14:35:08 [Ashok]
Arnaud: Spec was published on March 11 ... Last Call ends April 2
14:35:18 [Ashok]
... UCR was also published
14:35:41 [Ashok]
s/Publictaions/Publications/
14:37:51 [Ashok]
Topic: f2f
14:38:15 [Ashok]
Arnaud: I have set up a page ... please indicate attendance
14:39:08 [Ashok]
Topic: Spec issues
14:39:26 [Ashok]
Arnaud: Start with ISSUE-94
14:39:42 [MiguelAraCo]
MiguelAraCo has joined #ldp
14:39:52 [Ashok]
Arnaud: Sandro you have proposal for this
14:40:18 [Ashok]
Sandro: I sent email with proposal
14:40:44 [Arnaud]
http://lists.w3.org/Archives/Public/public-ldp-wg/2014Feb/0058.html
14:41:32 [sandro]
http://lists.w3.org/Archives/Public/public-ldp-wg/2014Feb/0064.html
14:42:02 [SteveS]
ericP I created ACTION-135 for fixing up docs hosted at namespace URI
14:42:03 [sandro]
that's my proposal, basically.
14:42:18 [ericP]
SteveS, roger and thanks
14:43:09 [Ashok]
Sandro: discusses motivation and proposal
14:43:56 [Ashok]
... not so hard to implement
14:44:42 [Ashok]
.. lossless paging
14:47:12 [Ashok]
Sandro: Some may want snapshot paging ... we don't need that under paging
14:48:05 [Ashok]
Ted: Discusses database options for paging etc.
14:48:40 [Ashok]
... client can specify option he wants
14:49:24 [TallTed]
https://en.wikipedia.org/wiki/Isolation_%28database_systems%29
14:49:34 [JohnArwe]
which of the 4 links are required under this proposal?
14:49:57 [TallTed]
"Read uncommitted" is basically what Sandro's talking about
14:50:24 [JohnArwe]
q+ to ask which links are required
14:50:42 [sandro]
client MUST get every triple in intersect(g0, g1, ... gn) and MAY get any triples in the union (g0, g1, .. gn)
14:50:56 [Arnaud]
ack JohnArwe
14:50:56 [Zakim]
JohnArwe, you wanted to ask which links are required
14:51:10 [Ashok]
Ted: Please rend the article I quoted. Good coverage of the database options
14:51:31 [Ashok]
JohnArwe: Asks about which links are required
14:52:35 [JohnArwe]
Sandro: "if you do fwd paging, X; if you do bwd paging, Y"
14:52:36 [Ashok]
Sandro: We want to allow folks to do forward paging only or backward paging only
14:52:42 [TallTed]
and... https://en.wikipedia.org/wiki/Cursor_%28databases%29
14:53:04 [Ashok]
Eric: I argued for singly linked list vs. doubly linked and we got pushback
14:53:24 [TallTed]
we're very much reinventing DBMS
14:53:50 [Ashok]
Yes, Ted, but don't have locking :-)
14:54:19 [bblfish]
TallTed, we're improving DBMS is anything, since they are not RESTful or have URIs
14:54:23 [TallTed]
locking is a later implementation
14:54:42 [Ashok]
Discussion on which pointers are needed
14:54:59 [Arnaud]
STRAWPOLL: add Sandro's requirement that client MUST get every triple in intersect(g0, g1, ... gn) and MAY get any triples in the union (g0, g1, .. gn), where gx represents the graph at different times while the client is traversing the pages
14:55:14 [TallTed]
DBMS are neither RESTful nor not. that's orthogonal.
14:55:14 [TallTed]
URIs are "super-keys" -- universally usable, beyond table-bound DBMS keys
14:55:32 [bblfish]
they are not, so we're not re-inventing them
14:55:44 [TallTed]
if anything, this might be considered a next stage of DBMS -- but ignoring what exists and reinventing it is not effective use of time
14:55:46 [sandro]
gx is the the state of the graph when the client does GET x
14:55:56 [Ashok]
+1
14:55:56 [ericP]
+1
14:55:58 [sandro]
+1
14:56:03 [TallTed]
+1
14:56:15 [deiu]
+0 (needs testing)
14:56:21 [bblfish]
+1
14:56:21 [SteveS]
+.5
14:56:22 [JohnArwe]
+0.5 nothing jumping out as a problem, but Definitely requires more thought
14:56:46 [Ashok]
Topic: 209
14:57:47 [Ashok]
Eric: We are trying to get shortcut on 303/200 pattern we are trying to push TAG's position
14:58:23 [Ashok]
... MarkN is content to use Atom where paging info is in the payload
14:58:58 [Ashok]
... they think we are complication HTTP for sake of simplifying our app.
14:59:54 [Ashok]
... or use SPDY ... which gives 303 and content of 303 ... if on save server
15:00:04 [bblfish]
q+
15:00:04 [Ashok]
s/save/same/
15:00:29 [Ashok]
Eric: We need to agree that HTTP 2.0 will be deployed
15:01:18 [Ashok]
... we may have situations where we do not have origin
15:02:00 [Ashok]
... but client may not always enforce same-origin
15:02:45 [Arnaud]
ack bblfish
15:03:12 [Ashok]
Henry: SPDY is meant to be semantically equivalent
15:04:05 [Ashok]
... I don't think SPDT give you what you asked for and a bunch of other info
15:04:39 [Ashok]
s/SPDT/SPDY/
15:05:13 [Ashok]
EricP: Would appreciate help with usecases etc.
15:05:31 [bblfish]
ericP: please send us pointers to those list
15:06:31 [Ashok]
Arnaud: ADJOURNED
15:06:40 [codyburleson1]
codyburleson1 has left #ldp
15:06:41 [JohnArwe]
-JohnArwe
15:09:51 [bblfish]
I kind of agree with ericP here. If some requests only make sense on certain resources, then having clients do a lot of Prefers...
15:13:00 [bblfish]
The Location response-header field defines the exact location of the resource that was identified by the Request-URI. For 3xx responses, the location must indicate the server's preferred URL for automatic redirection to the resource. Only one absolute URL is allowed.
15:13:05 [bblfish]
http://www.w3.org/Protocols/HTTP/1.0/spec.html#Location
15:20:21 [Yves]
better take the definition here: http://tools.ietf.org/html/draft-ietf-httpbis-p2-semantics-26#section-7.1.2
15:21:41 [bblfish]
thanks
15:25:37 [bblfish]
http://java.dzone.com/articles/using-spdy-and-http
15:25:55 [bblfish]
Using SPDY and HTTP Transparently Using Netty
15:31:58 [sandro]
silence.
15:32:00 [sandro]
???
15:32:14 [sandro]
hanging up. bye folks.
15:35:31 [Zakim]
restarting zakim-bot in 2 minutes to recover bridge connection
17:09:51 [Arnaud]
Arnaud has joined #ldp
17:19:50 [deiu]
deiu has joined #ldp
18:47:04 [SteveS]
SteveS has joined #ldp
20:21:58 [jmvanel]
jmvanel has joined #ldp
21:27:56 [SteveS_]
SteveS_ has joined #ldp
23:38:08 [SteveS]
SteveS has joined #ldp