IRC log of ldp on 2014-07-21

Timestamps are in UTC.

13:59:15 [RRSAgent]
RRSAgent has joined #ldp
13:59:15 [RRSAgent]
logging to http://www.w3.org/2014/07/21-ldp-irc
13:59:17 [trackbot]
RRSAgent, make logs public
13:59:17 [Zakim]
Zakim has joined #ldp
13:59:19 [trackbot]
Zakim, this will be LDP
13:59:19 [Zakim]
ok, trackbot; I see SW_LDP()10:00AM scheduled to start in 1 minute
13:59:20 [trackbot]
Meeting: Linked Data Platform (LDP) Working Group Teleconference
13:59:20 [trackbot]
Date: 21 July 2014
14:00:17 [Zakim]
SW_LDP()10:00AM has now started
14:00:22 [JohnArwe]
JohnArwe has joined #ldp
14:00:24 [Zakim]
+JohnArwe
14:00:36 [JohnArwe]
trackbot, start meeting
14:00:38 [trackbot]
RRSAgent, make logs public
14:00:40 [trackbot]
Zakim, this will be LDP
14:00:40 [Zakim]
ok, trackbot, I see SW_LDP()10:00AM already started
14:00:41 [trackbot]
Meeting: Linked Data Platform (LDP) Working Group Teleconference
14:00:41 [trackbot]
Date: 21 July 2014
14:01:03 [Zakim]
+ +33.9.50.97.aaaa
14:01:04 [Zakim]
+[IPcaller]
14:01:11 [Zakim]
+Sandro
14:01:13 [Arnaud]
zakim, aaa is me
14:01:13 [Zakim]
sorry, Arnaud, I do not recognize a party named 'aaa'
14:01:16 [codyburleson]
Zakim, IPcaller is me.
14:01:16 [Zakim]
+codyburleson; got it
14:01:17 [Zakim]
+??P16
14:01:21 [Zakim]
+[IBM]
14:01:21 [Ashok]
Ashok has joined #ldp
14:01:24 [Arnaud]
zakim, aaaa is me
14:01:24 [Zakim]
+Arnaud; got it
14:01:29 [SteveS]
Zakim, [IBM] is me
14:01:29 [Zakim]
+SteveS; got it
14:01:31 [nmihindu]
Zakim, ??P16 is me
14:01:31 [Zakim]
+nmihindu; got it
14:01:36 [nmihindu]
Zakim, mute me
14:01:36 [Zakim]
nmihindu should now be muted
14:02:10 [Zakim]
+Ashok_Malhotra
14:02:49 [Zakim]
+ +43.660.274.aabb
14:02:55 [JohnArwe]
zakim, who's on?
14:02:55 [Zakim]
I don't understand your question, JohnArwe.
14:03:02 [MiguelAraCo]
MiguelAraCo has joined #ldp
14:03:07 [sergio]
Zakim, +43.660.274.aabb is me
14:03:07 [Zakim]
+sergio; got it
14:03:12 [JohnArwe]
zakim, who's on the call?
14:03:12 [Zakim]
On the phone I see JohnArwe, Arnaud, codyburleson, Sandro, nmihindu (muted), SteveS, Ashok_Malhotra, sergio
14:04:04 [Zakim]
+ericP
14:04:16 [Zakim]
+ +1.857.928.aacc
14:04:29 [betehess]
Zakim, aacc is Alexandre
14:04:29 [Zakim]
+Alexandre; got it
14:05:07 [JohnArwe]
regrets: andrei
14:05:53 [betehess]
slides for LD Patch walk-through are at http://lists.w3.org/Archives/Public/public-ldp-wg/2014Jul/att-0068/ldpatch.html
14:06:27 [JohnArwe]
frac
14:06:34 [JohnArwe]
scribe: johnarwe
14:06:54 [JohnArwe]
topic: Minutes of last meeting - July 14
14:07:07 [JohnArwe]
Proposal: Approve the minutes of the 14 July teleconf: http://www.w3.org/2013/meeting/ldp/2014-07-14
14:08:04 [JohnArwe]
Arnaud: since trackbot was not started on time, reconstructed from private irc log
14:08:19 [JohnArwe]
resolved: minutes approved w/o objections
14:08:30 [JohnArwe]
topic: Next meeting
14:08:43 [JohnArwe]
Arnaud: July 28 ... any objections?
14:08:56 [JohnArwe]
resolved: July 28 next meeting
14:09:05 [JohnArwe]
topic: actions + issues
14:09:08 [JohnArwe]
action-143?
14:09:08 [trackbot]
action-143 -- Nandana Mihindukulasooriya to Validate the JSON-LD examples of the primer -- due 2014-07-12 -- PENDINGREVIEW
14:09:08 [trackbot]
http://www.w3.org/2012/ldp/track/actions/143
14:09:08 [codyburleson]
Regrets, in advance, for next meeting (July 28th); I'm on vacation.
14:09:33 [nmihindu]
yes, I corrected few errors and validated them.
14:09:50 [JohnArwe]
resolved: close action-143
14:10:08 [JohnArwe]
action-77?
14:10:08 [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
14:10:08 [trackbot]
http://www.w3.org/2012/ldp/track/actions/77
14:10:58 [JohnArwe]
Ashok: enough review has happened, ok to close. got some feedback from TallTed.
14:11:07 [JohnArwe]
resolved: close action-77
14:11:23 [JohnArwe]
Arnaud gloats that all actions now closed.
14:12:20 [JohnArwe]
topic: LD Patch Format Walk-through
14:12:39 [betehess]
http://lists.w3.org/Archives/Public/public-ldp-wg/2014Jul/att-0068/ldpatch.html
14:12:51 [JohnArwe]
link to slides above
14:13:06 [JohnArwe]
Andrei presents slides
14:13:18 [JohnArwe]
s/Andrei/Alexandre/
14:13:23 [SteveS]
betehess is presenting
14:15:07 [Zakim]
+OpenLink_Software
14:15:13 [TallTed]
Zakim, OpenLink_Software is temporarily me
14:15:13 [Zakim]
+TallTed; got it
14:15:14 [TallTed]
Zakim, mute me
14:15:14 [Zakim]
TallTed should now be muted
14:15:55 [JohnArwe]
ashok: what is a pathological graph here? betehess: graphs containing bnodes that cannot be addressed
14:16:11 [JohnArwe]
arnaud: document what it means in here
14:16:19 [JohnArwe]
...i.e. as limitations
14:16:34 [sandro]
+1 include section on LIMITATIONS, showing what kind of bnode structures CANT be patched
14:20:33 [ericP]
even turtle doesn't technically need '.'s
14:21:00 [ericP]
(i haven't tried without so perhaps some ambiguity would arise)
14:21:10 [JohnArwe]
sandro: wrt syntax on page 5, perhaps could allow quads with relatively simple changes)
14:22:28 [JohnArwe]
s/changes)/changes/
14:23:46 [ericP]
q+ to ask about /schema:attendee[/foaf:name!]/schema:name
14:24:18 [JohnArwe]
Arwe: need to define "pure" equality; offline email said it means the object, any datatype, and any lang tag need to all match
14:24:31 [JohnArwe]
...+ exact lexical match
14:24:38 [sandro]
question - recursive constraints? sequential constraints?
14:25:04 [Arnaud]
ack ericP
14:25:04 [Zakim]
ericP, you wanted to ask about /schema:attendee[/foaf:name!]/schema:name
14:25:31 [JohnArwe]
betehess: ericp's example would mean exactly one foaf:name
14:25:35 [sandro]
q+
14:25:52 [Arnaud]
ack sandro
14:26:28 [JohnArwe]
sandro: recursive constraints? alex: yes. and sequential? yes.
14:27:01 [Ashok]
q+
14:27:26 [Arnaud]
ack Ashok
14:27:37 [JohnArwe]
sandro: timbl historically used caret for inverse instead of minus. Alex: authors don't care; left it as in Pierre-Antoinne's solution; wg members can make proposal to change rather than authors responding informally.
14:27:46 [sandro]
how about aligning with http://www.w3.org/TR/sparql11-query/#pp-language ???
14:28:01 [sandro]
(which uses ^ instead of - )
14:28:22 [ericP]
sandro, http://www.w3.org/TR/2013/REC-sparql11-query-20130321/#rPathEltOrInverse says "^" for inverse
14:28:37 [sandro]
and dont use ! for this
14:28:45 [sandro]
q+
14:28:51 [JohnArwe]
ashok: pathological you said have unaddressable bnodes, right? alexandre: used to traverse graph, could reference bnode w/o knowing its ID, using path expression ... that's exactly why that's they're in the syntax.
14:29:26 [ericP]
i guess there's some precedent for using "#" in formal maths, but that'd be nuts
14:29:36 [sandro]
Ashok, a pathological graph: <a> <b> _:x,_:y.
14:29:41 [JohnArwe]
ashok: are there instances where that's still not enough? alex: yes, if the conditions are not unique. recent article published that explores use of bnodes in web datasets.
14:29:48 [JohnArwe]
...will followup on mailing list.
14:30:02 [Arnaud]
ack sandro
14:30:53 [ericP]
+1 to re-using SPARQL where possible (hoping there's something left to use in place of '!')
14:30:58 [JohnArwe]
sandro: you should align with sparql property path syntax whereever you overlap, like inverse path. and avoid using bang, since they use it with a different syntax. don't gratuitously differ. Alex: we can review together.
14:31:25 [JohnArwe]
slide 7
14:32:08 [JohnArwe]
slide 8
14:32:45 [ericP]
Bind ?grandparent <#> /gen:parent .
14:32:51 [ericP]
Bind ?grandparent <#> /gen:parent . # now a grandparent?
14:33:17 [ericP]
sorry, s/<#>/?grandparent/
14:33:57 [sandro]
Suggest ".." instead of ">" as slice operator.
14:33:58 [JohnArwe]
betehess: if not careful, updatelist can lead to graphs 'pointing to nowhere'
14:34:33 [JohnArwe]
ericp: can use a bound variable within its own (subsequent) binding? alexandre: yes
14:34:38 [JohnArwe]
slide 9
14:35:12 [sandro]
q+ to ask why this is better than SPARQL UPDATE ?
14:35:29 [sandro]
q+ to ask why this is better than SPARQL UPDATE or a profile of SPARQL UPDATE
14:37:38 [JohnArwe]
alexandre: very important for specifying operational semantics, i.e. math meaning given to the patch.
14:38:01 [Arnaud]
ack sandro
14:38:01 [Zakim]
sandro, you wanted to ask why this is better than SPARQL UPDATE ? and to ask why this is better than SPARQL UPDATE or a profile of SPARQL UPDATE
14:39:12 [JohnArwe]
sandro: like this in general. what's advantage of this over (some profile of) sparql update? most people's problem with sparql is the potentially exponential performance, and this proposal has the same complexity.
14:39:28 [Yves]
14:39:38 [JohnArwe]
betehess: don't think complexity is exponential; provide example please.
14:41:07 [JohnArwe]
sandro: social graph; /foaf:knows repeated 5x; that has 100million triples. Alex: no, not true. ... sandro: if branching factor is 1K, doesn't this branch rapidly? alex: no, don't remember other branches. ... more words...
14:41:26 [Ashok]
Sandro, why is this worse than XPath?
14:44:05 [JohnArwe]
meta-question to arnaud: there are other topics on the agenda, what is the time allocated to this current topic vs the later one(s)?
14:44:34 [JohnArwe]
arnaud: take offline to mailing list
14:44:55 [JohnArwe]
slide 9: status "almost ready" ...for what? Alex: adding operational semantics.
14:45:54 [JohnArwe]
Arnaud: wg had 1 resolution to investigate this. we've fulfilled that. sandro's questions important. now need to decide if overall approach of _this proposal_ is ok with WG.
14:46:14 [JohnArwe]
Alex: agree; if overall approach is ok, ready to publish FPWD
14:47:06 [SteveS]
q+ to talk about how we addressed this before
14:47:40 [codyburleson]
We are actually using a patch format designed by Sandro, so we kind of feel like honoring his opinion on this matter (whatever it may be). And then, our one concern is just the fact that you have to know yet another language to understand LDP.
14:47:42 [Arnaud]
ack Steves
14:47:42 [Zakim]
SteveS, you wanted to talk about how we addressed this before
14:48:04 [JohnArwe]
sandro: would like someone to articulate why they'd use this proposal but would not use sparql update.
14:49:15 [JohnArwe]
arnaud: concerned about spending lots of time fixing details if anyone in the wg has major issues with approach.
14:50:33 [JohnArwe]
ericp: feels like people would have to write another parser, awkward to dup content in "competing" stds aside from an isomorphism. Alex: did not want to rely on sparql implementation(s).
14:51:19 [JohnArwe]
arnaud: Alexandre, can you write the "why not sparql update" section to flush out?
14:51:21 [betehess]
https://www.w3.org/2012/ldp/wiki/LDP_PATCH_Proposals
14:51:59 [JohnArwe]
sandro: fine if we want to publish FPWD saying very explicitly that we recognize overlap and invite comments on that specifically.
14:53:04 [JohnArwe]
... could also name a subset of sparql update equivalent to this, ericp. If it's not hard to exactly define the subset.
14:53:11 [TallTed]
+1 LIMITATIONS section (and I recommend including some of the discussion/experience/justification for why these exist, and possibly what the other options were, and why they were not pursued)
14:53:11 [TallTed]
also +1 FPWD with something like "while working through this, we found it to basically respecify SPARQL Update... without relying on a back-end SPARQL implementation. does anyone actually need this?"
14:53:41 [JohnArwe]
Alex: issues with managing the dependencies on syntax etc. Thinks it's harder than Sandro/EricP appear to think/
14:53:42 [sandro]
https://www.w3.org/2012/ldp/wiki/LDP_PATCH_Proposals#EricP.27s_proposal
14:55:12 [JohnArwe]
discussion... not exactly ericp's proposal, really a version of that which "looks more like LDPatch"
14:55:40 [JohnArwe]
sandro: missing variables and delete that I'd really like. thinking of long string literals.
14:55:52 [sandro]
sandro: I want "delete <s> <p> *" and "delete <s> * * "
14:55:58 [bblfish]
bblfish has joined #ldp
14:57:17 [JohnArwe]
alexandre: path is optional. bind it. sandro: don't want to transmit the literal at all. very hard to do with your proposal, although could be added pretty easily. Alex: if we change binding to be to a set of nodes, should fix that.
14:58:11 [JohnArwe]
Action: betehess to make updates and handle sparql update reasoning
14:58:12 [trackbot]
Created ACTION-144 - Make updates and handle sparql update reasoning [on Alexandre Bertails - due 2014-07-28].
14:58:12 [Zakim]
-Arnaud
14:58:33 [Arnaud]
shoot
14:58:46 [JohnArwe]
should we wait for your chairness?
14:59:05 [Arnaud]
no, wha'ts the status of paging?
14:59:29 [Zakim]
+Arnaud
14:59:35 [JohnArwe]
cody: arnaud said I need to provide a pure-html (no respec) version, put that under TR dir? Steve S: will help.
14:59:51 [nmihindu]
yes. I am.
15:00:21 [codyburleson]
Thanks
15:01:23 [codyburleson]
Cody will submit the pure HTML (final) Best Practices document today into the TR directory. Will be on vacation next week. Nandana will help Cody keep an eye on any process to push that document through.
15:02:01 [JohnArwe]
topic: paging
15:02:21 [JohnArwe]
late-breaking email with list of "must be fixed by LC" changes
15:02:30 [betehess]
thanks, bye
15:02:32 [Zakim]
-Alexandre
15:02:34 [Zakim]
-Sandro
15:02:34 [JohnArwe]
discussion of process; hope to work most out on list w/o formal issues
15:02:35 [Zakim]
-TallTed
15:02:36 [Zakim]
-Arnaud
15:02:37 [Zakim]
-sergio
15:02:38 [Zakim]
-ericP
15:02:39 [Zakim]
-Ashok_Malhotra
15:02:40 [SteveS]
I had a few comments as well on paging
15:02:43 [Zakim]
-SteveS
15:02:48 [JohnArwe]
arnaud: bulk of agenda next week to be paging
15:02:49 [Zakim]
-codyburleson
15:02:52 [Zakim]
-JohnArwe
15:02:54 [Zakim]
-nmihindu
15:02:54 [Zakim]
SW_LDP()10:00AM has ended
15:02:54 [Zakim]
Attendees were JohnArwe, +33.9.50.97.aaaa, Sandro, codyburleson, Arnaud, SteveS, nmihindu, Ashok_Malhotra, sergio, ericP, +1.857.928.aacc, Alexandre, TallTed
15:14:25 [bblfish]
bblfish has joined #ldp
16:10:32 [SteveS]
SteveS has joined #ldp
16:28:24 [codyburleson]
codyburleson has joined #ldp
16:52:51 [bblfish]
bblfish has joined #ldp
16:58:38 [Zakim]
Zakim has left #ldp
17:03:57 [bblfish]
bblfish has joined #ldp
17:25:24 [stevebattle11114]
stevebattle11114 has joined #ldp
17:54:11 [SteveS_]
SteveS_ has joined #ldp
17:59:56 [bblfish]
bblfish has joined #ldp
18:21:29 [bblfish]
bblfish has joined #ldp
18:24:26 [bblfish]
bblfish has joined #ldp
18:48:54 [bblfish]
bblfish has joined #ldp
20:17:51 [bblfish]
bblfish has joined #ldp
22:13:22 [bblfish]
bblfish has joined #ldp