12:58:23 RRSAgent has joined #shapes 12:58:23 logging to http://www.w3.org/2015/05/19-shapes-irc 12:58:25 RRSAgent, make logs rdf-data-shapes 12:58:25 Zakim has joined #shapes 12:58:27 Zakim, this will be SHAPES 12:58:27 I do not see a conference matching that name scheduled within the next hour, trackbot 12:58:28 Meeting: RDF Data Shapes Working Group Teleconference 12:58:28 Date: 19 May 2015 12:59:41 hknublau has joined #shapes 13:00:02 aryman has joined #shapes 13:00:52 hi there 13:01:05 beware, we couldn't get Zakim for this meeting 13:01:11 and we have to use WebEx 13:01:12 pfps has joined #shapes 13:01:27 zakim, who is on the phone? 13:01:27 sorry, pfps, I don't know what conference this is 13:01:29 On IRC I see pfps, aryman, hknublau, Zakim, RRSAgent, kcoyle, Dimitris, elf-pavlik, rhiaro_, Arnaud, trackbot, ericP 13:01:30 we will use IRC for logging and queueing as usual 13:01:36 zakim, this is shapes 13:01:36 sorry, pfps, I do not see a conference named 'shapes' in progress or scheduled at this time 13:01:42 but for voice we use webex 13:01:47 that won't work peter 13:01:58 grrr 13:02:15 zakim doesn't know anything about webex 13:04:33 the host has *ultimate* control on WebEx, and everyone else has squat 13:04:55 Labra has joined #shapes 13:04:56 are you on webex peter? 13:05:12 I see one "Call-in User_3" 13:06:46 Which webex? 13:07:41 which web interface?? 13:08:01 http://www.webex.com/ 13:08:14 and click join 13:08:17 https://mit.webex.com/mw0401lsp13/mywebex/default.do?siteurl=mit 13:08:29 the web interface wasn't in the email from Eric and thus I didn't put it on the meeting page 13:08:31 and you can use the audio there, no need to phone in 13:08:46 no, it wasn't in eric's email -- i sent a follow-up mail 13:08:54 hsolbrig has joined #shapes 13:09:34 hsolbrig: log on to https://mit.webex.com/mw0401lsp13/mywebex/default.do?siteurl=mit 13:09:49 Arnaud has changed the topic to: RDF Data Shapes WG: https://www.w3.org/2014/data-shapes/ - Next agenda: https://www.w3.org/2014/data-shapes/wiki/F2F3#Agenda Use WebEx for audio connection - see F2F page for details 13:11:01 WebEx is *not* very Linux compatible 13:11:53 pfps, it doesn't work from Linux, but there is an android app that works very well 13:12:26 That's astonishing - WebEx audio is known to be problematic 13:14:09 Screen sharing can require *lots* of bandwidth, which may not be possible for all 13:18:06 chair: Arnaud 13:18:12 scribenick: aryman 13:19:48 agenda: https://www.w3.org/2014/data-shapes/wiki/F2F3#Day_1_-_Tuesday_May_19 13:20:09 but no muting from IRC using Zakim 13:20:20 aryman: I think we're getting feedback from you - can you mute? 13:20:34 iovka has joined #shapes 13:21:30 @kcoyle I muted. Is that better? 13:21:47 aryman: yes, thx 13:22:09 http://www.w3.org/2015/Talks/0519-shacl-egp/ 13:22:15 TOPIC: ShEx proposal deep-dive 13:24:17 Why all the marketing stuff? I was expecting something technical. 13:26:45 ericP is presenting the ShEx/SHACL deep dive 13:30:31 slide 7 13:30:48 agenda+ How does ShEx compile to SPARQL? 13:33:02 sorry, I thought I was muted 13:34:50 q+ I'm confused: a triple constraint can be datatype? 13:35:11 q+ i'm confused - a triple constraint can be datatpe? 13:35:31 q+ 13:36:40 ack pfps 13:37:48 DNF is your friend 13:38:31 s/DNF/BNF/ 13:39:32 agenda+ this presentation does not appear to allow conjunction in valueClass 13:40:41 q+ What does the .* mean? (on 11) 13:40:59 q+ 13:41:44 ack aryman 13:44:19 q+ 13:44:46 ack pfps 13:46:26 pfps: the ShEx language in this presentation does not match the language in the semantics document 13:46:43 iovka: yes, there are some differences 13:47:30 pfps: in the presentation, value class and conjunction appear to be missing 13:48:13 http://www.w3.org/2005/01/yacker/uploads/ShEx2?lang=perl&markup=html#prod-ShEx2-tripleConstraint 13:48:21 ericP: conjunction is not part of a triple constraint, but appears elsewhere 13:48:37 conjunction is in the valueClass position 13:49:26 pfps: I am concerned that the syntax is different than the published semantics 13:51:59 detailed discussion of semantics 13:53:54 moving on to discussion of closed shapes 13:55:03 A big difference between the presentation and the language in the semantics is that the semantics permits negated shapes whereas the presentation has negated tripleConstraints 13:57:02 Arnaud: how are shapes associated with data? 13:58:01 ericP: two mechanisms 1) the data points to shape, e.g. instanceShape, classShape, 2) defined by the application (uses XSD and WSDL analogy) 13:58:07 agenda+ how does ShEx associate a node with a shape? 13:58:07 http://w3c.github.io/data-shapes/data-shapes-primer/#h-associations 14:02:34 What is the contract between the semantics and the semantic actions? 14:06:12 What about backtracking? Are actions undone? 14:07:12 iovka: in the semantics doc, semantic actions have no side effects 14:07:40 ericP: in addition we can use semantic actions with side effects 14:07:50 q+ 14:09:36 ericP: we don't have semantics for actions with side effects 14:09:56 pfps: if no semantics for side effects, then it is out of scope 14:10:18 ack aryman 14:10:27 one sec 14:11:09 can't find the window!! 14:11:11 go on 14:11:56 found it! 14:11:59 may I ask? 14:12:10 q+ 14:12:16 let him finish that slide and we'll get to you 14:13:20 ack aryman 14:14:46 The semantics (now) needs multiple models, so determining whether to call an action with side effects becomes very problematic 14:15:44 aryman: semantic actions with side effects is useful, e.g. to generate code, but we need deterministic semantics, e.g. like ANTLR 14:15:51 q+ 14:16:11 ack pfps 14:16:36 ericP: our implementation does this in two passes, first generating a tree of recognized triples, second traversing the tree depth first 14:19:06 I don't believe either of the examples work in the semantics, which gives the interface to the semantic extensions in terms of three arguments, graph, focus node, and language 14:20:09 the focus node is ?this 14:20:30 s/is ?this/is linked to ?this/ 14:21:16 The examples include other variables that do not seem to be free 14:21:45 simonstey has joined #shapes 14:22:38 I think that usually eric uses ?o for the focus node, and in the second example on slide 18 he mixed ?o and ?this, which both stand for the focus node 14:23:03 sorry, I'm wrong ! 14:23:59 q+ 14:24:10 ack aryman 14:25:22 the question is when is the semantics doc going to be stable? 14:25:24 aryman: when will your semantics doc be frozen so we can review it in detail? 14:26:20 iovka: will announce the version when it is ready for review 14:27:06 agenda? 14:27:16 ACTION: iovka to announce a stable version of the semantics document so we can have a review cycle 14:27:17 Created ACTION-25 - Announce a stable version of the semantics document so we can have a review cycle [on Iovka Boneva - due 2015-05-26]. 14:27:44 Zakim, next agendum 14:27:44 agendum 1. "How does ShEx compile to SPARQL?" taken up [from pfps] 14:29:10 q+ 14:29:55 q+ 14:30:23 pfps: Is every aspect of ShEx compilable to SPARQL? 14:30:44 ericP: Not if valueShape is present 14:31:04 q- 14:31:50 q- 14:31:52 pfps: How can you assure the translation to SPARQL is correct? 14:32:04 q+ 14:32:12 ericP: Just wrote some test cases. Is it possible? 14:32:40 ack iovka 14:32:41 pfps: There is some work on translating OWL constraints to SPARQL 14:33:19 iovka: Without recursion, ShEx can be translated to SPARQL 14:33:46 iovka: We could potentially prove the translated is correct 14:34:05 s/translated/translation/ 14:34:11 The treatment of negation in the semantics does not appear to have an obvious translation into SPARQL. 14:34:52 ericP: iovka has done some analysis of complexity 14:35:18 iovka discusses some complexity results 14:35:49 How is VIRTUAL handled in the semantics? 14:36:13 Take a 15 minute break now 14:36:21 Resume at 10:50 AM 14:41:30 karen, i just updated the title of http://www.w3.org/2015/Talks/0519-shacl-egp/#(22) to clarify that it was for your 2nd use case 14:44:43 http://www.slideshare.net/HolgerKnublauch/shacl-specification-draft 14:50:31 +1 14:50:33 +! 14:50:43 +1 14:50:56 TOPIC: Holger's proposal deep-dive 14:56:26 agenda+ Status of recursive shapes 14:57:01 q+ 14:57:59 Does the prototype depend on TBC? 14:58:28 aryman: you need to mute - we get echo from you 14:58:32 aryman: sorry 14:58:53 ack aryman 15:00:28 hknublau: the prototyoe includes a stand-alone SHACL engine based on the Jena API, also an editor in TBC 15:09:01 s/prototyoe/prototype/ 15:15:29 Labra has joined #shapes 15:18:38 Labra has joined #shapes 15:20:30 dksharma has joined #shapes 15:22:57 +q 15:28:04 -q 15:28:30 sorry iovka, if I missed your request in a timely manner! 15:39:33 q+ 15:39:49 valueShape appears to be missing from slide 25 15:40:39 ack aryman 15:43:48 Apologies for having leave. Have to catch a plane. Fascinating talk, Holger and I want to pursue it further... 15:44:20 I see valueShape in Slide 26 in sh:PropertyConstraint 15:45:35 q+ 15:46:25 Shouldn't NativeConstraint be language neutral? 15:46:44 i.e have subclasses for SPARQL, JS. 15:46:53 ack aryman 15:48:40 +q 15:49:13 ack iovka 15:49:33 hknublau: NativeConstraint can have more than one language, all being equivalent, the engine picks the language it supports 15:50:11 iovka: how do you ensure that alternate language strings are equivalent 15:50:52 hknublau: it is the responsibility of the constraint author to write equivalent definitions 15:51:25 Arnaud: this is a general problem with extensions 15:52:03 hknublau: we only defines SPARQL since there is no API for JS 15:52:05 q+ 15:52:16 ack aryman 15:52:58 why not use JSON-LD 15:54:01 Arthur is talking about the JSON encoding of SPARQL results, I think. 15:54:03 also use SPARQL JSON result format for output 15:54:42 Both, use JSON-LD as the input, use SPARQL JSON Result format for the output 15:55:07 q+ 15:55:20 ack pfps 15:55:57 pfps: If there are two defs, which one is the "controlling" def? 15:56:27 +q 15:56:32 holger: the SPARQL one 15:56:35 hklunblau: I suggest use SPARQL of the controlling def 15:56:41 s /of/as/ 15:56:59 ack iovka 15:57:33 iovka: Do you have SPARQL defs for all features? 15:57:48 hknublau: Yes 15:58:24 q+ to ask, per iovka's q, is there a defn for valueShape? 15:58:59 +q 16:00:18 q+ 16:00:43 What about security? For thrid-party templates? 16:00:52 s/thrid/third/ 16:02:28 Arnaud: Please explain why other W3C languages are "not designed for the Web"? 16:03:24 hknublau: Languages should have open vocabularies with defs associated with downloadable URIs 16:04:42 ack ericP 16:04:42 ericP, you wanted to ask, per iovka's q, is there a defn for valueShape? 16:04:55 -> http://w3c.github.io/data-shapes/shacl/#operation-validateNodeAgainstShape hasShape algorithm 16:05:46 ericP: discusses recursion and negation - does this proposal handle it correctly? 16:07:01 Arnaud: let's defer this general question and focus on clarification questions for Holger 16:07:37 q- 16:08:09 ack aryman 16:08:19 hknublau: I haven't made much use of recursion. I can put in a guard to prevent it. Can add negation. 16:08:50 aryman: if we allow 3rd party to provide javascript, that entails a security issue. 16:09:12 ... we need to have some trust mechanism for 3rd part extensions, or a sandbox mode 16:09:50 What about safety of third-party extensions? 16:10:20 hknublau: This is out of scope currently 16:10:38 Break for lunch 16:10:46 Resume at 1:00 PM 16:13:16 quit 16:13:34 iovka has left #shapes 16:18:33 elf-pavlik_ has joined #shapes 16:44:52 Labra has joined #shapes 16:47:55 michel has joined #shapes 16:54:52 kcoyle has joined #shapes 17:02:26 pfps has joined #shapes 17:02:27 scribenick: kcoyle 17:03:18 My presentation is https://www.w3.org/2014/data-shapes/wiki/Shacl-sparql-presentation 17:04:18 aryman has joined #shapes 17:04:21 similar to Holger's proposal; based on sparql, but is even closer to sparql 17:04:39 every constraint is translated to a single sparql query 17:05:25 standard sparql, no modifications 17:05:58 what is the presentation link? 17:06:08 http://w3c.github.io/data-shapes/shacl/#operation-validateNodeAgainstShape 17:06:12 shapes and classes are distinct; not related 17:06:35 https://www.w3.org/2014/data-shapes/wiki/Shacl-sparql-presentation 17:07:59 rdf vocab all mapped to sparql 17:08:20 could be used as less than full sparql engine 17:08:27 s/as/with 17:09:15 closed world constraints 17:09:52 +q 17:12:32 ack hknublau 17:13:10 holger: no relationship on shapes, constraints, classes: is classScope the same as in Holger's proposa.? 17:13:57 pfps: yes. you can say that constraints are classes, but classes and nodes and shapes are all separate; nothing operates over them 17:14:51 ... shouldn't create a node that is both a shape and a constraint, or class and constraint; but are not required to be disjoint 17:15:05 ... proposal is agnostic on their relationship 17:15:24 Arnaud: terminology: is this same as Holger? 17:15:48 pfps: no, different; constraint is the thing that you run and that has violations 17:16:10 ... a shape is like the thing in shex -- shapes are satisfied, don't have violations 17:16:54 ... conditions are the things that make the node to shape connection 17:17:21 ... constraint is combination of scope and shape 17:18:31 ... scope: can be a single node; all instance of a type; you can scope a condition to those nodes that satisfy a shape 17:20:04 ... shapes are either rdf-encoded or sparql 17:20:23 iovka has joined #shapes 17:20:55 ... rdf encoding doesn't (yet) cover all of sparql 17:21:25 ... shapes can refer to other shapes 17:21:39 ... any rdf can be added if there is a sparql translation 17:21:45 q+ 17:21:54 ack aryman 17:22:17 aryman: can you refer to shapes by name? would imply recursion 17:22:52 pfps: you can name shapes, but no recursion 17:23:01 ... aka, no cyclic references 17:24:21 +q 17:24:22 ... because cannot be done in a single translation to sparql 17:26:47 pfps: this is a constraints proposal, not shape recognition 17:27:34 aryman: gives example of test cases and defects 17:28:28 ack hknublau 17:28:36 recursive shapes are a syntax violation 17:29:39 hknublau: can combine class-based selection & shape based; if just have scope that is a shape can you insert sparql into constraint? if so, this only supports a subset of sparql 17:30:43 michel has joined #shapes 17:31:19 pfps: gives an answer in code, using ?THIS 17:32:20 ... Peter agrees with Holger, need to be careful with sparql as a scope; may not work 17:33:21 hknublau: problem is arbitrary sparql as a selector; peter agrees 17:34:24 Problem case, e.g. FILTER STRSTARTS(STR(?this), “http://example.org/ns#”) 17:36:47 proceeding with examples 17:37:14 UC1: the model is broken 17:39:03 4 examples 17:40:19 q+ 17:40:58 iovka has joined #shapes 17:41:01 classScope is not the norm in Linked Data APIs, i.e. validating PUT and POST requests 17:41:31 ack aryman 17:42:17 aryman: in linked data, classScope is not the norm 17:42:24 q+ to ask if the cycles (last block) is for cycles between a single pair of classes 17:43:15 (sorry for being late, technical problems ...) 17:43:47 UC2: enforcing cardinality - skip 17:43:53 ack ericP 17:43:53 ericP, you wanted to ask if the cycles (last block) is for cycles between a single pair of classes 17:44:28 pfps: requires rdfs closure; otherwise needs jplus sign 17:45:37 UC4: variations on a shape 17:46:03 pfps: something has a status that is either reported or varified 17:46:22 ... + two constraints that have shape scope, not class scope 17:48:01 ... if no type links, would say: everything whose status is reported has to have at least one (or two) reporters; once you have 17:48:17 ,,, shape scope you don't need type links 17:48:45 ... UC9 contract time intervals 17:50:48 ... UC23 schema.org constraints 17:52:31 ... shows transitivity as a violation; plus class scopes on Person -- uses some sparql for comparison of dates 17:53:41 ... common that you have to drop into sparql 17:54:18 UC33: validate medical procedure 17:54:45 pfps: bugs from shex primer, and choices example 17:59:15 +q 17:59:16 q+ 17:59:21 ack hknublau 17:59:47 q- 18:00:14 q+ 18:00:26 hknublau: make not be possible to create a single sparql query for all 18:02:31 pfps: this is about sub-optimal implementation of sparql; proposal could be implemented in other ways 18:03:04 ack aryman 18:03:45 aryman: single sparql query is semantics of constraint? 18:04:18 pfps: trans to single sparql query is a referent implementation 18:04:31 Problem cases with Sub-Selects (they cannot access any variables from the outside) 18:04:41 ... could be a problem with large constraints over large graphs 18:05:37 aryman: no error reporting mechanism ; in practice you would evaluate in chunks 18:06:18 pfps: this proposal does have error reporting; for high level language it's simple, but for raw sparql it is the sparql result 18:10:20 +q 18:11:15 q- 18:11:19 ack Labra 18:12:25 Labra: pfps proposal does not define high level language 18:12:43 Arnaud: how much is in the core? 18:13:18 Labra: Peter's proposal is more about how to constrain the language to be based on SPARQL than about what is in the language 18:13:19 pfps: so far no template mechanism; with templates, high level language is irrelevant; 18:13:44 ... high level language is just pointers to sparql translations 18:13:54 ... similar to Holger's proposal 18:15:05 TOPIC: next f2f meeting 18:15:23 ... where and when 18:15:58 +1 to lille 18:16:11 ... Iovka offered Lille (FR), not far from Paris 18:16:12 +1 to europe 18:16:38 ... when: cannot be same week as TPAC 18:16:53 WU vienna could host too 18:16:55 I'm not sure that I can travel to Europe until October due to budget limitations. 18:17:25 +1 for Lille 18:17:52 +1 Lille 18:18:02 you can also come through brussels 18:19:16 +1 Lille, -1 August, +1 mid- late-September 18:19:27 Europe has a better time zone for me than East Coast. 18:20:38 Arnaud: 8/19 September? 18:20:43 PROPOSED: next F2F in Lille on 8-10 September 18:21:04 +1 18:21:04 +1 18:21:04 +1 18:21:06 +1 for Europe but September will hard to commit from now (even for virtual participation) 18:21:12 +1 18:21:23 ah, won't be able to make that 18:21:39 what would work for you michel? 18:22:47 0 as budget limitations make it hard to travel now - I even had to turn down the DL workshop in Greece 18:23:34 PROPOSED: next F2F in Lille on 1-3 September? 18:23:48 scratch that 18:24:40 RESOLVED: next F2F in Lille on 8-10 September 18:24:44 I can't make it 18:24:55 in Australia 18:25:50 10-18 / 09 won't work for me so I might skip the last day Arnaud: ok, I can set up a poll to find out which week in september would work best 18:27:28 +1 for poll 18:27:43 s/RESOLVED/Not quite resolved/ 18:29:13 break for 15mn 18:46:04 kcoyle has joined #shapes 18:47:25 scribenick: kcoyle 18:48:08 Finished - deep dives; thanks to all 18:48:42 topic: Round of self-criticism 18:49:47 ericP: semantics are complicated; we don't have an elegant extensibility mechanism, just language refs 18:50:11 ... no templates in core language; 18:50:46 ... document status is "catching up"; document suite is bigger 18:51:54 ... prefixes and bases inherited from surrounding doc rather than literals 18:53:14 ... doesn't transform isomorphically to turtle 18:53:37 ... there are 'extra bits" 18:54:01 Arnaud: are there some user stories you would have trouble addressing? 18:54:38 ericP: we can always call out to extentions, so question is which can we meet in core language? we meet more than other languages but still not all 18:54:42 Huh? 18:54:44 q+ 18:54:48 q+ 18:54:51 ack pfps 18:55:05 pfps: larger than other proposals? 18:55:32 ericP: what you can do without sparql; our core language is larger than holger's, not sure re: peter's 18:56:08 ... decides, no, not larger than peter's 18:56:15 q+ 18:56:21 ack aryman 18:57:04 aryman: uneasy with recursion; value references can be deeply nested: are you confident you have clear semantics for that? 18:57:59 iovka: yes, i am confident because do not mix negation and recursion; disjunction does not cause a problem, but is harder to check 18:58:54 ... two kinds of disjunction - one of, some of 18:59:15 ack pfps 18:59:41 pfps: semantics are now more complicated - do you believe it is right? 19:02:25 iovka: yes, it is complicated; but I am confident that it works; now needs formal proof (can't really hear now - very soft) 19:03:49 Arnaud: all of the proposals have unfinished areas 19:04:30 what is the reason for the complexity (to Peter)? 19:05:00 next step is to write proofs and to provide a formal study of the computational complexity 19:05:18 pfps: very complex with many interacting parts; now multiple partial implementaiton semantics; hard to know what's going on 19:06:03 ... hard to have confidence that it can be fixed 19:07:40 iovka: allow negation onlhy on non-recursive shapes 19:08:36 ... what would be most helpful at this point? 19:09:12 pfps: dunno. 19:09:20 q+ 19:09:28 Arnaud: possible: test suite could help nail down corner cases 19:09:40 ack aryman 19:10:07 aryman: what level of rigor was applied to sparql? 19:13:22 hknublau: shacl is a chance to improve on spin; draft has gone through iterations, borrows from other approaches 19:14:10 ... weaknesses: reliance on prefixes from surrounding file, but RDF does not have API for prefixes 19:14:17 q+ 19:14:34 ... prefixes can be easily lost 19:14:45 OSLC defined an RDF voc for prefixes for use in a simple query syntax 19:14:48 ... need to be spelled out 19:15:33 ... needs pre-binding of variables as in Jena, but perhaps not supported by all dbs 19:16:30 ... reliance on sparql extensions ; open issue: inferencing - can we rely on it or not? 19:17:02 ... seems we can't rely on it, not always supported, can't activate programmatically. This doesn't affect core language 19:17:39 ... can users expect inferencing to be activated? have to tell engine which queries require inferencing 19:18:43