15:49:06 RRSAgent has joined #rdf-star 15:49:10 logging to https://www.w3.org/2024/10/31-rdf-star-irc 15:49:10 Zakim has joined #rdf-star 15:49:30 meeting: RDF-star WG biweekly focused meeting 15:49:34 agenda: https://www.w3.org/events/meetings/e7387703-d1cb-4215-a7e1-eec9b4af24d4/20241031T120000/ 15:49:34 clear agenda 15:49:34 agenda+ potential WG note explaining what we did -> 1 https://github.com/w3c/rdf-star-wg/issues/124 15:49:34 agenda+ Drop the requirement to support ill-typed literals with recognized datatype IRIs -> 2 https://github.com/w3c/rdf-concepts/issues/60 15:49:35 agenda+ Move RDF Dataset Merge to RDF Semantics or RDF Concepts -> 3 https://github.com/w3c/sparql-query/issues/155 15:49:36 agenda+ Additional "needs discussion" issues -> 4 https://github.com/orgs/w3c/projects/20/views/6 15:49:39 agenda+ Any Other Business (AOB), time permitting 15:49:44 RRSAgent, draft minutes 15:49:45 I have made the request to generate https://www.w3.org/2024/10/31-rdf-star-minutes.html TallTed 15:49:48 RRSAgent, make logs public 15:50:32 previous meeting: https://www.w3.org/2024/10/25-rdf-star-minutes.html 15:50:34 next meeting: https://www.w3.org/2024/11/01-rdf-star-minutes.html 15:55:50 AZ has joined #rdf-star 15:57:59 present+ 15:58:04 Chair: ktk 15:58:13 present+ 15:58:24 AZ: is scribing working for you today? 15:58:38 yep 15:58:42 great tnx 15:58:46 tl has joined #rdf-star 15:59:02 scribe+ 15:59:33 draggett has joined #rdf-star 15:59:33 s/scribe+/scribe: AZ 15:59:33 RRSAgent, draft minutes 15:59:35 I have made the request to generate https://www.w3.org/2024/10/31-rdf-star-minutes.html TallTed 15:59:49 present+ 16:00:22 present+ 16:01:02 s/AZ: is scribing working for you today?// 16:01:02 s/yep// 16:01:02 s/great tnx// 16:01:08 RRSAgent, draft minutes 16:01:09 I have made the request to generate https://www.w3.org/2024/10/31-rdf-star-minutes.html TallTed 16:01:16 present+ 16:01:48 Dominik_T has joined #rdf-star 16:02:00 present+ 16:02:06 regrets+ Ora 16:02:08 present+ 16:02:09 present+ 16:02:11 pfps has joined #rdf-star 16:02:25 present+ 16:02:30 niklasl has joined #rdf-star 16:02:39 enrico has joined #rdf-star 16:02:44 present+ 16:03:06 present+ 16:03:36 ktk: Ora has a meeting so I'm chairing 16:03:45 Zakim, open item 1 16:03:46 agendum 1 -- potential WG note explaining what we did -> 1 https://github.com/w3c/rdf-star-wg/issues/124 -- taken up [from agendabot] 16:03:55 gkellogg has joined #rdf-star 16:04:31 pfps: most WG do something along this line 16:04:40 present+ 16:04:45 ... we haven't yet decided to do it but I think it would be a good idea 16:05:17 ... I just noticed there is a "what's new" deliverable 16:05:32 4 possible places New, primer, concepts, note - depends on the ambition - e.g. patterns and anti-patterns 16:05:53 https://w3c.github.io/rdf-new/spec/ 16:06:12 ... In fact there is one (see URL above) 16:06:37 ... we'll do it 16:06:43 q? 16:07:01 ktk: we still have to see who can work on this 16:07:10 Zakim, next item 16:07:10 agendum 2 -- Drop the requirement to support ill-typed literals with recognized datatype IRIs -> 2 https://github.com/w3c/rdf-concepts/issues/60 -- taken up [from agendabot] 16:07:42 james has joined #rdf-star 16:07:52 q+ 16:08:12 q+ 16:08:23 ack pfps 16:08:24 pfps: I agree with what Andy says in the issue 16:08:35 ... the wording should change from MUST to SHOULD 16:08:48 Souri has joined #rdf-star 16:08:57 present+ 16:09:12 AndyS: it depends what "support" really means here 16:09:16 present+ 16:09:42 ack AndyS 16:09:51 ... I don't think ill typed literal making the whole graph iinvalid is very useful 16:09:58 scribe+ 16:10:30 AZ: I also want to ask what is ment by "support". If you have a system that does not recognize a datatype IRI 16:10:41 ... if you want to move that to another triplestore, you might loose something. 16:11:02 ... I'm not sure what support means. It should pass as syntacticly correct. 16:11:13 q+ 16:11:13 q+ 16:11:16 s/loose something/lose something/ 16:12:20 ... By the semantics of illtyped literals, since RDF 1.1, if you have an ill-typed literal in a graph, it makes the graph inconsistent, unsatisfiable. 16:12:22 RDF concepts -- "The list of datatypes supported by an implementation is determined by its recognized datatype IRIs." seems to be the nearest to defining "support". 16:12:56 ... If you say this kind of graphs may not be supported, what about other kinds of inconsistencies. Should any such graph not be supported? 16:13:15 ... I'm not sure if I agree with this proposal. 16:13:17 scribe- 16:13:24 ack AZ 16:13:31 ack pfps 16:14:02 pfps: one option would be to tweak the wording 16:14:05 I have made the request to generate https://www.w3.org/2024/10/31-rdf-star-minutes.html TallTed 16:14:47 ack gkellogg 16:15:18 One option is that implementations MUST accept input documents with ill-typed literals and SHOULD include the resultant triple in the RDF graph. 16:15:19 gkellogg: it makes no sense to talk about an ill-type literal for non-recognised datatypes 16:15:30 ... it all depends on what "support" means 16:15:44 q+ 16:16:04 That is - parsing MUST NOT stop at an ill-typed literal but the system MAY choose to not include the triple in the resultant graph. 16:16:06 ... I think the idea is to be able to only retain well-typed literals 16:16:24 I would add that if an implementation drops the triple then it MUST produce a warning. 16:16:44 q+ 16:16:44 ... it would be reasonable for RDF systems to not deal with ill-typed literals 16:16:45 ack TallTed 16:17:07 TallTed: the current text is "MUST accept", not "MUST support" 16:17:33 ... "accept" means it can evolve 16:17:47 ... triplestores should be able to take any literals 16:18:18 ... but then it may deal with the literal for some processes adequatel 16:18:49 q+ 16:18:50 q+ 16:18:54 ... you can do almost anything with RDF and unless there is a strong argument against that, we should keep it like this 16:19:21 ktk: how are different implementations dealing with this? 16:19:24 ack ktk 16:19:58 q+ 16:20:20 AndyS: in SPARQL, there are cases when you need to assign a value, so it does not work with ill-typed literals but that a SPARQL process 16:20:26 ack AndyS 16:20:42 ... there could be wording to make this a little more flexible with "MAY" 16:20:53 ... it's difficult to make it a "MUST" 16:20:59 agreed that it is difficult to require a warning 16:21:22 james: We are very accepting (in our implem) and it has been very useful 16:21:43 ack james 16:21:47 ... I think it should be a "MUST" for reasons of interoperability 16:21:54 ack Souri 16:21:55 "SHOULD accept" -- MUST for warnings is a bit strange. We don't have a "warning" mechanism in the specs. 16:21:57 ... but it's personal opinion 16:22:42 Souri: when we find an illtyped literal, we separate it 16:23:05 ... we continue even if we find error and they get reported 16:23:27 q+ 16:23:27 ... we do not accept it in that form, so for us, a MUST would not work 16:23:28 q+ 16:23:57 AndyS: choosing the datatypes you choose to handle is something you do when you use the data 16:24:04 ack AndyS 16:24:07 ... at loading time, you may not have decided 16:24:43 TallTed: I'm concerned to hear that some implementations are not conformant 16:25:00 q+ 16:25:03 q+ 16:25:05 ack TallTed 16:25:10 ... It's blocking evolution, because there may be new datatypes supporting in the future 16:25:39 q+ 16:25:45 ... The reasoning I see is that the proposal is done because there are implementations that are not conformant 16:26:12 +1 for evolution (with the caveat that I prefer opt-in "drop unrecognized" modes to avoid sending inexplicable data onward). 16:26:14 q+ to say that implementations that reject unrecognized datatypes are broken but ones that do not fully accept known ill-typed literals are not so bad 16:26:19 ack Souri 16:26:42 q- because Souri covered my poin 16:26:58 Souri: if we have an xsd:integer with "abc" lexical form, we don't accept it, but if you have ex:mytype, we don't do anything 16:27:22 q? 16:27:35 ... we report the problem and users can decide what to do with this problem 16:27:38 ack pfps 16:27:38 pfps, you wanted to say that implementations that reject unrecognized datatypes are broken but ones that do not fully accept known ill-typed literals are not so bad 16:27:41 q? 16:28:17 ack james 16:28:38 james: we do 2 kind of things, one on the values to do efficient operation, and one that just take any literal transparently 16:29:08 ... in the past, we did not do anuything with time, then it evolved to handle it appropriately 16:29:28 q? 16:29:54 ktk: what do we do with this issue? We don't really have a conclusion 16:29:56 "MUST accept" is current text 16:30:26 q+ 16:30:40 ack Souri 16:31:13 Souri: in Oracle, we don't want to have, e.g., 31st February, so we reject it 16:31:24 q+ 16:31:48 ... we do not hide it, we report it 16:32:01 q+ 16:32:26 ... I would not like to have "MUST accept" 16:32:43 ack TallTed 16:32:43 q+ 16:33:18 TallTed: not accepting data is bad but you can handle the ill-typed literals after they are loaded 16:33:57 q+ 16:34:22 ... in the future, there could be a change that makes a lexical form acceptable 16:34:56 ack tl 16:35:25 tl: I like the idea that there are several phases, 1st you parse and put in store, then other processes 16:35:48 ... then the user can be informed of problematic literals 16:36:14 ... you would get an error if you use reasoner on the data 16:37:03 AndyS: I find the use cases of rejecting or not rejecting both reasonable 16:37:28 ... the problem is when an entire graph is rejected 16:37:35 ack AndyS 16:38:16 My pref is change "MUST accept" to "SHOULD accept". All the described handling cases seem reasonable for their different cases. 16:38:17 ack Souri 16:38:19 Souri: w do not reject entire graph, just the triples with ill-typed literals 16:38:47 +1 for SHOULD 16:38:49 ... the earlier the problems can be pointed out the better 16:39:07 ... customers are also happy with this 16:40:09 Strawpoll: "Implementations MUST accept ill-typed literals" gets changed to "Implementations SHOULD accept ill-typed literals" 16:40:15 +1 16:40:16 +1 16:40:19 +1 16:40:20 +1 16:40:21 +1 16:40:26 +1 16:40:26 +1 16:40:33 -0.3141592 16:40:33 0 16:40:44 -0.5 16:40:52 0 16:40:53 q+ 16:40:58 +0.5 (I might prefer some "SHOULD by default, MUST if asked to accept"...) 16:41:32 q+ 16:41:33 q+ 16:41:38 ack TallTed 16:41:42 TallTed: if we make this change, we have to be really clear how errors are dealt wit 16:41:49 s/wit/with/ 16:42:18 AndyS: I don't think we should go into how errors and warnings are handled 16:42:35 An ill typed literal is not a syntax error. 16:42:36 ack AndyS 16:42:59 An ill typed literal conforms to syntax. 16:43:03 +1 to AndyS 16:43:04 ... there's an historical example (??) where specs mentioned what to do with errors and it took a large space, and was eventually dropped 16:43:10 ack niklasl 16:43:58 niklasl: I had experienced cases of systems that reject things that I would have like be accepted because things evolved 16:44:14 ... although I'm sympathetic to the arguments (thus my +0.5 vote) 16:44:33 ... it could be something that users can opt-in or out 16:44:54 q? 16:45:11 ktk: there could be a note that explain what pitfalls etc occur and how to deal with them 16:45:22 Zakim, next item 16:45:22 agendum 3 -- Move RDF Dataset Merge to RDF Semantics or RDF Concepts -> 3 https://github.com/w3c/sparql-query/issues/155 -- taken up [from agendabot] 16:45:26 ... we can continue this offline 16:46:12 gkellogg: not a strong opinion, but RDF Semantics seams the right place for this definition 16:46:52 ... this is currently defined in SPARQL Query 16:46:58 https://github.com/w3c/sparql-query/pull/152#discussion_r1735203153 16:46:59 https://github.com/w3c/sparql-query/pull/152 -> Pull Request 152 Use RDF Dataset definition from RDF Concepts (by rubensworks) [spec:substantive] 16:47:11 q+ 16:47:35 pfps: I don't see a problem about putting it in RDF semantics 16:47:46 ack pfps 16:48:00 ... it makes sense to have a section on dataaset merging in smenatics 16:48:05 +1 to pfps (add to section 10) 16:48:12 s/dataaset/dataset/ 16:48:23 s/smenatics/semantics/ 16:48:33 q? 16:48:37 pfps: graph merge is in semantics 16:48:48 +1 to putting Dataset Merge in same document as Graph Merge 16:48:53 https://www.w3.org/TR/sparql12-query/#sparqlDataset "Definition: RDF Dataset Merge" 16:48:57 it would likely go in section 10. 16:49:32 https://www.w3.org/TR/sparql12-query/#sparqlDataset 16:49:44 pfps: don't close the issue until we have a proposed content 16:49:59 Document RDF semantics section 4.1Shared blank nodes, unions and merges 16:50:23 ktk: 5 min left, let's jump to AOB 16:50:34 s/4.1Shared/4.1 Shared/ 16:51:09 enrico: I want to look at the writing in RDF 1.2 Semantics 16:51:31 ... what's the process for updating the doc 16:52:36 ktk: Semantic TF tomorrow is also 1 hour earlier for the Europeans 16:52:53 RRSAgent, draft minutes 16:52:55 I have made the request to generate https://www.w3.org/2024/10/31-rdf-star-minutes.html TallTed 16:57:08 s/anuything/anything/ 16:57:14 s/iinvalid/invalid/ 16:57:23 s/illtyped/ill-typed/ 16:57:33 s/smenatics/semantics/ 16:57:37 s/syntacticly/syntactically/ 16:57:53 s/seams/seems/ 16:58:05 RRSAgent, draft minutes 16:58:07 I have made the request to generate https://www.w3.org/2024/10/31-rdf-star-minutes.html ktk 16:59:47 Zakim, leave 16:59:47 leaving. As of this point the attendees have been ktk, AZ, tl, draggett, TallTed, Dominik_T, AndyS, gtw, pfps, enrico, niklasl, gkellogg, Souri, james 16:59:47 Zakim has left #rdf-star 16:59:53 RRSAgent, leave 16:59:53 I see no action items