14:58:15 RRSAgent has joined #rdf-wg 14:58:15 logging to http://www.w3.org/2013/04/24-rdf-wg-irc 14:58:17 RRSAgent, make logs world 14:58:17 Zakim has joined #rdf-wg 14:58:19 Zakim, this will be 73394 14:58:19 ok, trackbot; I see SW_RDFWG()11:00AM scheduled to start in 2 minutes 14:58:20 Meeting: RDF Working Group Teleconference 14:58:20 Date: 24 April 2013 14:59:32 SW_RDFWG()11:00AM has now started 14:59:41 +GavinC 15:00:39 +Guus 15:00:42 AZ has joined #rdf-wg 15:00:43 +OpenLink_Software 15:00:50 pfps has joined #rdf-wg 15:00:58 zakim, who is here? 15:00:58 On the phone I see GavinC, Guus, OpenLink_Software 15:00:59 Zakim, OpenLink_Software is temporarily me 15:01:00 On IRC I see pfps, AZ, Zakim, RRSAgent, Guus, TallTed, gavinc, SteveH, yvesr, Arnaud, davidwood, gkellogg, cygri, manu, manu1, ericP, mischat, sandro, trackbot 15:01:00 +TallTed; got it 15:01:00 Zakim, mute me 15:01:01 TallTed should now be muted 15:01:26 +pfps 15:01:55 TallTed has changed the topic to: RDF WG -- current agenda: http://www.w3.org/2011/rdf-wg/wiki/Meetings:Telecon2013.04.24 15:02:16 +AZ 15:02:42 +??P16 15:02:47 zakim, I am ??P16 15:02:48 +gkellogg; got it 15:03:13 +Sandro 15:03:39 +1 to using Futures 15:04:13 patH has joined #rdf-wg 15:04:19 +[GVoice] 15:04:32 scribe: gavinc 15:04:37 Zakim, [GVoice] is me 15:04:37 +ericP; got it 15:04:43 gkellogg: we may need another last call for JSON-LD 15:04:55 +PatH 15:04:59 +Arnaud 15:05:13 -1 to that, if we can find any possible alternative. 15:05:18 gkellogg: only thing worth spending some time on, is perhaps WHATWG DOM spec as normative reference? 15:05:21 (and I think we can) 15:06:04 I may have to leave slightly early today. 15:06:09 Topic: Admin 15:06:26 zwu2 has joined #rdf-wg 15:06:40 zakim, who is here? 15:06:40 On the phone I see GavinC, Guus, TallTed (muted), pfps, AZ, gkellogg, Sandro, ericP, PatH, Arnaud (muted) 15:06:42 On IRC I see zwu2, patH, pfps, AZ, Zakim, RRSAgent, Guus, TallTed, gavinc, SteveH, yvesr, Arnaud, davidwood, gkellogg, cygri, manu, manu1, ericP, mischat, sandro, trackbot 15:07:00 PROPOSED: Accept minutes of the 17 April Telecon, https://www.w3.org/2013/meeting/rdf-wg/2013-04-17 15:07:03 minutes look fine 15:07:07 + +1.650.265.aaaa 15:07:13 RESOLVED: Accept minutes of the 17 April Telecon, https://www.w3.org/2013/meeting/rdf-wg/2013-04-17 15:07:17 zakim, +1.650.265.aaaa is me 15:07:18 +zwu2; got it 15:07:24 subtopic: Action item review 15:07:25 zakim, mute me 15:07:25 zwu2 should now be muted 15:07:51 Guus: Two by pfps, one by patH. Close them 15:08:03 CLOSE ACTION-252 15:08:03 Closed ACTION-252 Propose wording in concepts to reflect blank node semantics changes. 15:08:05 CLOSE ACTION-253 15:08:05 Closed ACTION-253 Propose changes for what Concepts says about ill-typed literals. 15:08:07 CLOSE ACTION-254 15:08:07 Closed ACTION-254 Propose solution for ISSUE-118. 15:08:39 zakim, who is on th ephone 15:08:39 I don't understand 'who is on th ephone', pfps 15:08:46 zakim, who is on the phone? 15:08:48 On the phone I see GavinC, Guus, TallTed (muted), pfps, AZ, gkellogg, Sandro, ericP, PatH, Arnaud (muted), zwu2 (muted) 15:08:58 Close ACTION-243 15:08:58 Closed ACTION-243 Provide license text, asking submitters of comments 21 and 28 if they are willing to share their submitted tests with no IP restrictions and under the proposed license.. 15:09:08 topic: Issues related to Concepts and Semantics 15:09:14 ISSUE-107? 15:09:14 ISSUE-107 -- Revised definition of blank nodes -- open 15:09:14 http://www.w3.org/2011/rdf-wg/track/issues/107 15:09:24 Zakim, who's muted? 15:09:24 I see TallTed, Arnaud, zwu2 muted 15:09:25 suptopic: Revised definition of blank nodes 15:09:45 pfps: Fixing up blank nodes, situation hasn't changed in a while. semantics is terse 15:09:53 ... concepts is very sparse on blank nodes as well 15:10:07 ... proposal is add just enough to cover RDF graphs sharing blank nodes 15:10:08 q+ 15:10:16 ... semantics refers to concepts saying they can happen 15:10:23 q+ 15:10:38 ... concepts has to say some stuff, they only can show up if the syntax says they do 15:10:39 +davidwood 15:10:51 there appear to be some wording issues -- the difference between "sharing blank nodes" and "sharing blank node identifiers" is significant 15:10:54 -ericP 15:11:12 ack AZ 15:11:32 AZ: Had some issues with the wording of the note pfps wants to add 15:11:39 +[GVoice] 15:11:50 AZ's message: http://lists.w3.org/Archives/Public/public-rdf-wg/2013Apr/0137.html 15:11:58 Souri has joined #rdf-wg 15:12:09 +Souri 15:13:06 AZ: RDF 2005 Does not define only merge, also says you can do union 15:13:17 ack AZ 15:13:24 pfps: I suppose that the documents... 15:13:26 s/2005/2004/ 15:13:42 patH: is AZ objecting to the email or the text? 15:13:53 ... sounds like the email 15:14:00 AZ: Yes. 15:14:27 pfps: I suppose AZ is technically correct, but I don't think that's how people read the documents 15:14:42 q? 15:14:51 ... in RDF 2004 if you combine two graphs your supposed to standardize apart any shared blank node 15:15:20 ... I think we now need to say that it's fine that RDF graphs can share blank nodes, and they will be the same blank node. And in other cases they won't share blank nodes. 15:15:33 the trouble with "underlying assumptions" is that they often don't get written into the spec ... because they're just "understood to be there" 15:15:34 AZ: This is based on your understanding of what people understand of the specification 15:15:58 ... there is a paragraph that describes union vs. merge. 15:16:50 Guus: AZ your mainly objecting to how we talk about RDF 2004, not the proposed text. 15:17:01 AZ: Yes. 15:17:34 +1 lets agree to on the new plan, and then figure out how to describe what the change is/was, if any. 15:17:47 AZ: The way to clarify bnode scropes, is have something that talks about the scope of bnodes. 15:17:51 +q 15:18:26 Guus: If I read pfps message, it says this is the required changes to RDF semantics, and RDF concepts. 15:18:27 http://www.w3.org/TR/sparql11-update/#def_datasetUnion 15:19:36 pfps: There is still an issue with the relationship between RDF 2004 and RDF 1.1 15:19:51 ... this isn't part of what's being decided today 15:20:16 Zakim, unmute me 15:20:16 TallTed should no longer be muted 15:21:14 -q 15:21:29 PROPOSED: Close ISSUE-107 as per http://lists.w3.org/Archives/Public/public-rdf-wg/2013Apr/0115.html, up until "Note:" (We're not characterizing the relationship to RDF 1.0) 15:21:44 +1 15:21:45 +1 15:21:46 +1 15:21:46 +1 15:21:46 +1 15:21:50 +1 15:21:52 +1 15:21:58 +1 15:22:03 +1 15:22:04 +1 15:22:08 RESOLVED: Close ISSUE-107 as per http://lists.w3.org/Archives/Public/public-rdf-wg/2013Apr/0115.html, up until "Note:" (We're not characterizing the relationship to RDF 1.0) 15:22:14 +1 15:22:17 +1 15:22:23 close issue-107 15:22:23 Closed ISSUE-107 Revised definition of blank nodes. 15:22:26 (then, to replace what's in the note, I would like to see, in Concepts, something about bnode [identifiers] scope) 15:22:28 ACTION pfps to edit Semantics per the previous proposal 15:22:28 Created ACTION-255 - Edit Semantics per the previous proposal [on Peter Patel-Schneider - due 2013-05-01]. 15:23:06 sandro: There is very clear evidence that graph nodes could be shared, and others who thought they couldn't 15:23:07 ok to say that there was an ambiguity (but not that it was assumed) 15:24:01 pat: Probably no one contemplated blank nodes could be coincidence by chance 15:24:43 ACTION gavinc link TriG to new text in RDF Concepts 15:24:43 Error finding 'gavinc'. You can review and register nicknames at . 15:24:52 ACTION gcarothe link TriG to new text in RDF Concepts 15:24:52 Created ACTION-256 - Link TriG to new text in RDF Concepts [on Gavin Carothers - due 2013-05-01]. 15:25:22 ACTION davidwood add resolution of issue-107 to RDF Concepts 15:25:23 Created ACTION-257 - Add resolution of issue-107 to RDF Concepts [on David Wood - due 2013-05-01]. 15:25:29 subtopic: What's the consequence of a literal being ill-typed? 15:25:39 ISSUE-109? 15:25:39 ISSUE-109 -- What's the consequence of a literal being ill-typed? -- open 15:25:39 http://www.w3.org/2011/rdf-wg/track/issues/109 15:25:52 Guus: Proposal from pfps http://lists.w3.org/Archives/Public/public-rdf-wg/2013Apr/0117.html 15:26:13 pfps: Illtyped literals do not denote, therefor an ill-typed literal is false 15:26:19 pfps: ill typed literals do not denote, therefor any triple including such a literal is false in all interpretations. 15:26:20 ... requires a change to concepts 15:27:34 Proposed: Close ISSUE-109 by making ill-typed literals not denote and requiring that RDF implementations accept ill-typed literals and produce RDF graphs from them, possibly producing a warning. 15:27:55 +1 15:27:56 PROPOSED: Close ISSUE-109 by making ill-typed literals not denote and requiring that RDF implementations accept ill-typed literals and produce RDF graphs from them, possibly producing a warning. http://lists.w3.org/Archives/Public/public-rdf-wg/2013Apr/0117.html 15:27:56 +1 15:27:56 +1 15:27:58 +1 15:27:58 +1 15:27:59 +1 15:28:01 +1 15:28:04 +1 15:28:04 +1 15:28:07 +1 15:28:17 +1 15:28:26 +1 15:28:34 RESOLVED: Close ISSUE-109 by making ill-typed literals not denote and requiring that RDF implementations accept ill-typed literals and produce RDF graphs from them, possibly producing a warning. http://lists.w3.org/Archives/Public/public-rdf-wg/2013Apr/0117.html 15:28:36 close issue-109 15:28:36 Closed ISSUE-109 What's the consequence of a literal being ill-typed?. 15:28:54 ACTION davidwood update RDF Concepts to reflect issue-109 15:28:54 Created ACTION-258 - Update RDF Concepts to reflect issue-109 [on David Wood - due 2013-05-01]. 15:29:21 subtopic: xsd:string 15:29:30 issue-126? 15:29:30 ISSUE-126 -- xsd:string -- open 15:29:30 http://www.w3.org/2011/rdf-wg/track/issues/126 15:30:03 pfps: xsd:string does not allow all unicode code points. it doesn't allow code point 0, ASCII null 15:30:21 ... technicaly this changes the status of literals that contain ASCII nulls 15:30:28 q+ 15:30:33 q- 15:30:57 ... any triple containing ASCII null, will now be ill typed and thus false. 15:31:10 "\u0000"^xsd:string is ill-typed 15:31:10 q+ 15:31:25 pfps: if your using ascii null, you should have your head exaimined 15:31:32 ack davidwood 15:31:36 davidwood: good catch, thanks pfps 15:31:41 "Currently a literal with type 15:31:41 xsd:string containing the #x0 character is an ill-typed literal, but is 15:31:41 syntactically permissable." 15:31:42 ... however ... 15:32:08 ... it isn't posible in any syntax 15:32:33 pfps: Should be said that it is an acceptable RDF Literal 15:32:59 davidwood: but it's not permissable in any syntax 15:33:20 pfps: can't pull nulls directly into Turtle, but can use escapes 15:33:21 in any case, it is syntactically permissable per the abstract syntax 15:33:24 tbaker has joined #rdf-wg 15:33:40 pfps: something that turns into a null character can't be put into xsd:string 15:33:44 http://www.w3.org/2013/TurtleTests/LITERAL1_all_controls.ttl has NULLs 15:34:21 q? 15:34:24 ack ericP 15:34:24 q+ 15:34:26 inaudible? 15:34:27 weak 15:34:45 Zakim, please dial ericP-mobile 15:34:45 ok, ericP; the call is being made 15:34:57 ack ericP 15:35:07 Zakim, please dial ericP-mobile 15:35:07 ok, ericP; the call is being made 15:35:35 Zakim, please dial ericP-mobile 15:35:35 ok, ericP; the call is being made 15:36:01 q- 15:36:12 pfps: http://www.w3.org/2013/TurtleTests/LITERAL1_all_controls.ttl produces a fine, but a semantics error 15:36:22 pfps: http://www.w3.org/2013/TurtleTests/LITERAL1_all_controls.ttl produces a fine RDF graph, but a semantics error 15:36:29 +EricP.a 15:36:32 http://www.w3.org/2013/TurtleTests/LITERAL1_all_controls.ttl is a semantic error, not a syntax error. 15:36:40 -ericP 15:36:56 sandro: turtle has to accept it with another data type! 15:37:18 How about "is syntactically permissable" goes to "is acceptable as an RDF literal" as a replacement? 15:37:27 Guus: would like to move on to 118? 15:38:01 davidwood: very few people are stupid enough to use ascii nulls in strings 15:38:05 -EricP.a 15:38:31 +EricP 15:39:26 does XML 5th Ed allow control characters like ctl-a and ctl-g ? 15:39:26 no 15:39:41 ericP: I acctually raised this back in April 15:40:16 ... this is great if you have a PNG or something else with 0s in it. 15:42:00 ... the issue is that plain literal could/was used for a blobs, not just for strings in the past. 15:42:36 ... these things also exist in SPARQL update. So SPARQL Update works with 0s in what were plain literals 15:42:47 sandro: You want a string of characters, and you also want a blob 15:42:49 +q 15:43:19 sandro: you often need two diffrent types, one for characters and one for blobs 15:43:26 ericP: They have to be characters 15:43:34 ack gavinc 15:43:46 The concept of "character" has morphed significantly since Unicode... 15:44:05 +q 15:44:19 ack gavinc 15:45:00 To quote from unicode.org, "The very term character is rather ambiguous" 15:45:02 gavin: +1 sandro, I don't think you want to treat the same datatype as both a bag of bits and a sequence of character. 15:45:23 gavin: Eg Python 15:45:31 gkellogg: and Ruby and Perl 15:45:44 ericP: I don't see how these apply. 15:45:55 davidwood: This is why the concept of a unicode code point was developed 15:46:10 +1 david 15:46:11 gavinc: All the functions that apply are completely different. Are you counting character or counting bytes...? 15:46:55 q+ 15:47:17 ericP: I don't believe any of these problems apply to us. The problem in these languages is because you didn't know if it was a byte or a character. In our situation, we can write each byte as a character. 15:47:24 ack pfps 15:47:31 gavin: No! 15:47:34 pfps: No! 15:47:46 pfps: Plain literals are unicode strings, not sequences of bits. 15:48:18 pfps: Each element is a unicode codepoint - it's binary representation is not stated. 15:48:33 eric: I have code points 0 to 255, it's a very natural encoding 15:48:43 eric: It works fine! 15:48:46 use case = naughty programmer. 15:50:11 pfps: little endian, etc, problems. 15:50:22 eric: hex encoding has the same problem 15:50:24 pfps: true 15:50:39 issue-118? 15:50:39 ISSUE-118 -- Simplifying datatype semantics -- open 15:50:39 http://www.w3.org/2011/rdf-wg/track/issues/118 15:51:24 pfps: One way forward requires no change. I'm happy for someone to argue the other way. 15:51:37 Guus, it's okay for us to stay on Zakim today. 15:51:45 People can write 0s! Just not in xsd:string 15:52:05 ericP: people will want to write zeros, in plain literals. 15:52:23 ericP: We make things hard for them by disallowing them. 15:52:44 pat: 2004 spec says plain literals were co-extant with strings. maybe nobody noticed that. 15:53:05 guus: amended proposal? 15:54:03 gavin: There is no problem. RDF Plain Literal was always defined in terms of char from XML. 15:54:19 eric: Trig, SPARQL, Turtle, allows zeros 15:54:31 gavin: Just like now -- they're allowed in the syntax, just not in RDF. 15:54:41 "A character is an atomic unit of text. Each character has a Universal Character Set (UCS) code point [ISO/IEC 10646] (or, equivalently, a Unicode code point [UNICODE]) that MUST match the Char production from XML [XML] thus ensuring compatibility with XML Schema Datatypes [XML Schema Datatypes]. Code points are sometimes represented in this document as U+ followed by a four-digit hexadecimal value of the code point. " 15:54:57 patH: It sounds like implementations have slack at the edges. That situation will probably continue. 15:55:31 sandro: Sounds like we need to add some RDF (in)Consistency Test(s). 15:56:03 pfps: replace "syntactically valid" with "acceptable RDF literal" 15:57:16 http://www.w3.org/TR/rdf-plain-literal/ 15:57:22 gavin: This is not a change. 15:57:42 rdf:PlainLiteral is not part of RDF Concepts 15:58:12 The first rec version of rdf:PlaiinLiteral is from 2009 15:58:23 sandro: That's not the spec we should be referring to. 15:58:35 http://www.w3.org/TR/rdf-concepts/ 15:59:10 "Plain literals have a lexical form and optionally a language tag as defined by [RFC-3066], normalized to lowercase." in Sec.6.5 of RDF Concepts 16:00:20 (sorry, "All literals have a lexical form being a Unicode [UNICODE] string, which SHOULD be in Normal Form C [NFC].") 16:00:21 note that i've run http://www.w3.org/2013/TurtleTests/manifest.ttl in situ 16:00:27 -EricP 16:00:56 PROPOSED: Accept http://lists.w3.org/Archives/Public/public-rdf-wg/2013Apr/0118.html, changing "is an ill-typed literal, but is syntactically permissable." to "is an ill-typed literal". 16:01:10 +1 16:01:18 +1 16:01:18 +1 16:01:20 +1 16:01:22 +1 16:01:23 +0 not thrilled, but I don't see any alternative 16:01:24 +1 16:01:26 +1 16:01:26 +1 16:01:34 +1 16:01:39 +0 16:01:40 +1, with the provisio that there may be some further proposal on how to handle plain literals 16:02:08 RESOLVED: Accept http://lists.w3.org/Archives/Public/public-rdf-wg/2013Apr/0118.html, changing "is an ill-typed literal, but is syntactically permissable." to "is an ill-typed literal". 16:03:01 issue-126? 16:03:01 ISSUE-126 -- xsd:string -- open 16:03:01 http://www.w3.org/2011/rdf-wg/track/issues/126 16:03:22 guus: This was ISSUE-126, now closed. 16:03:23 close issue-126 16:03:23 Closed ISSUE-126 xsd:string. 16:03:33 ACTION davidwood to modify RDF Concepts in respect to ISSUE-126, see http://lists.w3.org/Archives/Public/public-rdf-wg/2013Apr/0118.html 16:03:33 the minutes will need fixing to change the previous resolution from number 118 to number 126 16:03:33 Created ACTION-259 - Modify RDF Concepts in respect to ISSUE-126, see http://lists.w3.org/Archives/Public/public-rdf-wg/2013Apr/0118.html [on David Wood - due 2013-05-01]. 16:03:36 subtopic: issue-118 16:03:49 subtopic: issue-118 (get rid of datatype maps) 16:03:59 a partial mapping, complicated?! 16:04:12 pat: We can get rid of datatype maps by just making it part of the interpretation map 16:04:23 +1 16:04:36 pat: I made this change, but got some pushback. It seems editorial to me. 16:05:02 pat: I've suggested changes to the concepts document to bring it in line with this. 16:05:17 pat: I think it's simpler, and more directly intuitive. 16:05:23 pat: But I wont push on it hard. 16:05:24 q? 16:05:24 +1 to new language 16:05:31 +1 I love it 16:05:35 q? 16:05:46 yes I sent 16:05:48 AZ, response? 16:06:34 AZ: I'm not resisting this change. But I'm not sure that we need to make the change. 16:06:55 az: while I agree with what pat said, and I'm not strongly resisting this change, I'm reluctant to make a change where the positive motivation is not motivated by hard evidence. It does create a problem: how do you specify what a datatype IRI that is no XSD maps to? 16:07:35 q? 16:07:45 s/not resisting/not resisting strongly/ 16:07:56 pat: This isn't really an inconsistency between 2004 and 2013. 16:08:32 pat: I don't think inventing a mathematic thing called a datatype map actually solved any problem. 16:08:43 q? 16:08:50 q+ 16:09:18 ack sandro 16:09:22 pat: The only role the datatype map played was to require the interpretation to be the same at those points 16:09:22 q+ 16:09:54 sandro: I've never liked datatype maps. 16:09:57 ack AZ 16:10:40 az: The role of the datatype map is to say what datatype the IRI denotes. 16:10:48 sandro: (rant) 16:11:36 q+ 16:11:50 ack patH 16:12:07 sandro: in general when you're writing code that hardcodes the meaning of IRIs, ... that's true for datatypes or predicates or anything. 16:12:23 sandro and az disagree about what datatype maps and iris and interpretations work 16:12:59 pat: one less technical notion that people have to understand. the nonmath folks see one-more-mapping and groan (and maybe give up on RDF). 16:13:10 pat: and the equations boil down from 5 to 1. 16:13:21 semantics contain s much more complicated things than "a mapping" 16:13:55 pat: I'll be improving the exposition of Semantics. That might help with AZ's objection. 16:14:16 pat: eg partial mapping on a set is the same as a total mapping on a subset 16:14:38 guus: maybe a full proposal for issue-118 for next week? 16:14:47 guus: let's resolve it then, please. 16:14:54 pat: Good for me! 16:15:22 Turtle Test suite, it's out there... but we haven't resolved PrEfIX 16:15:25 guus: Any Other Business? Anything Urgent? 16:15:30 toddlered? 16:15:55 gavinc: Turtle Test Suite is up, but we need to resolve the Feature At Risk. 16:16:00 -PatH 16:16:23 .. I'm hesitant to ask for new implementation reports until we have decided on that feature. 16:16:41 .. ALSO, n-triples was published as a NOTE, which means it's DONE by has no normative requirements. 16:16:56 .. But the old n-triples was a REC. So that's weird. 16:17:11 sandro: We could rescind the old one? 16:17:29 sandro: since it's becoming obsolete. 16:17:41 We should also probably reference the N-Triples test suite. 16:18:12 AOB: who's monitoring the WG list? Martin Nally from IBM sent an email to the list on RDF/JSON but because he's not subscribed his email was put on hold, I would appreciate if whoever has control let his email through 16:18:31 action: gavin to gather data for resolving the turtle feature-at-risk. 16:18:31 Created ACTION-260 - Gather data for resolving the turtle feature-at-risk. [on Gavin Carothers - due 2013-05-01]. 16:18:52 -zwu2 16:19:01 -gkellogg 16:19:02 -Souri 16:19:08 -pfps 16:19:32 -AZ 16:19:33 Oh! ... and we need to split the XML and HTML datatypes out of RDF Concepts 16:19:43 gavinc, why? 16:19:55 We agreed to a few weeks ago ;) 16:20:14 why, gavinc? Because of W3C Dependency rules? That's no longer a good reason. 16:20:21 (things are changing) 16:20:29 DOM requirements, and lacking a stable specification to attach to HTML fragement parsing 16:20:48 we should talk about that more, gavinc, because pulling it out. 16:20:49 If we can update RDF Concepts as fragement parsing stablizes that's fine 16:20:53 -Guus 16:21:30 -Arnaud 16:21:35 -TallTed 16:21:44 davidwood, poke me if you get lost 16:21:55 Cheers, gavinc 16:21:56 hg clone https://dvcs.w3.org/hg/rdf/ 16:22:15 -Sandro 16:22:16 -davidwood 16:22:18 -GavinC 16:22:18 SW_RDFWG()11:00AM has ended 16:22:18 Attendees were GavinC, Guus, TallTed, pfps, AZ, gkellogg, Sandro, ericP, PatH, Arnaud, zwu2, davidwood, Souri 16:22:45 Yeah, I don't care about it being seprate. I just want to make sure that rdf:HTML is implentable :D 16:22:55 It isn't exactly at the moment 16:23:01 at least not the way anyone would expect 16:23:09 ... why? 16:23:36 the fragment parsing part of HTML5 will happily create tags around your fragment :D 16:23:54 which is likely to be surprising 16:23:59 Ah. 16:24:12 the