Chatlog 2012-09-06

From RDFa Working Group Wiki
Jump to: navigation, search

See CommonScribe Control Panel, original RRSAgent log and preview nicely formatted version.

13:18:17 <RRSAgent> RRSAgent has joined #rdfa
13:18:17 <RRSAgent> logging to http://www.w3.org/2012/09/06-rdfa-irc
13:18:19 <trackbot> RRSAgent, make logs world
13:18:19 <Zakim> Zakim has joined #rdfa
13:18:21 <trackbot> Zakim, this will be 7332
13:18:21 <Zakim> ok, trackbot; I see SW_RDFa()10:00AM scheduled to start in 42 minutes
13:18:22 <trackbot> Meeting: RDF Web Applications Working Group Teleconference
13:18:22 <trackbot> Date: 06 September 2012
13:32:43 <MacTed> MacTed has joined #rdfa
13:46:17 <ShaneM> ShaneM has joined #rdfa
14:00:15 <ivan> zakim, dial ivan-voip
14:00:18 <Zakim> ok, ivan; the call is being made
14:00:23 <Zakim> SW_RDFa()10:00AM has now started
14:00:26 <Zakim> +Ivan
14:00:43 <niklasl> niklasl has joined #rdfa
14:01:00 <Zakim> +scor
14:01:24 <Zakim> +??P36
14:01:27 <niklasl> zakim, I am ??P36
14:01:32 <Zakim> +niklasl; got it
14:01:55 <Zakim> +??P0
14:01:57 <manu1> zakim, I am ??P0
14:01:57 <Zakim> +manu1; got it
14:02:44 <Zakim> +[OpenLink]
14:02:51 <MacTed> Zakim, [OpenLink] is temporarily me
14:02:52 <manu1> zakim, who is on the phone?
14:02:52 <MacTed> Zakim, mute me
14:02:56 <scor> scor has joined #rdfa
14:03:08 <Zakim> +MacTed; got it
14:03:12 <Zakim> On the phone I see Ivan, scor, niklasl, manu1, MacTed
14:03:14 <Zakim> MacTed should now be muted
14:04:25 <ivan> scribenick: ivan
14:04:29 <ivan> scribe: Ivan
14:04:32 <ivan> chair: Manu
14:04:34 <manu1> Agenda: http://lists.w3.org/Archives/Public/public-rdfa-wg/2012Sep/0000.html
14:05:05 <ivan> q+
14:05:37 <manu1> Topic: FPWD for HTML+RDFa 1.1
14:05:41 <scor> q+ (after FPWD)
14:05:44 <Zakim> +McCarron
14:05:47 <ivan> ivan: new charter is approved
14:06:24 <manu1> ivan: Since this is a new charter with new deliverables, with patent policy issues - everyone is required to rejoin the group within 45 days.
14:06:26 <manu1> ack ivan
14:06:50 <manu1> ivan: It's an administrative issue, but we need to do this.
14:07:53 <manu1> New draft for HTML+RDFa that is ready for FPWD: http://www.w3.org/2010/02/rdfa/drafts/2012/WD-rdfa-in-html-20120911/
14:09:14 <manu1> ivan: I need to look at the status section, I'll do that later today.
14:14:32 <manu1> ivan: Well, I just looked at the status of the document, it's fine as-is. So, next step is to do a formal transition request.
14:14:48 <manu1> ivan: I already talked with Thomas, we can use the same short name, so that's clear.
14:16:51 <manu1> PROPOSAL: Publish HTML+RDFa 1.1 ( http://www.w3.org/2010/02/rdfa/drafts/2012/WD-rdfa-in-html-20120911/ ) as a First Public Working Draft on Tuesday, September 11th 2012.
14:16:54 <ivan> +1
14:16:55 <ShaneM> +1
14:16:56 <manu1> +1
14:16:57 <niklasl> +1
14:17:11 <scor> +1
14:17:14 <MacTed> +1
14:17:16 <ivan> RESOLVED:  Publish HTML+RDFa 1.1 ( http://www.w3.org/2010/02/rdfa/drafts/2012/WD-rdfa-in-html-20120911/ ) as a First Public Working Draft on Tuesday, September 11th 2012.
14:17:39 <ivan> ISSUE-140?
14:17:39 <trackbot> ISSUE-140 -- How is the OBJECT element processed when containing the @data attribute? -- pending review
14:17:39 <trackbot> http://www.w3.org/2010/02/rdfa/track/issues/140
14:17:41 <manu1> Topic: ISSUE-140: OBJECT and @data attribute
14:17:48 <manu1> http://www.w3.org/2010/02/rdfa/track/issues/140
14:18:06 <ivan> manu1: whether or not we should process @data on the object element
14:18:11 <ivan> … if yes, what are the results
14:18:19 <ivan> … gregg did an overview
14:18:35 <ivan> … he said last year we resolved
14:18:46 <ivan> …. not to support the attribute
14:18:54 <ivan> … the test manifest comments this one out
14:18:59 <ivan> q+
14:19:12 <manu1> ack ivan
14:20:17 <ivan> ivan: is the @data/object still out?
14:20:30 <ivan> manu: the idea was to use different media elements in the browsers
14:20:46 <ivan> … today we have audio and video elements, those are the attributes that take over the object
14:20:52 <ivan> … it is still used for plugins
14:21:00 <ivan> … but with the big push against plugins
14:21:12 <ivan> … then I think that still holds
14:21:23 <ivan> … so html5 is getting rid of objects this way
14:21:31 <niklasl> q+
14:21:33 <ivan> … anyone knows anything else?
14:21:35 <manu1> ivan: The argument that we used to not support @data - the usage of the @data attribute and OBJECT element is on its way out - argument made by you, Manu.
14:21:41 <manu1> ack niklasl
14:21:42 <ivan> ack niklasl 
14:22:05 <ivan> niklasl: I think it makes sense, the @data attribute was around for a long time and we did not need it in xhtml1.1 either
14:22:11 <ivan> +1 to niklasl 
14:22:24 <manu1> PROPOSAL: Do not support the @data attribute in HTML+RDFa 1.1.
14:22:25 <scor> +1
14:22:27 <ivan> +1
14:22:28 <manu1> +1
14:22:29 <niklasl> +1
14:22:31 <MacTed> +1
14:22:32 <ShaneM> +1
14:22:34 <scor> +1
14:22:38 <ivan> RESOLVED: Do not support the @data attribute in HTML+RDFa 1.1.
14:22:49 <trackbot> ISSUE-140 How is the OBJECT element processed when containing the @data attribute? closed
14:22:56 <ivan> issue-141?
14:22:56 <trackbot> ISSUE-141 -- How many of the possible datatypes for @datetime should be supported? -- open
14:22:56 <trackbot> http://www.w3.org/2010/02/rdfa/track/issues/141
14:23:14 <manu1> Topic: ISSUE-141: Datatypes for @datetime
14:23:44 <Zakim> -McCarron
14:23:49 <manu1> http://www.w3.org/2010/02/rdfa/track/issues/141
14:24:06 <ivan> manu: these are the time related types that we support
14:24:18 <ivan> … gregg mentioned a number of resolutions
14:24:40 <ivan> -> http://lists.w3.org/Archives/Public/public-rdfa-wg/2012Aug/0002.html Gregg's analysis and list of types
14:25:04 <ivan> … we also resolved that the text content in the absence of a day-time attribute is used
14:25:13 <ivan> … the test suite has all the necessary tests already
14:25:20 <ivan> … it is consistent with html5
14:25:40 <ivan> … the suite also looks also at the inner text of the elements
14:25:50 <ivan> … 3 out of 4 passes these tests, too
14:26:05 <ivan> … the issue seems to be resolved
14:26:12 <ivan> … any other input on this?
14:26:27 <ivan> (silence)
14:27:04 <manu1> PROPOSAL: Support processing the @datetime attribute and the contents of the TIME element. Support the following datatypes: xsd:date, xsd:time, xsd:dateTime, xsd:duration, xsd:gYear, and xsd:gYearMonth.
14:27:12 <manu1> +1
14:27:13 <ivan> +1
14:27:14 <niklasl> +1
14:27:17 <scor> +1
14:27:31 <MacTed> +1
14:27:40 <ivan> RESOLVED: Support processing the @datetime attribute and the contents of the TIME element. Support the following datatypes: xsd:date, xsd:time, xsd:dateTime, xsd:duration, xsd:gYear, and xsd:gYearMonth.
14:27:53 <manu1> gkellogg: +1
14:28:13 <ivan> ISSUE-137?
14:28:13 <trackbot> ISSUE-137 -- HTML+RDFa should normatively declare media types and describe how to identify relative to XHTML+RDFa 1.1 -- open
14:28:13 <trackbot> http://www.w3.org/2010/02/rdfa/track/issues/137
14:28:32 <Zakim> +McCarron
14:28:45 <trackbot> ISSUE-137 -- HTML+RDFa should normatively declare media types and describe how to identify relative to XHTML+RDFa 1.1 -- open
14:28:45 <trackbot> http://www.w3.org/2010/02/rdfa/track/issues/137
14:28:50 <manu1> Topic: ISSUE-137: Normative media types for HTML+RDFa
14:29:00 <manu1> http://www.w3.org/2010/02/rdfa/track/issues/137
14:29:52 <ivan> manu: shane made a quick review, the current text already says that html+rdfa1 should be labeled text/html
14:30:11 <ivan> … there is also a text on the usage of @version (even if it is non comforting per html5)
14:30:48 <ivan> … processors must read the content of @version and use it to switch; if the value is not correct then the latest version of rdfa1.1 should be used
14:30:58 <ivan> … a non-normative text should be added:
14:31:19 <ivan> "Note: Some processors may not be able to detect the media type of the 
14:31:19 <ivan> document being processed because of system limitations.  In these cases 
14:31:20 <ivan> the default processing rules in [RDFA-CORE] [2] section 4.1 - RDFa 
14:31:21 <ivan> Processor Conformance - take precedence."
14:32:08 <ivan> manu: the problem is that this may not answer Alex's concern
14:32:18 <ivan> … but the previous texts may answer
14:32:36 <ivan> ... the spec says something about non-xml mode
14:32:44 <ivan> … does it say something about xhtml5?
14:32:57 <ivan> (manu looks up spec text)
14:33:06 <ivan> shane: I do not think there is a text there
14:33:21 <ivan> … I wanted to avoid normative text
14:33:42 <ivan> manu: do we actually state this in rdfa1.1 core
14:34:02 <ivan> ivan: I do not think so
14:34:34 <ivan> manu: going back to the original issue
14:34:48 <ivan> … the spec specifies for text/html
14:35:03 <ivan> … but there is no mention for application/xhtml+xml
14:35:14 <ivan> … alex proposes a text:
14:35:23 <ivan> "HTML+RDFa documents should be labeled with Internet Media Types "text/html" or "application/xhtml+xml" as defined in [RFC3236]."
14:35:37 <ivan> q+
14:36:10 <ivan> manu: the problem is with xhtml 
14:36:28 <ivan> … can we say something about the choice xhtml1.1 and xhtml5
14:36:42 <ivan> shane: core says you must use a media type
14:36:59 <ivan> … if you cannot use a media type then you can do something else
14:37:10 <niklasl> from rdfa 1.1. core: A conforming RDFa Processor may use additional mechanisms (e.g., the DOCTYPE, a file extension, the root element, an overriding user-defined parameter) to attempt to determine the Host Language if the media type is unavailable. These mechanisms are unspecified.
14:37:13 <manu1> ack ivan
14:38:33 <manu1> ivan: RDFa 1.1 Core does say what Niklas' put in IRC. I can understand Alex's issue because I've hit this issue myself. What I do right now is I try to look at the DOCTYPE of the XHTML5 document, if the doctype is one of the DTDs that we have specified for RDFa 1.0 or RDFa 1.1, then I fall back on that type of document, otherwise I do XHTML5. The @version attribute is a different question...
14:38:35 <manu1> ...here, it may switch between RDFa 1.0 and RDFa 1.1.
14:39:43 <manu1> ivan: However, that's not what we're talking about... we're talking about what the host language is. The text that Alex proposes is fine, but it's not enough. 
14:41:51 <manu1> ivan: for XHTML - first, we need to look at the Media Type and see that it's 'application/xhtml+xml'. Second, we look at the DTD, if there is one - if it's one we have defined for XHTML1+RDFa 1.1 or XHTML1+RDFa 1.0, we use that. If that does not match, it is XHTML5.
14:42:04 <manu1> ivan: I think that's an unambiguous algorithm, that's what we should say.
14:43:04 <ivan> shane: the core spec is consistent with this
14:43:30 <ivan> … this extra detail sounds normative to me
14:43:41 <ivan> … would it change the behavior of an existing document?
14:44:05 <ivan> q+
14:44:41 <manu1> ack ivan
14:45:54 <manu1> ivan: I can live with the rules that are above - if you have a document that uses the correct DTD, there won't be any change for those documents.
14:46:03 <manu1> ivan: We are fine as far as that's concerned.
14:47:49 <niklasl> q+
14:48:32 <manu1> shane: What happens when a badly authored XHTML+RDFa 1.1 document (without a doctype or @version) that is using XHTML1 terms gets processed with these new rules?
14:48:40 <manu1> manu: The terms are dropped.
14:48:45 <manu1> shane: Yes, that's my concern.
14:49:06 <manu1> ivan: Well, the sad truth is that this is not defined behavior today, I spent some time figuring out how to do this correctly.
14:49:40 <MacTed> Zakim, unmute me
14:49:40 <Zakim> MacTed should no longer be muted
14:49:41 <manu1> shane: As long as we're doing this with out eyes open, I'm okay with it. These badly authored documents will have their terms dropped.
14:50:04 <manu1> ack niklasl
14:50:04 <MacTed> Zakim, mute me
14:50:05 <Zakim> MacTed should now be muted
14:50:06 <ivan> ted: if you were sloppy generating the document, that is what you get, so do not be sloppy...
14:50:16 <ivan> niklasl: that sounds adequate to me as well
14:51:54 <ivan> niklasl: what about using xmlns as a switch?
14:52:03 <ivan> ivan: that is legal in xhml5
14:52:20 <ivan> shane: what about @version attribute
14:52:27 <ivan> … that has defined values for our processor
14:53:20 <ivan> manu: that would allow authors to use a non-dtd based mechanism for switching
14:53:41 <ivan> … but we are getting into rathole a.k.a. sniffing
14:53:54 <ivan> … would anyone object to use ivan's attribute?
14:54:03 <ivan> s/attribute/algorithm/
14:54:13 <ivan> manu1: we cannot use the @version, 
14:54:44 <ivan> … if we put in value for xhtml5+rdfa.1.1 for the value, that would lead to problems with the html wg, because we'd add a new attribute
14:55:03 <ivan> … if we refer to the value referring to xhtml1.1
14:55:45 <ivan> shane: you suggest that we do not have the option to define a string for the @version attribute to define xhtml5
14:56:14 <ivan> … I always thought we could do that...
14:56:45 <ivan> manu: I wonder whether this is an issue for javascript processors at all
14:56:56 <ivan> … is there a difference between that
14:56:57 <niklasl> .. according to this, js can get at the doctype: http://stackoverflow.com/questions/6088972/get-doctype-of-an-html-as-string-with-javascript
14:57:16 <ivan> q+
14:57:40 <manu1> ack ivan
14:59:48 <ShaneM> ShaneM has joined #rdfa
15:00:50 <manu1> PROPOSAL: Add normative text that specifies how to detect an XHTML+RDFa document in HTML+RDFa 1.1. The algorithm is to first, look at the Media Type and see that it's 'application/xhtml+xml'. Second, look at the DTD, if there is one - if it's one we have defined for XHTML1+RDFa 1.1 or XHTML1+RDFa 1.0, we use that. If that does not match, it is XHTML5. Place a warning in the spec stating that documents that don't contain a DTD, don't have @version, and are served as XHTML will default to XHTML5.
15:01:07 <ivan> +1
15:01:09 <ShaneM> +1
15:01:12 <manu1> +1
15:01:16 <scor> +1
15:01:17 <niklasl> +1
15:01:22 <MacTed> +1
15:01:31 <manu1> shane: Make sure to point this back to the relevant sections of RDFa Core 1.1
15:01:35 <manu1> gkellogg: +1
15:01:37 <manu1> RESOLVED: Add normative text that specifies how to detect an XHTML+RDFa document in HTML+RDFa 1.1. The algorithm is to first, look at the Media Type and see that it's 'application/xhtml+xml'. Second, look at the DTD, if there is one - if it's one we have defined for XHTML1+RDFa 1.1 or XHTML1+RDFa 1.0, we use that. If that does not match, it is XHTML5. Place a warning in the spec stating that documents that don't contain a DTD, don't have @version, and are served as XHTML will default to XHTML5.
15:02:36 <Zakim> -Ivan
15:02:51 <ShaneM> q+ to talk about the Role Attribute
15:03:07 <manu1> Topic: RDFa to JSON-LD implementation
15:03:26 <Zakim> -MacTed
15:04:01 <manu1> niklasl: My implementation (in CoffeeScript) of an RDFa to JSON-LD processor is going well. Not too complicated to implement.
15:04:14 <manu1> niklasl: It ties what you ask for back to elements in the DOM, which is what the Microdata API does.
15:04:26 <manu1> niklasl: So, it computes views for the data in the document.
15:04:48 <niklasl> https://github.com/niklasl/rdfa-lab/wiki
15:05:22 <manu1> manu: When will it be production-ready?
15:06:13 <manu1> niklasl: Well, the issue is that the API is something I've created... it will be production ready once four test cases are fixed. The body of the logic and the API shouldn't be considered stable - the outer parts will be just as usable as green turtle implementation. Really, we need to hammer out the RDFa DOM API.
15:06:22 <manu1> Topic: Role Attribute spec
15:07:05 <manu1> shaneM: The @role attribute spec went into CR and has come out without receiving any comments. There are a number of RDFa processors that have done @role attribute processing. I want to encourage implementers to add support for it. There are tests for it in the RDFa Test Suite. I'm doing an implementation report now.
15:07:23 <manu1> shanem: So, there are enough implementations that do it to get out of CR.
15:07:36 <manu1> manu: I intend to add support when I can.
15:07:44 <manu1> shanem: PyRDFa and Ruby distiller support it.
15:08:05 <manu1> shanem: Who did the clojure one? Niklas.
15:08:15 <manu1> shanem: If you want to add support, that would be great.
15:08:54 <tinkster> tinkster has joined #rdfa
15:09:00 <Zakim> -McCarron
15:09:02 <Zakim> -manu1
15:09:03 <Zakim> -scor
15:09:07 <Zakim> -niklasl
15:09:07 <Zakim> SW_RDFa()10:00AM has ended
15:09:07 <Zakim> Attendees were Ivan, scor, niklasl, manu1, MacTed, McCarron
15:09:09 <manu1> niklasl: since this is element-centric, it makes most sense to put it there. You can navigate the DOM by @role, which is exactly what it's used for.
# SPECIAL MARKER FOR CHATSYNC.  DO NOT EDIT THIS LINE OR BELOW.  SRCLINESUSED=00000253