Chatlog 2010-09-23

From RDFa Working Group Wiki
Jump to: navigation, search

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

14:00:00 <RRSAgent> RRSAgent has joined #rdfa
14:00:01 <RRSAgent> logging to
14:00:02 <trackbot> RRSAgent, make logs world
14:00:02 <Zakim> Zakim has joined #rdfa
14:00:04 <trackbot> Zakim, this will be 7332
14:00:04 <Zakim> ok, trackbot; I see SW_RDFa()10:00AM scheduled to start now
14:00:05 <trackbot> Meeting: RDFa Working Group Teleconference
14:00:06 <trackbot> Date: 23 September 2010
14:00:14 <Steven> zakim, dial steven-617
14:00:14 <Zakim> ok, Steven; the call is being made
14:00:15 <Zakim> SW_RDFa()10:00AM has now started
14:00:16 <Zakim> +Steven
14:01:27 <ivan> zakim, dial ivan-voip
14:01:33 <Zakim> ok, ivan; the call is being made
14:01:35 <Zakim> +Ivan
14:02:08 <Zakim> +??P19
14:02:14 <manu1> zakim, I am ??P19
14:02:14 <Zakim> +manu1; got it
14:03:21 <Zakim> + +49.631.205.75.aaaa
14:03:41 <Benjamin> zakim, aaaa is Benjamin
14:03:41 <Zakim> +Benjamin; got it
14:04:53 <ShaneM> ShaneM has joined #rdfa
14:05:11 <Zakim> +ShaneM
14:06:54 <manu1> Agenda:
14:06:58 <manu1> Chair: Manu
14:07:28 <manu1> Regrets: MarkB
14:07:37 <manu1> Present: Ivan, Steven, Benjamin, Manu, Shane
14:08:30 <manu1> Topic: RDFa-related news
14:08:57 <manu1> Benjamin: There was a question about further annotating stuff in RDFa in XML3D at a recent talk I gave?
14:09:13 <manu1> Ivan: They could put the data in SVG metadata tag if they want to do that.
14:09:20 <manu1> Manu: There are two very major content industries that are looking at deploying their next major metadata formats in XHTML+RDFa. The information may go public later this year and during the early part of next year.
14:09:39 <Benjamin> scribenick: Benjamin
14:09:40 <manu1> scribe: Benjamin
14:16:03 <manu1> Topic: ISSUE-41: Under-defined processor graph vocabulary
14:16:14 <manu1>
14:16:44 <Benjamin> manu1: Talked to Mark about this issue. He is concerned that we don't have the time required to deal with this issue. Agreed that an RDFa Processor Graph should exist in RDFa Core.
14:17:08 <Benjamin> ... also agreed about the vocabulary to be out of scope here.
14:18:05 <ivan> q+
14:18:13 <Benjamin> ... implementors are interested in this vocabulary. Therefore is it still out of scope?
14:18:55 <manu1> Shane: I disagree that it's out of scope - it is very much in scope.
14:18:57 <Benjamin> ivan: The vocabulary is not out of scope, but the group could not reach a consensus about what should go into the vocabulary.
14:19:53 <Benjamin> ... we have other issues to solve and should not re-open the issue about the vocabulary.
14:20:58 <Steven> -> What do they do with my RDFa?
14:25:10 <Benjamin> manu1: Well, how does this sound like for a plan to proceed: Those that are interested in RDFa Processor Graph interoperability will hack on an RDFa Processor Graph vocabulary on the wiki. Once we have general agreement, we'll see if we want to integrate with RDFa Core. If we can't reach consensus, we may publish it as a note, for those that are interested in an RDFa Processor graph vocabulary.
14:25:40 <Benjamin> manu1: Ivan will add the vocabulary he already defined to the wiki and will send it around the mailing list.
14:26:58 <Benjamin> manu1: we close issue 41 as it is not critical path to RDFa Core and those that are interested will implement the Processor Graph Vocabulary that we define outside of the RDFa Core spec.
14:27:04 <manu1> Topic: ISSUE-45: @cite and @longdesc support
14:27:14 <manu1>
14:27:46 <Benjamin> manu1: We have been asked to make RDFa generate triples when we see longdesc and cite
14:28:19 <ivan> q+
14:28:27 <Benjamin> ... do we we support the @longdesc and @cite attributes?
14:29:37 <Benjamin> Shane: We always meant to support these attributes in XHTML2. It's simple to implement.
14:31:09 <Benjamin> Steven: Metadata elements in HTML such as title should conatin markup, so there was always this notion that we would support stuff like this.
14:31:16 <Benjamin> Ivan: I am against it
14:32:21 <Benjamin> ... it will require too much effort to discuss with little benefit.
14:33:17 <Zakim> -Benjamin
14:34:21 <Zakim> +Benjamin
14:35:00 <Benjamin> Manu: I don't think we should support @longdesc or @cite - it complicates the RDFa processing rules, gives us two more attributes that authors have to remember produce triples, and it is not clear how much benefit it would provide. This doesn't mean that other's can't modify the RDFa Processing rules as it's pretty clear what triples should be generated - just don't think that it needs to go in RDFa Core. So if this is an accessibility issue, an accessibility compliant RDFa Processor could parse @role, @longdesc and @cite and generate triples for it - no problem. We'll put this out to the mailing list, but it doesn't sound like there would be major objections if we didn't support this in RDFa Core. Other specs can define this if they want to.
14:35:23 <Benjamin> manu1: Any objections if we close this issue via the mailing list?
14:35:25 <Benjamin> No objections noted.
14:35:29 <Benjamin> Topic: ISSUE-16: RDF Collections
14:35:29 <trackbot> ISSUE-16 Determine if there is a good way of supporting RDF Collections notes added
14:37:34 <manu1>
14:37:34 <ivan> -> Ivan's proposal
14:39:57 <Benjamin> Shane: Can't I generate triples about RDF Collections by using e.g., @typeof rdf:Seq?
14:40:55 <Benjamin> Ivan: Yes, you can. rdf:List is a bit more difficult to do in RDFa. You have to build up the list manually which is very complicated in RDFa
14:41:16 <manu1> q+ to discuss linked data cloud
14:41:28 <Zakim> -Steven
14:42:01 <Benjamin> q+
14:42:04 <ivan> q+
14:42:22 <manu1> ack ivan
14:42:23 <Steven> zakim, dial steven-617
14:42:24 <manu1> ack manu1
14:42:24 <Zakim> ok, Steven; the call is being made
14:42:26 <Zakim> +Steven
14:42:26 <ivan> q+
14:42:27 <Zakim> manu1, you wanted to discuss linked data cloud
14:42:43 <ivan> ack Benjamin 
14:43:01 <manu1> ack ivan
14:43:34 <manu1> Benjamin: These features are very useful - they just have bad support thus far.
14:43:36 <Benjamin> manu1: We don't need to support this feature because no one is using SEQ, ALT, BAG - if you check the Linked Data Cloud Cache, there is very little usage of these constructs.
14:44:52 <Benjamin> q+
14:44:55 <manu1> q+ to say that we don't disallow this stuff
14:45:51 <Benjamin> Ivan: The LOD cache is not representative of all systems that use RDF sequences, alt or bags. OWL is using lists extensively, so does XMP (which is embedded in almost every photoshop file/image created). There are many applications that depend or will depend on this stuff.
14:45:54 <manu1> ack Benjamin
14:46:14 <manu1> ack manu1
14:46:14 <Zakim> manu1, you wanted to say that we don't disallow this stuff
14:47:16 <Benjamin> manu1: We don't disallow usage of this stuff. The issue is just to make it easier in RDFa to express lists and similar structures.
14:50:18 <Benjamin> ivan: We should separate the ISSUE into its two components: The first is if RDF lists, sequences and bags is something that we want to provide a syntax for in RDFa. The second is to ask if the technical solution outlined in the wiki is the solution that we want.
14:51:16 <Benjamin> Steven: I'm of two minds, but it would be nice to support it
14:52:34 <Benjamin> Shane: If we have the energy, we should solve the problem
14:52:39 <ivan>
14:52:54 <Benjamin> manu1: I don't think we should address this particular problem, you can already express the items in RDFa, I think it'll take too long to come to consensus on this.
14:54:57 <manu1> q+ to discuss LC timeline
14:56:23 <Benjamin> manu1: Our last call is in 5 weeks. We don't have enough time do make such a major change.
14:57:21 <ShaneM> q+ to discuss schedule
14:57:25 <manu1> ack manu1
14:57:25 <Zakim> manu1, you wanted to discuss LC timeline
14:57:44 <Benjamin> manu1: based on the timeline, should we skip this or not?
14:58:16 <manu1> ack ShaneM
14:58:17 <Zakim> ShaneM, you wanted to discuss schedule
14:59:20 <Benjamin> ivan: it's nearly finished and nobody spoke against the solution in the maining list. It's already implemented in my PyRDFa processor, so we know it works.
14:59:44 <Benjamin> shane: I don't want to change the schedule, we should deal with it if we have the time.
15:00:04 <manu1> q+ to end the call
15:00:13 <Benjamin> ivan: First priority should be the format of profile files
15:00:30 <Benjamin> manu1: three major issues are left: RDFa Profile file format, how to generate triples for object literals that are clearly URLs, and this RDF Collections issue.
15:01:20 <Benjamin> manu1: We'll handle the remaining issues in that order and stick to our timeline.