Chatlog 2011-02-01

From RDFa Working Group Wiki
Jump to: navigation, search

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

14:15:32 <RRSAgent> RRSAgent has joined #rdfa
14:15:32 <RRSAgent> logging to http://www.w3.org/2011/02/01-rdfa-irc
14:15:34 <trackbot> RRSAgent, make logs world
14:15:34 <Zakim> Zakim has joined #rdfa
14:15:36 <trackbot> Zakim, this will be 7332
14:15:36 <Zakim> ok, trackbot; I see SW_RDFa(RFDaWG)10:00AM scheduled to start in 45 minutes
14:15:37 <trackbot> Meeting: RDFa Working Group Teleconference
14:15:37 <trackbot> Date: 01 February 2011
14:16:05 <manu> Chair: Manu
14:28:10 <manu> Present: Ivan, Nathan, Knud, Manu, MarkB, ShaneM, Toby
14:45:39 <manu> Agenda: http://lists.w3.org/Archives/Public/public-rdfa-wg/2011Jan/0156.html
14:59:33 <Knud> Knud has joined #rdfa
14:59:50 <manu> zakim, code?
14:59:50 <Zakim> the conference code is 7332 (tel:+1.617.761.6200 tel:+33.4.26.46.79.03 tel:+44.203.318.0479), manu
15:00:03 <Zakim> SW_RDFa(RFDaWG)10:00AM has now started
15:00:08 <Zakim> + +1.540.961.aaaa
15:00:13 <manu> zakim, I am aaaa
15:00:13 <Zakim> +manu; got it
15:00:41 <ShaneM> ShaneM has joined #rdfa
15:02:03 <Zakim> +Knud
15:04:43 <manu> scribenick: manu
15:04:47 <manu> Scribe: Manu
15:05:58 <manu> Topic: Call Start-up
15:06:09 <Zakim> +ShaneM
15:06:11 <manu> Knud: We should have a 5 minute break, halfway through the call.
15:07:49 <manu> Knud: What are we hoping to get through on the call today?
15:08:06 <manu> Manu: Hopefully, everything - next step for all issues should be an official WG response.
15:08:13 <ivan> zakim, dial ivan-voip
15:08:13 <Zakim> ok, ivan; the call is being made
15:08:14 <Zakim> +Ivan
15:09:16 <tinkster> I still can't get on call.  Will keep trying and participate IRC only for now. 
15:09:25 <manu> Topic: ISSUE-71: RDFa Core 1.1 LC comments from Shelley Powers
15:09:33 <tinkster> Could you skip over the issues I was triaging until I can get on?
15:10:11 <manu> Toby: http://www.w3.org/2006/tools/wiki/Zakim-SIP
15:10:24 <markbirbeck> Hi guys. I'm going to have a real problem this afternoon, I'm afraid. We have a release scheduled for tomorrow, and I need to be doing a lot of talking with other people. Would it be possible to keep an eye on IRC, and call in only if needed? That way I can still talk to people here. :)
15:11:06 <markbirbeck> (Yesterday we were ahead of schedule...today we are behind. :( )
15:11:27 <manu> http://lists.w3.org/Archives/Public/public-rdfa-wg/2011Jan/0137.html
15:14:13 <manu> Shane: What about the accessibility issues?
15:14:58 <manu> Manu: We just need to make sure we're not doing anything wrong from an a11y standpoint.
15:21:00 <ivan> http://cmsmcq.com/mib/?p=1113
15:21:32 <manu> Shane: What about linking to triples?
15:22:06 <manu> Manu: It would be nice to forward-link to RDF-specific terms in all sections before section 3. We should also change examples like this:
15:25:58 <manu> <div about="http://dbpedia.org/resource/Albert_Einstein">
15:26:00 <manu>   <div rel="dbp:citizenship">
15:26:01 <manu>     <span about="http://dbpedia.org/resource/German_Empire"></span>
15:26:03 <manu>     <span about="http://dbpedia.org/resource/United_States"></span>
15:26:05 <manu>   </div>
15:26:07 <manu> </div>
15:22:06 <manu> Manu: To examples like this (note the addition of descriptive text)
15:26:42 <manu> <div about="http://dbpedia.org/resource/Albert_Einstein">Albert Einstein was a
15:26:44 <manu>   <div rel="dbp:citizenship">citizen of the
15:26:45 <manu>     <span about="http://dbpedia.org/resource/German_Empire">German Empire</span> and the
15:26:47 <manu>     <span about="http://dbpedia.org/resource/United_States">United States</span>.
15:26:48 <manu>   </div>
15:26:50 <manu> </div>
15:27:41 <manu> Shane: Ok, all looks good.
15:28:03 <manu> Manu: I will write the official response, then.
15:30:09 <manu> Topic: ISSUE-63: Case-insensitive term matching is non-deterministic at times
15:30:15 <manu> http://www.w3.org/2010/02/rdfa/track/issues/63
15:30:45 <manu> Manu: Shane's wording was: 'In the event that multiple terms are defined that differ only in case, it is undefined which term is matched when a reference is made which does NOT match case-sensitively.'
15:31:29 <manu> tinkster, any issue with the wording that shane proposed for ISSUE-63?
15:32:15 <tinkster> fine by me.
15:32:26 <ShaneM> I put this in:  <p class='note' id="C22" about="#C22" resource="http://www.w3.org/2010/02/rdfa/meetings/2011-01-20#ISSUE__2d_63__3a__Case__2d_insensitive_term_matching" rel='bibo:affirmedBy'>In the event that multiple <code>term</code>s are defined that differ only in case (e.g., 'Agent', 'agent', and 'AGENT'), if a reference is made which DOES NOT match case-sensitively (e.g., typeof='AGENt'), the results are UNSPECIFIED.</p>
15:33:42 <manu> Ivan: This is an issue where nobody will be completely happy, the text looks good, lets use it and move on from there.
15:33:56 <manu> Manu: Sounds good to me, let's move on.
15:34:14 <manu> Shane: I put this in the document already.
15:34:18 <manu> Ivan: This issue is DONE!
15:34:55 <manu> Topic: ISSUE-46: Should plain literals that match fully qualified IRIs be automatically converted to IRIs
15:35:02 <ivan> ISSUE-46?
15:35:02 <trackbot> ISSUE-46 -- Should plain literals that match fully qualified IRIs be automatically converted to IRIs -- open
15:35:02 <trackbot> http://www.w3.org/2010/02/rdfa/track/issues/46
15:35:10 <manu> http://www.w3.org/2010/02/rdfa/track/issues/46
15:35:46 <manu> Ivan: It came back up when handling Harry Halpin's LC comment: http://lists.w3.org/Archives/Public/public-rdfa-wg/2011Jan/0121.html
15:36:22 <manu> Manu: People wanted to know if they could do this <meta property="foo:bar" content="http://example.org/baz" />
15:36:24 <Zakim> -ShaneM
15:36:30 <Zakim> +??P0
15:36:36 <webr3> Zakim, I am ??
15:36:36 <Zakim> +webr3; got it
15:36:46 <tinkster> SIP seems to be working. Sound on my computer is not though. (Or at least input audio.)
15:37:05 <manu> zakim, who is on the call?
15:37:05 <Zakim> On the phone I see manu, Knud, Ivan, webr3
15:37:06 <Zakim> +??P7
15:37:16 <ShaneM> zakim, I am P7
15:37:16 <Zakim> sorry, ShaneM, I do not see a party named 'P7'
15:37:18 <ivan> shane, another pointer for colours: http://www.visionaustralia.org.au/info.aspx?page=628
15:37:22 <ShaneM> zakim, I am ??P7
15:37:22 <Zakim> +ShaneM; got it
15:38:42 <manu> Manu: The people that we've been talking with have said that this is not a big issue for them.
15:39:07 <webr3> I'd assert: "people know what to expect by understanding what the property means" (which is how sem web should be, also)
15:40:16 <ivan> <html prefix="foo: URIFORFBNAMESPACE><head><meta propery="foo:FBPREDICATE".../>
15:40:22 <manu> Ivan: Ok, let's get a few more arguments on the table.
15:42:15 <tinkster> I have looked at a lot of implementations of Open Graph Protocol and they nearly *all* just search for the XPath //head/meta[property="og:foo"] (or its equivalent via DOM crawling)
15:42:44 <webr3> tinkster, likewise, true of almost every "curie" I've ever seen used outside semweb
15:44:07 <ShaneM> toby, are you saying that they are looking for a literal string?
15:44:24 <tinkster> yup
15:44:28 <ShaneM> wow
15:45:09 <webr3> ShaneM, that's v common.. from RDF to atom to opengraph and beyond
15:45:24 <webr3> rss* not rdf lol
15:45:46 <webr3> "dc:title" as a string token, not a qname/curie etc
15:45:55 <manu> Manu: ok, well I'll prepare a response to Harry on ISSUE-46
15:46:07 <manu> ACTION: Manu to prepare official WG response to Harry on ISSUE-46
15:46:07 <trackbot> Created ACTION-50 - Prepare official WG response to Harry on ISSUE-46 [on Manu Sporny - due 2011-02-08].
15:51:46 <manu> Nathan: I have a general concern about how this stuff is used in the field - don't people use IRIs in plain literals?
15:55:43 <tinkster> People do have valid reasons for putting IRIs (and things that look like they might be IRIs but might not be really) in plain literals. I've said before and I'll say it again, that Facebook's <http://ogp.me/ns#url> property takes a literal value quite sensibly.
15:56:32 <webr3> if you have http://example.org/foo in the wild, in @content an element, anywhere, people /expect/ it to be treated as a uri/url/iri
15:58:48 <tinkster> Maybe, but do they expect it to be treated as <http://example.org/foo> or as "http://example.org/foo"^^xsd:anyURI? They're different things - the latter is not just a funny way of saying the former.
15:59:50 <Knud> I think not only were spaces the problem, but also different IRI schemes. Pretty much anything /(.*?):(.*?)/ can be an IRI.
16:00:54 <webr3> "fb:me" matches that
16:01:14 <ShaneM> note that 'shane' is also an IRI
16:01:41 <tinkster> ":-)" matches that.
16:02:32 <ShaneM> heh.  http://www.w3.org/TR/xmlschema-2/#anyURI
16:03:17 <Knud>  there were suggestions to only accept certain IRI schemes, but I think people didn't like that back then.
16:03:34 <ShaneM> knud, you are right. 
16:05:09 <manu> Ivan: I think it was good that this issue was raised again, we cannot determininistically see whether or not something is an IRI - there is no way of doing that.
16:05:27 <manu> Manu: Also, keep in mind that there are other solutions to this problem, such as type coercion.
16:07:44 <manu> PROPOSAL: Response to ISSUE-46 should point out that there is no algorithm to deterministically identify IRIs.
16:07:52 <ivan> +1
16:07:53 <webr3> +1
16:07:55 <Knud> +1
16:07:56 <manu> +1
16:07:57 <ShaneM> +1
16:08:05 <manu> RESOLVED: Response to ISSUE-46 should point out that there is no algorithm to deterministically identify IRIs.
16:08:15 <manu> [Five minute break]
16:08:21 <Zakim> -webr3
16:08:43 <Zakim> -Knud
16:09:34 <Zakim> -ShaneM
16:09:47 <manu> zakim, who is on the call?
16:09:47 <Zakim> On the phone I see manu, Ivan
16:10:34 <Zakim> -manu
16:10:57 <Zakim> +??P2
16:10:58 <Zakim> -??P2
16:10:58 <Zakim> +??P2
16:11:02 <manu> zakim, I am ??P2
16:11:02 <Zakim> +manu; got it
16:14:36 <Zakim> +??P0
16:14:44 <ShaneM> zakim, ??P0 is ShaneM
16:14:44 <Zakim> +ShaneM; got it
16:14:47 <manu> If there is a deterministic way to determine an IRI, then our argument falls.
16:14:54 <tinkster> No, :-) doesn't match IRI - it just matches Knud's expression. (His regular expression, not necessarily his facial one.)
16:14:55 <manu> zakim, who is on the call?
16:14:56 <Zakim> On the phone I see Ivan, manu, ShaneM
16:15:05 <webr3> toby, lol
16:15:12 <manu> That may be his current facial expression - hard to know.
16:15:27 <Zakim> +[IPcaller]
16:15:34 <webr3> Zakim, i am IPcaller
16:15:34 <Zakim> ok, webr3, I now associate you with [IPcaller]
16:16:56 <Zakim> +Knud
16:18:19 <Zakim> + +44.785.583.aabb
16:18:22 <Zakim> - +44.785.583.aabb
16:18:32 <ShaneM> <h2 rel='title'>http://whatever</h2>
16:18:58 <Zakim> + +44.785.583.aacc
16:19:04 <tinkster> Zakim, aacc is me
16:19:04 <Zakim> +tinkster; got it
16:19:08 <tinkster> Zakim, mute me
16:19:08 <Zakim> tinkster should now be muted
16:19:26 <ivan> <h2 property="title">http://whatever</h2>
16:19:30 <manu> <h2 property='myurl'>http://whatever/</h2> => <> myurl <http://whatever/>
16:19:53 <ivan> <h2 property="title" datatype="">http://whate4ver</h2>
16:20:12 <webr3> href="a"
16:21:20 <Knud> manu: only a deterministic way of determining an IRI would save us
16:22:44 <tinkster> Agent/agent vs aGenT.
16:23:03 <manu> PROPOSAL: ISSUE-63 official WG response should refer to the official draft spec text on how to handle case-insensitive term matching.
16:23:28 <manu> PROPOSAL: ISSUE-63 official WG response should refer to the editors draft spec text on how to handle case-insensitive term matching.
16:23:48 <tinkster> refer to versioned URI?
16:24:20 <Knud> ivan: cannot refer to draft spec, because that changes all the time
16:24:54 <manu> PROPOSAL: ISSUE-63 - case-insensitive term matching is non-deterministic.
16:24:57 <tinkster> +1 happy with that. Implementors can always try to agree on a strategy later.
16:25:03 <ivan> +
16:25:04 <manu> +1
16:25:07 <ivan> +1
16:25:07 <Knud> +1
16:25:10 <webr3> +1
16:25:16 <manu> RESOLVED: ISSUE-63 - case-insensitive term matching is non-deterministic.
16:25:16 <ivan> for the records: the proposed text is: >In the event that multiple <code>term</code>s are defined that differ only in case (e.g., 'Agent', 'agent', and 'AGENT'), if a reference is made which DOES NOT match case-sensitively (e.g., typeof='AGENt'), the results are UNSPECIFIED.
16:25:49 <manu> PROPOSAL: Editorial suggestions in ISSUE-71 draft response is fine, send official response to Shelley Powers.
16:26:03 <Knud> +1
16:26:10 <manu> +1
16:26:13 <ShaneM> +1
16:26:16 <webr3> +1
16:26:18 <tinkster> +1
16:26:29 <ivan> +1
16:26:31 <manu> RESOLVED: Editorial suggestions in ISSUE-71 draft response is fine, send official response to Shelley Powers.
16:26:56 <ivan> ISSUE-65?
16:26:56 <trackbot> ISSUE-65 -- Feedback from Michael Hausenblas on RDFa Core 1.1 -- open
16:26:56 <trackbot> http://www.w3.org/2010/02/rdfa/track/issues/65
16:26:58 <manu> Topic: ISSUE-65: Last Call feedback from Michael Hausenblas
16:27:00 <manu> http://www.w3.org/2010/02/rdfa/track/issues/65
16:27:25 <manu> http://www.w3.org/2010/02/rdfa/meetings/2011-01-13#ISSUE__2d_65__3a__Michael_Hausenblas__27__LC_Comments
16:28:16 <Knud> manu: we want an official WG LC response for each issue
16:28:37 <Knud> shane: I'll do that
16:29:01 <Knud> ivan: I added the diagram to the document
16:29:11 <manu> ACTION: Shane to write official response to Michael Hausenblas for ISSUE-65
16:29:11 <trackbot> Created ACTION-51 - Write official response to Michael Hausenblas for ISSUE-65 [on Shane McCarron - due 2011-02-08].
16:29:33 <ivan> ISSUE-67?
16:29:33 <trackbot> ISSUE-67 -- RDFa Core 1.1 LC comments from Henri Sivonen -- open
16:29:33 <trackbot> http://www.w3.org/2010/02/rdfa/track/issues/67
16:29:34 <manu> Topic: ISSUE-67: RDFa Core 1.1 LC comments from Henri Sivonen
16:29:52 <manu> http://www.w3.org/2010/02/rdfa/track/issues/67
16:30:27 <webr3> : http://lists.w3.org/Archives/Public/public-rdfa-wg/2011Jan/0140.html
16:30:34 <tinkster> Zakim, unmute me
16:30:34 <Zakim> tinkster should no longer be muted
16:31:02 <Knud> tinkster: Henry doesn't want CURIEs in RDFa at all
16:31:16 <Knud> s/Henry/Henri
16:31:28 <Knud> … : we cannot really do that
16:31:53 <Knud> (going through the editorial changes proposed)
16:32:07 <webr3> -> HS: The concept of "processor graph" seems to be an open-ended loophole of non-interoperability. - issue
16:34:48 <webr3> "foo: http://" valid, "foo:   http://" invalid?
16:36:17 <Knud> (going over to the "won't fix" section)
16:38:11 <Knud> ivan: I have no comments on the won't fix section
16:38:49 <manu> Manu: Looks good to me.
16:38:53 <Knud> … : Toby's answers are good
16:39:33 <Knud> tinkster: some editorial changes have not been addressed yet
16:42:01 <Knud> "The concept of "processor graph" seems to be an open-ended loophole of non-interoperability."
16:42:22 <Knud> manu: we don't really say exactly what is in the pg
16:43:04 <Knud> … : we could respond by saying that the WG will have a test suite that will clarify this
16:43:23 <Knud> tinkster: we could define a vocabulary, but not insist that it is being used
16:43:48 <Knud> ivan: we talked about a vocab before, I had a proposal, but that was rejected
16:44:18 <Knud> manu: there were objections because it might take us too long to agree on a vocab (by mark?)
16:44:24 <webr3> q+
16:44:37 <Knud> ivan: I had a wiki page with a proposal
16:44:38 <webr3> Zakim, I am IPcaller
16:44:38 <Zakim> ok, webr3, I now associate you with [IPcaller]
16:45:11 <manu> ack webr3
16:45:14 <manu> ack webr3
16:45:32 <Knud> manu: we could say: we are having a PG graph vocab on a wiki and we will have examples in the test suite
16:46:23 <Knud> ivan: the spec defines error conditions, but not how errors should be raised
16:46:25 <manu> http://www.w3.org/TR/rdfa-core/#accessing-the-processor-graph
16:47:32 <ShaneM> and also http://www.w3.org/2010/02/rdfa/sources/rdfa-core/Overview-src.html#dfn-processor_graph
16:48:59 <ShaneM> q+ to talk about appendix B
16:49:37 <ivan> q+
16:49:43 <manu> ack shanem
16:49:43 <Zakim> ShaneM, you wanted to talk about appendix B
16:49:50 <webr3> q-
16:49:57 <ShaneM> v
16:49:57 <Knud> tinkster: we require a PG, but we say very little regarding what should be in it
16:50:02 <ShaneM> http://www.w3.org/2010/02/rdfa/sources/rdfa-core/Overview-src.html#vocabulary
16:51:11 <Knud> ShaneM: the reference to "processor graph" here confuses me
16:51:21 <Knud> manu: this should read "profile"
16:52:24 <manu> ack ivan
16:52:33 <Knud> manu: there are several use cases for the PG
16:52:41 <ivan> http://www.w3.org/2010/02/rdfa/wiki/Processor_Graph_Vocabulary
16:53:08 <Knud> ivan: that's the link to my PG vocabulary proposal
16:54:29 <Knud> … : we could say the implementation SHOULD use this vocabulary and MAY add any other triples it likes
16:55:11 <tinkster> Zakim, mute me
16:55:11 <Zakim> tinkster should now be muted
16:55:40 <tinkster> Zakim, unmute me
16:55:40 <Zakim> tinkster should no longer be muted
16:56:50 <Knud> Nathan: happy with this proposal. Will look at the proposed vocabulary.
16:57:49 <Knud> tinkster: just because there are use cases for a PG, we don't have to require it
17:00:00 <ShaneM>          An ERROR must be generated when the document fails to be fully          processed as a result of non-conformant host language markup.                    An ERROR must be generated when a referenced RDFa Profile is not           recognized, as described in RDFa Profiles.                  A WARNING must be generated when a CURIE prefix fails to be          resolved.         A WARNING must be generated when a Term fails to be resolved.       
17:00:23 <Knud> ivan: these are all the error conditions we have
17:00:59 <Knud> … : this is necessary to get back to the user - how can we guarantee that other than with a PG?
17:01:22 <webr3> ERROR states are full failures yes? can't continue processing?
17:01:27 <Knud> tinkster: should be left application-specific. Could also be a callback function, etc.
17:01:40 <ShaneM> webr3: no I dont think so
17:02:04 <manu> ACTION: Ivan to update and integrate Processor Graph Vocabulary into the RDFa Core spec: http://www.w3.org/2010/02/rdfa/wiki/Processor_Graph_Vocabulary
17:02:05 <trackbot> Created ACTION-52 - Update and integrate Processor Graph Vocabulary into the RDFa Core spec: http://www.w3.org/2010/02/rdfa/wiki/Processor_Graph_Vocabulary [on Ivan Herman - due 2011-02-08].
17:03:03 <manu> ACTION: Ivan to modify RDFa Core 7.6.1 to make Processor Graph support optional, but if Processor Graph support is included, the Processor Graph Vocabulary MUST be used.
17:03:04 <trackbot> Created ACTION-53 - Modify RDFa Core 7.6.1 to make Processor Graph support optional, but if Processor Graph support is included, the Processor Graph Vocabulary MUST be used. [on Ivan Herman - due 2011-02-08].
17:04:23 <manu> PROPOSAL: RDFa Core should include a Processor Graph Vocabulary and make Processor Graph support optional.
17:04:29 <ivan> +1
17:04:30 <manu> +1
17:04:37 <tinkster> +1
17:04:47 <ShaneM> +1
17:04:47 <Knud> +1
17:04:53 <manu> RESOLVED: RDFa Core should include a Processor Graph Vocabulary and make Processor Graph support optional.
17:04:56 <tinkster> OKK.
17:04:58 <webr3> +1 (and if processor graph is present, must use grpah vocab)
17:06:24 <ivan> zakim, drop me
17:06:24 <Zakim> Ivan is being disconnected
17:06:24 <ShaneM> ACTION: Shane Introduce a definition for the PREFIX datatype that is referenced in the schema implementation of XHTML+RDFa
17:06:24 <trackbot> Created ACTION-54 - Introduce a definition for the PREFIX datatype that is referenced in the schema implementation of XHTML+RDFa [on Shane McCarron - due 2011-02-08].
17:06:25 <Zakim> -Ivan
17:08:25 <Knud> manu: is there some DOM spec that supresses whitespace?
17:11:19 <manu> http://www.w3.org/TR/xmlschema-1/#d0e1654
17:12:10 <Knud> shane: in XHTML whitespace must be preserved, in XML whitespace collapsing is permissable
17:14:40 <webr3> "white space preservation is defined by the host language"?
17:15:09 <Knud> tinkster: we should point that out in the spec, preferably with a reference (as manu put in)
17:15:26 <Zakim> -tinkster
17:16:05 <manu> ACTION: Toby to draft official RDFa WG response for ISSUE-67
17:16:05 <trackbot> Created ACTION-55 - Draft official RDFa WG response for ISSUE-67 [on Toby Inkster - due 2011-02-08].
17:16:05 <webr3> ACTION: Nathan Rixham review http://www.w3.org/2010/02/rdfa/wiki/Processor_Graph_Vocabulary and give Ivan feedback
17:16:05 <trackbot> Created ACTION-56 - Rixham review http://www.w3.org/2010/02/rdfa/wiki/Processor_Graph_Vocabulary and give Ivan feedback [on Nathan Rixham - due 2011-02-08].
17:16:29 <manu> Topic: ISSUE-73 and ISSUE-78 - Default RDFa Profile
17:15:59 <Knud> manu: I agree with your proposal, we will discuss this here on the call
17:16:07 <ivan> -> http://lists.w3.org/Archives/Public/public-rdfa-wg/2011Jan/0160.html Ivan's proposal
17:16:12 <trackbot> Sorry, couldn't find user - Introduce
17:16:32 <manu> http://lists.w3.org/Archives/Public/public-rdfa-wg/2011Jan/0160.html Ivan's proposal
17:18:10 <webr3> see-also - ACTION-511 TAG
17:21:20 <manu> http://w3.org/rdf-profiles/2010
17:21:50 <manu> http://w3.org/rdf-profiles/2014-rdfa
17:23:22 <webr3> "the rdfa profile may be updated in the future" ?
17:26:56 <manu> Every time you remove a profile, Shane kills a triple.
17:30:17 <manu> http://w3.org/rdf-profile
17:31:00 <Knud> webr3: The main problem is removing definitions from a profile, regardless of its URI or if we use a new default profile.
17:31:24 <ShaneM> <html profile='datedURIforProfile'>
17:31:32 <ShaneM> that's the only way to be sure
17:31:53 <Knud> … : People in their documents need to reference a stable profile
17:36:50 <manu> Manu: You can't remove or change any prefixes or terms in a profile at a particular URL. Appending is fine. If you have to change a URL, it MUST be backwards-compatible.
17:42:15 <manu> RDFa Core 1.1 => http://w3.org/rdf/profiles/2010
17:42:18 <manu> RDFa Core 2.0 => http://w3.org/rdf/profiles/2015
17:42:40 <ShaneM> youre not missing much
17:42:57 <manu> RDF default profile would be => http://w3.org/rdf/profiles
17:43:15 <manu> RDFa Core 1.1 would use => http://w3.org/rdf/profiles/2011
17:43:15 <manu> RDFa Core 2.0 would use => http://w3.org/rdf/profiles/2015
17:43:23 <manu> The latest working prefixes/terms => http://w3.org/rdf/profiles
17:43:43 <manu> Nobody would ever use: http://w3.org/rdf/profiles - except for discussion on what the latest prefixes/terms should be.
17:43:54 <manu> RDFa processor implementations would ALWAYS use the dated URI
17:44:15 <manu> Nathan: We could just not define: http://w3.org/rdf/profiles
17:44:23 <ivan> if I understand things right I do not agree. An implementation that is caching would always use a non-dated URI
17:44:49 <ivan> dated uri for a default profile makes only sense if you want to keep to 'history'
17:45:03 <ShaneM> ivan - you are missing too much context 
17:45:09 <ivan> sorry...:-(
17:48:20 <manu> http://w3.org/rdf-default-profile
17:50:20 <manu> lots of discussion on approaches
17:52:21 <webr3> http://prefix.cc/dc
17:55:05 <webr3> I assert that "dc" can't go in the default profile, because it is ambigous
17:55:17 <manu> Why do you think it's ambiguous?
17:55:27 <ivan> webr3: this is something the DCMI people have to tell us
17:55:42 <webr3> -> http://prefix.cc/dc
17:55:59 <webr3> it's not what dcmi say, it's how the web population uses the string "dc"
17:56:04 <Knud> web3: this is based on how the prefix "dc" is actually used
17:56:30 <manu> People can override w/ xmlns:dc thought, right?
17:56:58 <Knud> but what do we set it to in the profile?
17:57:10 <webr3> yes, what Knud said
17:57:47 <ivan> this is something that the dcmi community have to tell us
17:58:03 <ivan> dcmi community -> DCMI as a holder of that vocabulary
17:58:21 <webr3> "dc:title"
17:59:35 <Knud> but webr3, if dc is so ambiguous, then we just cannot include it in the default profile at all
17:59:47 <webr3> yes, what Kund said again :)
18:00:06 <webr3> s/kund/Kund (sorry)
18:00:32 <Zakim> -ShaneM
18:01:54 <manu> Nathan: We need a default profile that processors should implement, if people want it to be stable, they must reference it directly.
18:02:04 <Knud> so this is not a technical problem we can solve, it's a social problem?
18:03:03 <webr3> we should encourage people to reference a stable profile if they want stability, and not /rely/ on the default profile
18:03:09 <Knud> what Nathan says sounds as though we should not have a default profile after all
18:03:34 <Knud> then the default profile is useless
18:03:35 <ivan> zakim, dial ivan-voip
18:03:35 <Zakim> ok, ivan; the call is being made
18:03:37 <Zakim> +Ivan
18:03:44 <webr3> as in, the default profile we provide for peopel to reference, not for people to depend on as having unambigous forever meaning (mapping of string to uris, we can't promise that, for social reasons)
18:04:53 <webr3> should I write it all up in a mail to the list?
18:06:04 <Zakim> -Knud
18:17:14 <manu> Yes, please do.
18:17:15 <manu> Ivan: We shouldn't shoot down the idea of RDFa profiles just because Dublin Core has this ambiguity issue.
18:17:15 <manu> Nathan: What happens if we need to change FOAF in the future?
18:17:16 <manu> Manu: Then we would use foaf2 or foaf3 as the prefix.
18:18:00 <manu> Manu: We'll pick this up on the mailing list later this week.
18:19:13 <ivan> zakim, drop me
18:19:13 <Zakim> Ivan is being disconnected
# SPECIAL MARKER FOR CHATSYNC.  DO NOT EDIT THIS LINE OR BELOW.  SRCLINESUSED=00000352