daily pastebin goal

Structured Data in HTML Meeting at SemTech 2011

a guest Jun 8th, 2011 641 Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
  1. [Wed 15:48]     <tantek>        hello manu
  2. [Wed 15:48]     <sandro>        kavi: We wanted to make life simpler. If we accept more things, what are the upsides to doing that.
  3. [Wed 15:48]     <manu-db>       hi tantek :)
  4. [Wed 15:48]     <tantek>        kavi - life is always simpler in a dictatorship ;)
  5. [Wed 15:48]     <manu-db>       I've been meaning to e-mail you about this whole schema.org thing... heard you on that live radio broadcast a few days ago and couldn't help but nod my head in agreement.
  6. [Wed 15:49]     <tantek>        I appreciated your blog post also manu.
  7. [Wed 15:49]     <tantek>        who is speaking right now?
  8. [Wed 15:49]     <sandro>        giovanni: The point is this. All it takes to make everyone in this room, etc, happy is to add ONE LINE to the spec, "We will support this vocab ALSO IN RDFa." Then everyone can keep doing what they are doing, all the research and the projects. To support this takes ONE LINE of code, there are existing parsers. No extra parsers to write. Add that sentence, and we're done.
  9. [Wed 15:49]     <sandro>        kavi: It's not one line of code.
  10. [Wed 15:49]     <danbri>        hey tantek, i didn't see Microformats 2 post, but a big +1 to 'Proposal: simplify all microformats to flat sets of properties.'
  11. [Wed 15:49]     <tantek>        yeah it's not one line of code
  12. [Wed 15:49]     <hober> srsly
  13. [Wed 15:49]     <sandro>        kavi: My question is really what are the use cases. "Make everyone on the web happy"?
  14. [Wed 15:49]     <tantek>        danbri - microformats two has been openly discussed since I first brainstormed it on the wiki May 2010
  15. [Wed 15:50]     <sandro>        Giovanni: I have a long list of use cases not addressed by schema.org
  16. [Wed 15:50]     <tantek>        and we held an open session discussing it / introducing microformats 2.0 at SXSW 2011 in March
  17. [Wed 15:50]     <tantek>        Giovanni sounds upset
  18. [Wed 15:50]     <danbri>        i'm sure it has, I'm not complaining! I just missed it
  19. [Wed 15:50]     <tantek>        danbri - true enough - easy to miss all the awesome open standards work going on
  20. [Wed 15:50]     <tantek>        :)
  21. [Wed 15:50]     <sandro>        Giovanni: I don't understand why you announced it like this. I'm not a partisan, but if we can JUST say RDFa is okay, then everyone is happy.
  22. [Wed 15:51]     <sandro>        kavi: I am asking for the use cases so we can talk through them
  23. [Wed 15:51]     <vulcan_wsmith> Giovanni: Is laying it out there
  24. [Wed 15:51]     <tantek>        Giovanni - RDFa doesn't make everyone happy.
  25. [Wed 15:51]     <tantek>        having *choice* makes everyone happy
  26. [Wed 15:51]     <danbri>        They could say microformats are OK too
  27. [Wed 15:51]     <sandro>        Giovannl: It's really none of Google's businesses, actually.
  28. [Wed 15:51]     <tantek>        kavi - we ask for the use cases at microformats.org also
  29. [Wed 15:51]     <sandro>        Giovanni: You can get into whatever business you want, but let other people have their businesses, too.
  30. [Wed 15:51]     <tantek>        Kavi, you or Google should not be the arbiter of use-cases
  31. [Wed 15:52]     <tantek>        that's what open standards communities are for
  32. [Wed 15:52]     <tantek>        not one company
  33. [Wed 15:52]     <tantek>        or two companies
  34. [Wed 15:52]     <danbri>        it's not for a couple of big companies to decide by fiat, and say in same breath 'but we'll monitor uptake of the specs we've just overshadowed to see if we got it wrong'
  35. [Wed 15:52]     <tantek>        danbri - agreed
  36. [Wed 15:52]     * danbri        for the record is entirely not angry
  37. [Wed 15:52]     <sandro>        unk1: I don't want to get into syntacitc bickering; it's about the process, where you three sat down and came up with this alone.
  38. [Wed 15:52]     * tantek        has never seen danbri get angry at anything ;)
  39. [Wed 15:52]     <tantek>        (wonders what it would take)
  40. [Wed 15:52]     <sandro>        ivan: repeatinging, there are two issues here: syntax and vocab management.
  41. [Wed 15:52]     <danbri>        count yourself lucky ;)
  42. [Wed 15:52]     <tantek>        :)
  43. [Wed 15:53]     <sandro>        ivan: 10 more minutes on syntax
  44. [Wed 15:53]     <danbri>        is anyone relaying questions from IRC?
  45. [Wed 15:53]     <sandro>        unk1: Microformats says have a discussion first. You did that with hRecipe, so I'm surprsed to see you didnt go through that here. That'a the difference in phsilophy
  46. [Wed 15:54]     <tantek>        Google (Kavi in particular!) successfully worked with the open community on both hReview-aggregate and hRecipe - openly.
  47. [Wed 15:54]     <sandro>        kavi: This (richsnippets) team has been doing this for 2 years now. Starting with microformats and RDFa. We didnt want to pick a winner, force something on the world. Lots of tradeoffs. We saw people doing both. Most of the adoption was microformats.
  48. [Wed 15:54]     <tantek>        "historically most of the adoption has been microformats" (did I hear that right?)
  49. [Wed 15:54]     <hober> tantek: you did
  50. [Wed 15:54]     <danbri>        Q+ please "Would you agree that the uptight, patent-mentioning terms of service could discourage innovation and creative use of schema.org-formatted data? Can the ToS be relaxed?"
  51. [Wed 15:54]     <sandro>        kavi: Going and talking with sites, what drove people was How Simple Is This. And the things we think are at the simplest end of the spectrum, are at the most complex for userrs.
  52. [Wed 15:55]     <tantek>        How simple?
  53. [Wed 15:55]     <tantek>        simple for publishers/authors matters more than simple for developers
  54. [Wed 15:55]     <tantek>        I'm not buying the more confusion over time point - citation required.
  55. [Wed 15:55]     <sandro>        kavi: We were expecting consolitdation. Then microdata came up as well, and we did a 3rd version. There was no consensus being reached over time, both in syntax and vocab. it was getting worse, not better. Multiple standards, without support for one, and this looked bad.
  56. [Wed 15:56]     <tantek>        if anything we *had* started to see some level of convergence at least on vocabularies among open communities
  57. [Wed 15:56]     <sandro>        kavi: So we asked how we could drive up adoption
  58. [Wed 15:56]     <sandro>        kavi: So there is the question about what we picked, and the question about the process.
  59. [Wed 15:56]     <sandro>        kavi: We made a judgement call. It may not be the call most of the people in this room agreed with.
  60. [Wed 15:56]     <tantek>        Google picking a winner is not good for the open web.
  61. [Wed 15:56]     <bergie>        maybe there was a good reason for each three to exist?
  62. [Wed 15:56]     <sandro>        kavi: In terms of process, we knew a smaller group would be more likely to get it out the door.
  63. [Wed 15:56]     <tantek>        bergie - we're still exploring this space of technologies
  64. [Wed 15:57]     <tantek>        all 3 have strengths and weaknesses
  65. [Wed 15:57]     <sandro>        kavi: (philosophy of markup)
  66. [Wed 15:57]     <tantek>        and we've been learning from each others innovations
  67. [Wed 15:57]     <danbri>        amen re strengths/weaknesses
  68. [Wed 15:57]     <bergie>        tantek: exactly. So Google just pushing a particular format (and the least popular of them!) down everybody's throats is definitely bad
  69. [Wed 15:57]     <sandro>        kavi: Should we be working with our competitors? We wanted to simplify things for webmasters. Markup is hopelessly muddled. WHATEVER we pick, some people will be upset
  70. [Wed 15:57]     <danbri>        in some ways microdata is like the awkward offspring of microformats + rdfa designs
  71. [Wed 15:58]     <tantek>        danbri - I'd disagree with that
  72. [Wed 15:58]     <sandro>        kavi: The acheivement was to get something out there. We know it's not perfect. We can make it better. We hope this can be a step toward great adoption.
  73. [Wed 15:58]     <sandro>        David_Recordon: One of the concerns I heard, was if you want to be using schema.org, you can't put RDFa or Microformats in the page. That stops competition and evolution
  74. [Wed 15:58]     <danbri>        i remember seeing a tantek review somewhere that was quite admiring of microdata (did i imagine that?); but it came about thru initially a hixification of an htmlized RDFa -- until rdfa folk objected to use of the name for such a creature, and it got renamed Microdata
  75. [Wed 15:59]     * danbri        glad David's there too
  76. [Wed 15:59]     <sandro>        kavi: That was not the intent -- I hope that's changed. We were just trying to be pragmatic.
  77. [Wed 15:59]     <sandro>        unk1: You should have known better
  78. [Wed 15:59]     <tantek>        Kavi, first, Google should not be picking syntax winners (which was the philosophy in the initial Rich Snippets releases). Second, what is the scientific basis for picking the *least* adopted of the syntaxes? Looks completely unreasonable (scientifically) without presenting data to the contrary.
  79. [Wed 15:59]     <sandro>        kavi: That was an honest mistake, we'll fix that.
  80. [Wed 16:00]     <hober> see also http://search.twitter.com/search?q=%23schema (thanks for the live-tweeting, kevin)
  81. [Wed 16:00]     <sandro>        ivan: Kavi, you know a guy in ireland made a mirror page with an RDF version of the schema vocabs, with all the examples in RDFa 1.1 (soon, at least) --- Is it so that we (all of us) can tell the world that these alternative syntaxes work -- if you choose to use RDFa, then that's okay -- because the search engines will treat it the same?
  82. [Wed 16:00]     <sandro>        kavi: I guess this is a repeat
  83. [Wed 16:01]     <danbri>        (he's talking about http://schema.rdfs.org )
  84. [Wed 16:01]     <sandro>        kavi: THere is an underlying addumption that with RDFa you can do a lot of cool things, but with microdata and microformats you can't and we're closing the door to future innovations. I don't agree with that premise
  85. [Wed 16:01]     <manu-db>       Could someone ask Kavi: Do you plan on putting support into the Rich Snippets Testing Tool to display schema.org markup if it's expressed in RDFa?
  86. [Wed 16:02]     <tantek>        manu - have you had a chance to look at microformats 2.0 syntax? http://microformats.org/wiki/microformats-2
  87. [Wed 16:02]     <tantek>        I would be interested in your feedback
  88. [Wed 16:02]     <manu-db>       tantek - no, but we should really talk about that as well
  89. [Wed 16:02]     <sandro>        kavi: When we made this announcement, as three competitors, we couldn't talk about all the things we imagined doing with this data. But we all understand there is more that can be done. If you don't have something to ground the design tradeoffs, then it seems like we have subsumed the semantic web. We're trying to say this is just one thing we accept.
  90. [Wed 16:03]     <tantek>        manu - in short, many (most?) of the issues you encountered in developing hAudio have driven the design of microformats 2.0.
  91. [Wed 16:03]     <manu-db>       tantek - I've been trying to figure out ways to pull the Microformats and RDFa community closer together (it's the whole reason I joined the RDFa group in the first place - to bring both communities closer together)
  92. [Wed 16:03]     <sandro>        kavi: So now we get the opposite flak, "how can you have such small scope". This standard, like others, will have to evolve. People wont follow the spec at scale. All of these thigns will have to change as we get to 100K sites instead of 100.
  93. [Wed 16:03]     <manu-db>       tantek - great to hear - I've put it on my reading list, I will review and get back to you.
  94. [Wed 16:03]     <sandro>        kavi: (something about html)
  95. [Wed 16:03]     <tantek>        manu - one key aspect is to separate syntax from vocabulary
  96. [Wed 16:03]     <sandro>        kavi: Something is better than nothing. It's going to have to change.
  97. [Wed 16:03]     <tantek>        with microformats 2.0 we achieve that
  98. [Wed 16:04]     <tantek>        kavi, those all sound like rationalizations for ignoring/subsuming open standards community work
  99. [Wed 16:04]     <manu-db>       tantek - separate syntax from vocabulary - +1000 (I've always believed that Microformats community should focus on vocabs, not syntax)
  100. [Wed 16:04]     <sandro>        kavi: There is room for many opinions. This is a step on a long path. It's not meant to be a be-all end-all, in terms of either syntax or vocab.
  101. [Wed 16:04]     <sandro>        q: next steps?
  102. [Wed 16:04]     <tantek>        manu - I think there is still benefit to pursuing microformats-like syntax, hence microformats 2.0
  103. [Wed 16:04]     <sandro>        kavi: 1. collecting feedback and bug fixes. lots of stuff is coming in.
  104. [Wed 16:04]     <tantek>        manu but by separating that discussion from vocabulary, we can work on vocabulary as well
  105. [Wed 16:05]     <tantek>        kavi - why is it ok to put Google at the middle of this? Google is not W3C. Google is not IETF.
  106. [Wed 16:05]     <sandro>        kavi: 2. due to alienation, some disccusions were not happening. we're planning to reach out and have conversations.
  107. [Wed 16:05]     <manu-db>       tantek - yes, I agree... I care less about working on Syntax in uF community, more about working on vocabs in uF community.
  108. [Wed 16:05]     <sandro>        kavi: Some of the anger and frustrations I've seen: "why did you go create this new thing, when there's this other thing"
  109. [Wed 16:05]     * danbri        thinks microformats will benefit by making that vocab/syntax distinction
  110. [Wed 16:05]     <tantek>        manu - that's totally fine - I want to make it so that anyone can work on vocabulary in the microformats community *independent* of syntax
  111. [Wed 16:06]     <sandro>        kavi: We didn't try to re-invent at all, we just polished a bit. Obviously we didnt get it all right.
  112. [Wed 16:06]     <manu-db>       tantek - and I think that's a brilliant move.
  113. [Wed 16:06]     <tantek>        kavi's point of "[we're] collecting feedback and bug fixes. lots of stuff is coming in." indicates an implicit acceptance/value as Google at the middle of this.
  114. [Wed 16:06]     <tantek>        and that's the biggest problem I have with schema.org
  115. [Wed 16:06]     <tantek>        it's a subversion of all the open standards community work that *all* of us have done for years
  116. [Wed 16:07]     <tantek>        including tons of people *at Google*
  117. [Wed 16:07]     <danbri>        tantek - has anyone got a prototype microformats2 generic parser?
  118. [Wed 16:07]     <sandro>        gavin: ORA did some early ?data on the web. I've never really felt what the problem with RDFa was. What caused the creation of microdata in the first place? It felt like the RDFa WG addressed the problems, and then we still had microdata.
  119. [Wed 16:07]     <tantek>        people like Chris Messina on ActivityStrea.ms
  120. [Wed 16:07]     <manu-db>       I think that schema.org should be moved to Microformats 2 community (vocab part) or W3C (or both)
  121. [Wed 16:07]     <tantek>        and Joseph Smarr on PortableContacts.net
  122. [Wed 16:07]     <tantek>        danbri - working on it (generic microformats 2.0 parser)
  123. [Wed 16:07]     <tantek>        to produce both JSON and RDF triples (hopefully a bit cleaner than what microdata produces ;) )
  124. [Wed 16:08]     <sandro>        gavin: As a member of the HTML5 WG, I couldn't see the line. Inside google, was there a discussion when this was created? It's been hard to figure out which to use. But now it looks like Google is saying "use microdata". How did Google decide.
  125. [Wed 16:08]     <sandro>        kavi: There is no single or right answer to this question.
  126. [Wed 16:08]     <tantek>        manu - regarding moving schema.org - from a technical perspective I'm taking schema.org as brainstorming proposals
  127. [Wed 16:08]     <minmax>        tantek, it is.. but as a web developer, I have to say... finally ... sorry, w3c was neglecting web developers with rdfa for last X years...
  128. [Wed 16:08]     <sandro>        kavi: Certainly we could dive into the syntactic details.
  129. [Wed 16:08]     <tantek>        and working to see where it is possible to integrate the brainstorm proposals into various works in progress at microformats.org
  130. [Wed 16:08]     * danbri        tantek - if you want an alpha tester, give me a shout
  131. [Wed 16:09]     <sandro>        gavin: As a developer, until schema.org, it felt like two things doing the same thing, and no obvious reason to use one and not the other. And now there is schema.org.
  132. [Wed 16:09]     <tantek>        danbri - excellent - will do !
  133. [Wed 16:09]     <sandro>        unk2: You keep saying you could argue about usage, but no one is using microdata, and it does less than RDFa, and it's no simpler than RDFa 1.1, "but you could argue that it's the way to go".
  134. [Wed 16:09]     <manu-db>       minmax: How was RDFa WG ignoring web developers? We listened to every piece of input we got - primarily because we care about how web developers use this stuff, but also because we're required to!
  135. [Wed 16:10]     <tantek>        if this were the 1990s and Microsoft and Netscape had announced they'd developed the next version of HTML for all of us HTML-MN and dictated it - everyone would have flipped out
  136. [Wed 16:10]     <sandro>        kavi: Instead of just comparing microdata and RDFa and metatags in the head
  137. [Wed 16:10]     <sandro>        recordon: That's just RDFa
  138. [Wed 16:10]     <tantek>        but somehow in the 2010s and it's Google and Microsoft (+ Yahoo Search puppet) and they've developed the next version of semantics for the web and dictated it and people ARENT flipping out?!?
  139. [Wed 16:10]     <sandro>        kavi: I dont want to get into that
  140. [Wed 16:10]     <minmax>        manu-db, ... web developers _do_not_care_ about RDF... it's the wrong starting point...
  141. [Wed 16:11]     <sandro>        kavi: There has been relatively small adoption of RDFa or microdata, on the sites we crawl. The BROAD level theme is that the difference in adoption were not what mattered, but it had to be simple and cover just the use cases we wanted to adopt right away.
  142. [Wed 16:11]     <tantek>        minmax - more adoption of RDFa than microdata on the web would indicate that web developers cared *even less* about microdata than RDFa.
  143. [Wed 16:11]     <manu-db>       minmax: That is a broad, generalized statement - I'm a web developer and I care about RDF. There are many people out there that don't care, there are many that do care. The great thing about the Web is that we're all given a choice to care about what we want to care about.
  144. [Wed 16:11]     <sandro>        ivan: I don't think the discussion of which is better fits here
  145. [Wed 16:12]     <sandro>        unk2: Where do we go from here? cf Peter Mika's study.
  146. [Wed 16:12]     <minmax>        manu-db, I am just trying to defend the point from an entirely other perspective
  147. [Wed 16:12]     <tantek>        minmax - do you have comparable data for your perspective or is it anecdotal?
  148. [Wed 16:12]     <edsu>  manu-db: so using that logic, it's ok for google to care about microdata too?
  149. [Wed 16:12]     <sandro>        Ivan: Kavi you're not giving me a yes or a know. Will you allow RDFa and Microformats to live as equal formats for a few years at least?
  150. [Wed 16:12]     <minmax>        manu-db, and I agree it is overgeneralziation
  151. [Wed 16:12]     <minmax>        tantek, anecdotal...
  152. [Wed 16:13]     <minmax>        tantek, W3C failed to serve web developers with both RDF and RDFa... RDF was hijacked by enterprise market... so pushing it to web is not what web developers care about
  153. [Wed 16:13]     <manu-db>       minmax: That's fine, but your perspective sounds exclusionary to a bunch of people that care about Web development /and/ Graph-based data representations.
  154. [Wed 16:13]     <sandro>        kavi: The goal of schema.org was to reduce the confusion of having several standards that were equal citizens. We *were* letting them play out.
  155. [Wed 16:13]     <minmax>        manu-db, yeah, sorry, I take that back
  156. [Wed 16:13]     <sandro>        kavi: whether OGP is RDFa is going in circles
  157. [Wed 16:14]     <tantek>        folks - here's the data that Google published last time they presented this at *last year's SemTech*
  158. [Wed 16:14]     <tantek>        http://www.readwriteweb.com/archives/google_semantic_web_push_rich_snippets_usage_grow.php
  159. [Wed 16:14]     <manu-db>       edsu: Yes, perfectly fine for Google to care about Microdata - they should care! However, they should also care about RDFa and anything else Web developers *choose* to use.
  160. [Wed 16:14]     <sandro>        kavi: Within google we have people on all sides of these issues. I can say on behalf of Search Quality, that we did not and do not have a horse in this game. It is only a question of what will drive adoption the best. We may have made the wrong decision.
  161. [Wed 16:14]     <sandro>        kavi: This was our motivation.
  162. [Wed 16:14]     <tantek>        kavi, where's the updated data now?
  163. [Wed 16:15]     <danbri>        q: can we identify a common data model across microdata, microformats2 and rdfa such that we can share and mix vocabs across them?
  164. [Wed 16:15]     <tantek>        why not present updated data instead?
  165. [Wed 16:15]     <manu-db>       +1 to Danbri's question - make sure it's asked in person.
  166. [Wed 16:15]     <sandro>        kavi: Could we support more than one format, even if we're documenting just one? That could be a reasonable option. I tried to outline the factors we'd care about that. What kinds of things would it enable?
  167. [Wed 16:15]     <danbri>        (roughly something that smells a bit like rdf without some of the real-or-perceived clutter)
  168. [Wed 16:15]     <tantek>        danbri - that would likely result in trees of objects of property value pairs that get turned into simple triples just for the sake of RDF
  169. [Wed 16:15]     <danbri>        (ie. attribute/value pairs attached to web-identified thingies)
  170. [Wed 16:16]     <tantek>        basically, the common data model is hierarchies of objects, not triples
  171. [Wed 16:16]     <danbri>        the triple view isn't mandatory
  172. [Wed 16:16]     <danbri>        yeah
  173. [Wed 16:16]     <sandro>        ivan: Working through that line would be a way forward. W3C is a player on both formats, and would like the discussion to go on. How can these formats live and find the right messaging, okay for Google... blog items like "you bet on the wrong horse" should not happen.
  174. [Wed 16:16]     <manu-db>       danbri, tantek: Something like JSON-LD? http://json-ld.org/
  175. [Wed 16:16]     <edsu>  danbri: pasting some microdata/html into http://foolip.org/microdatajs/live/ suggests yes :)
  176. [Wed 16:16]     <manu-db>       (we can even make it look exactly like JSON)
  177. [Wed 16:17]     <danbri>        the rdfism leaks in when we ask whether type definitions get to have the last say about properties that can occur on that type
  178. [Wed 16:17]     <danbri>        yeah i was talking to foolip earlier :) great stuff
  179. [Wed 16:17]     <tantek>        folks, it's wrong for Google (or Google+Microsoft) to dictate vocabularies, let's not lose sight of that
  180. [Wed 16:17]     <welty> what is the "perceived clutter" - has anyone characterized that?
  181. [Wed 16:17]     <welty> is it just URIs?
  182. [Wed 16:17]     <tantek>        just like we wouldn't have accepted Netscape+Microsoft dictating HTML-NSMS
  183. [Wed 16:17]     <sandro>        unk3: I wanted to raise the formats issue a different way. the relationships of shcema.org and the formats and to extraction. it seems like schema.org has changed microdata a lot, like adding schema defintion, which microdata didnt have before. DO you expect microdata forlks to use that new schema language (which is a lot like RDFS).
  184. [Wed 16:18]     <sandro>        unk3: itemid in html5 vs schema.org ulr property. So the algo in html5 spec gives you a URL literal property, not the subject you might expect.
  185. [Wed 16:18]     <tantek>        unfortunately in the search space, Google and Microsoft are essentially potential bullies - and the release of schema.org is a bullying of open vocabulary efforts.
  186. [Wed 16:19]     <sophiap>       welty: I would like the structure to be flatter, less hierarchy
  187. [Wed 16:19]     <tantek>        welty - yes - that's been our experience with microformats too - flatter, less hierarchy works better
  188. [Wed 16:19]     <sandro>        kavi: We didnt actually document everything in microdata. In trying to be simple, ... if we could use fewer new attrs, (itemref, itemid, ... maybe those shouldnt have been left out?) In terms of inventing new things, to the best of my memory, we didnt actually change the syntax, but we organized things in a hierarchy
  189. [Wed 16:19]     <danbri>        tantek - do you have specific areas in mind where you think they might bias the vocab?
  190. [Wed 16:19]     <tantek>        frankly I was quite surprised by the hierarchy of schema.org - smelled very 1990s object hierarchy like - like lessons learned since (on the Web) were ignored by the schema.org "experts"
  191. [Wed 16:19]     <welty> the clutter associated with RDF (danbri's term)
  192. [Wed 16:19]     <minmax>        tantek, as a web developer I have to say... finally, :)
  193. [Wed 16:19]     <sandro>        kavi: We've been careful not to reuse property names across types, to give us flexibility.
  194. [Wed 16:20]     <welty> RDF is as flat as you want
  195. [Wed 16:20]     <tantek>        danbri - to put it bluntly, the "experts" that produced the schema.org vocabularies made so many "noob" mistakes that I can't take them seriously
  196. [Wed 16:20]     <danbri>        (the only obvious screwup i could find was spouse being defined as singular; hardly inclusive)
  197. [Wed 16:20]     <sandro>        (unk3 is mike l)
  198. [Wed 16:20]     <minmax>        RDF was largely hijacked by companies serving enterprise market...
  199. [Wed 16:20]     <tantek>        they're about 20 years out of date
  200. [Wed 16:20]     <tantek>        danbri - *tons* of mistakes in there - you weren't looking hard enough ;)
  201. [Wed 16:21]     * danbri        didn't go thru term by term yet, i got distracted by protovis - foaf.tv/tellyclub/schema.org/protovis-3.2/ex/den3.html
  202. [Wed 16:21]     <tantek>        the whole class hierarchy approach is 1990s java-think
  203. [Wed 16:21]     <tantek>        it's obsolete
  204. [Wed 16:21]     <tantek>        we know better now
  205. [Wed 16:21]     <tantek>        with *web-scale* markup requirements
  206. [Wed 16:21]     <manu-db>       danbri - yes, lots of issues in schema.org vocab - things that Microformats/RDFa community would've picked up on very quickly.
  207. [Wed 16:21]     <danbri>        ( the '90s heritage it cites is http://www.cyc.com/opencyc )
  208. [Wed 16:21]     <sandro>        ivan: From a w3c process point of view, but microdata and rdf1.1 are both still working drafts. the portions about the mapping to RDF have bugs reported -- in the WG they have to be settled. I don't think Google will have a problem with that. Neither of these documents is yet a standard.
  209. [Wed 16:22]     <manu-db>       Can somebody ask Kavi if Google or Microsoft plans to take part in either the Microdata or RDFa standardization work
  210. [Wed 16:22]     <manu-db>       We would love to have them on board in the RDFa Working Group
  211. [Wed 16:22]     <sandro>        ivan: Schema.org is out there, ... how do you envisage the process for the future whereby schema.org might be a place where new vocabs are developed. I place to make it a more open social process? Might be at w3c or not, but I'm curious about the process.
  212. [Wed 16:22]     <tantek>        Ivan - no it is unacceptable for Google+Microsoft to hijack open vocabulary development
  213. [Wed 16:22]     <tantek>        bullying should not be rewarded
  214. [Wed 16:23]     <sandro>        kavi: I don't have a great answer right now. I dont think any one company wants to own this in its entirety. By going with 3, we showed we weren't just doing it.
  215. [Wed 16:23]     * danbri        wonders if tantek sees some role for wikipedia/dbpedia for big vocabulary management
  216. [Wed 16:23]     <tantek>        "anyone can continue to develop vocabularies" - Kavi (did I hear that right?)
  217. [Wed 16:23]     <manu-db>       That's not a very re-assuring answer, Kavi.
  218. [Wed 16:23]     <danbri>        "anyone can continue to develop legacy vocabularies"
  219. [Wed 16:23]     <sandro>        kavi: (1) Any one CAN continue to develop vocabs. We might adopt them, maybe with the same syntax, maybe different syntax.
  220. [Wed 16:23]     <tantek>        danbri - microformats.org *is* very wikipedia-like (with process etc.) for vocabularies
  221. [Wed 16:23]     <tantek>        that's the point of microformats.org/wiki/process
  222. [Wed 16:23]     <sandro>        kavi: (2) (missed)
  223. [Wed 16:24]     <danbri>        yup, but wikipedia is wikipedia-scale ... can we (at some point) just cut over to use that?
  224. [Wed 16:24]     <sandro>        kavi: Then it leaves the question of where is the completely open discussion... We don't have an answer yet, but this is important. We'll need to sort out the stuff that's out there.
  225. [Wed 16:24]     <tantek>        danbri - no because wikipedia processes are for content, not vocabularies
  226. [Wed 16:24]     <tantek>        wikipedia processes are not scientific enough to actually do vocabulary development
  227. [Wed 16:24]     <manu-db>       The primary issue with all of this is: If the goal was to speed adoption, it's going to have the opposite effect. This announcement has created a great deal of confusion and uncertainty in the market.
  228. [Wed 16:25]     <tantek>        plus, microformats.org has a better license - Public Domain / CC0
  229. [Wed 16:25]     <tantek>        for standards
  230. [Wed 16:25]     * danbri        agrees re current state, but i think wikipedia is moving in our direction
  231. [Wed 16:25]     <manu-db>       It affects every one of us - Microformats, Microdata, /and/ RDFa.
  232. [Wed 16:25]     <tantek>        danbri - it's not and can't, the CC-SA / GNU license makes it infeasable long term
  233. [Wed 16:25]     <sandro>        unk1: Ours has an edit button, yours has a feedback button. The CORE of microformats is we reach agreement. YOU said "we did it in a closed room". You havent shown your work, your evidence, how others can get involved. This is the most worrying thing.
  234. [Wed 16:25]     <edsu>  manu-db: the omission of facebook at their table suggests to me that might've been part of the intent
  235. [Wed 16:25]     <hober> sandro: s/unk1/KevinMarks/
  236. [Wed 16:26]     <sandro>        kavi: That's a totally valid point. Microformats did a great job creating an open community.
  237. [Wed 16:26]     <sandro>        kavi: There's no good answer for why we didn't do that.
  238. [Wed 16:26]     <tantek>        good to hear that admission
  239. [Wed 16:26]     * manu-db       nods.
  240. [Wed 16:26]     <sandro>        kavi: Coming to microformats with a whole bunch of new things could have been an option. We did want to get something out there.
  241. [Wed 16:26]     <tantek>        at microformats we publicly document our research *while* doing development
  242. [Wed 16:27]     <tantek>        we've learned those lessons!
  243. [Wed 16:27]     <danbri>        is anyone from bing there?
  244. [Wed 16:27]     <tantek>        I don't think so
  245. [Wed 16:27]     <manu-db>       Nobody from Yahoo, either?
  246. [Wed 16:27]     <sandro>        kavi: The comment about "be big to be part of this", we that reaching consensus is generally very slow. We started with orgs that had deep vested interests, so we could reach consensus and get something out there.
  247. [Wed 16:27]     <tantek>        reaching oligopolic consensus is insufficient
  248. [Wed 16:27]     <tantek>        that completely goes against the *open web*
  249. [Wed 16:27]     <sandro>        unk1: It's easy if you define consensus as after excluding 90% of the people who care about it.
  250. [Wed 16:27]     <tantek>        please someone call bullshit on Kavi's consensus among companies line
  251. [Wed 16:28]     <sandro>        kavi: That's a valid criticism.
  252. [Wed 16:28]     <manu-db>       I agree - it may be fine for a first cut, but then you let the community review and give feedback (at the very least)
  253. [Wed 16:28]     <laroyo>        manu-db peter mika0x1c is here from Yahoo!
  254. [Wed 16:28]     <tantek>        manu - I disagree
  255. [Wed 16:28]     <tantek>        it's not even a good first cut
  256. [Wed 16:28]     <tantek>        no documentation of research
  257. [Wed 16:28]     <welty> there are no rules here - you do what works
  258. [Wed 16:28]     <tantek>        no use-cases
  259. [Wed 16:28]     <tantek>        nothing
  260. [Wed 16:28]     <manu-db>       tantek - I didn't say /good/ first cut
  261. [Wed 16:28]     <sandro>        iptc: re "being big enough" and "developing in a box". IPTC did reach out to them, and they did listen. in the news article area. they're not THAT evil, just talk to them. :-)
  262. [Wed 16:28]     <tantek>        just a schema hierarachy from on high
  263. [Wed 16:28]     <sandro>        iptc: which is not to say I'm in favor of it.
  264. [Wed 16:29]     <danbri>        'SEO' is not a use case? :)
  265. [Wed 16:29]     <tantek>        danbri - SEO is a source of spam
  266. [Wed 16:29]     <sandro>        unk1: hRecipe was a great example of how Google can do this.
  267. [Wed 16:29]     <tantek>        "I think the microformats process is a great one." - Kavi (did I hear that correctly? -t)
  268. [Wed 16:29]     <vulcan_wsmith> SEO = Direct Commercial Interest with the power to override every other use case
  269. [Wed 16:29]     <sandro>        kavi: The microformats process is a great one. Getting the details to scale is an open question. Microformats hasnt had tons of new formats recently. Maybe there has to be the proper social mechanisms.
  270. [Wed 16:29]     <vulcan_wsmith> which is a problem, but I don't know if it's google's problem
  271. [Wed 16:29]     <sandro>        unk1: microformats2
  272. [Wed 16:30]     <tantek>        doing good formats / vocabularies is *hard* folks
  273. [Wed 16:30]     <tantek>        most never get past the "do some research" phase
  274. [Wed 16:30]     <sandro>        andr: I heard people say 90% of the people are ignored. the point is not the people in this room. the process leading to RDFa didn't include all those voices either.
  275. [Wed 16:31]     <sandro>        andr: Think about what will bring the most value to those people, not in this room.
  276. [Wed 16:31]     <danbri>        tantek, where is the 'cut' line between vocabs that ought to be done at microformats.org, versus 'domain specific' vocabs done off in separate fora? (lifescience stuff? music ontology?)
  277. [Wed 16:31]     <danbri>        (not meant argumentatively; it seems a genuinely tricky thing to find balance)
  278. [Wed 16:31]     <sandro>        gavin: I think it's UNBELIEVABLE that Bing, Yahoo, and Google have now finally said that semantics in the document help! The fact that we're having this conversation is a good thing. I want to thank you for coming.
  279. [Wed 16:31]     <sandro>        (applause)
  280. [Wed 16:31]     <tantek>        danbri - there is no such cut
  281. [Wed 16:32]     <sandro>        giov: Why didnt facebook join in
  282. [Wed 16:32]     <tantek>        anyone is welcome to start and do research, publicly with public licenses at microformats.org
  283. [Wed 16:32]     <sandro>        kavi: We invited them.
  284. [Wed 16:32]     <singpolyma>    microformats hasn't had a lot of new formats because new formats are not the goal, good ones are :P
  285. [Wed 16:32]     <sandro>        recordon: We dont understand the value in yet another thing new, here.
  286. [Wed 16:32]     <manu-db>       andr, what do you mean "RDFa didn't include all those voices"? Open standards process - open issue tracker - yes, people complained, we made changes for many of those complaints - couldn't for others.
  287. [Wed 16:32]     <manu-db>       andr, to say that people couldn't comment is misrepresenting how the standards process works.
  288. [Wed 16:33]     <sandro>        kavi: How we make things more simple... We simplified in ways, but made things more complex in having more things you can mark up.
  289. [Wed 16:33]     <sandro>        kavi: Even if everyone agrees, there's going to be a long tail not adopting this.
  290. [Wed 16:33]     <tantek>        to be blunt it doesn't sound like we made much progress in this BOF
  291. [Wed 16:34]     <manu-db>       Did Kavi say that they asked Facebook to join and Facebook didn't join?
  292. [Wed 16:34]     <bergie>        yes
  293. [Wed 16:34]     <tlr>   yes
  294. [Wed 16:34]     <sandro>        ivan: People are gathering to use this room for the next sessions. Where do we go from here? Kavi, you can tell there are unhappy people.... Nobody wants unhappy people.
  295. [Wed 16:34]     <tantek>        the arrogance of Google+Microsoft taking a top-down vocabulary approach for the rest of the web is NOT acceptable
  296. [Wed 16:34]     <manu-db>       Did he elaborate on why?
  297. [Wed 16:34]     <bergie>        tantek: +1
  298. [Wed 16:34]     <welty> tantek: -1
  299. [Wed 16:34]     <danbri>        is this channel logged?
  300. [Wed 16:34]     <welty> lighten up
  301. [Wed 16:34]     <manu-db>       Nevermind, seeing David Recordon elaborating - didn't see value in yet another new thing.
  302. [Wed 16:34]     <tantek>        "contact me" is insufficient
  303. [Wed 16:35]     <tantek>        Kavi - bring your research to microformats.org
  304. [Wed 16:35]     <bergie>        manu-db: FB said they didn't see reason for yet another format that doesn't make anything easier
  305. [Wed 16:35]     <tantek>        and contribute it to the public domain / CC0
  306. [Wed 16:35]     <manu-db>       welty - Tantek makes a very valid point.
  307. [Wed 16:35]     <sandro>        kavi: That's the right question. It's not about who is happy, but offering something constructive. Anyeone in this room can contact me.
  308. [Wed 16:35]     <welty> which point?
  309. [Wed 16:35]     * tlr   keeping local logs; I think we'll put some minutes up on the web
  310. [Wed 16:35]     <tantek>        No - moving the discussion to Google is WRONG
  311. [Wed 16:35]     <sandro>        kavi: there is a dicussion group at google, linked from schema.org
  312. [Wed 16:35]     <tlr>   +1 to tantek
  313. [Wed 16:35]     <manu-db>       That this way of doing "standards development" is not how we do things on the *Open Web*
  314. [Wed 16:35]     <tantek>        manu-db +1
  315. [Wed 16:36]     <tantek>        Hey Kavi - do you see what you've done here?
  316. [Wed 16:36]     <welty> the point of the open web
  317. [Wed 16:36]     <sandro>        unk5: we'll talk about this more at the Rich Snippets session
  318. [Wed 16:36]     <welty> is that anyone can do anything
  319. [Wed 16:36]     <welty> and whatever works, wins
  320. [Wed 16:36]     <sandro>        tantek, Kavi hasn't been able to look at IRC, no.
  321. [Wed 16:36]     <manu-db>       part of that is the freedom to chose
  322. [Wed 16:36]     <tantek>        You've gotten a community leader of microformats.org (myself) and chair of W3C RDFa WG to *agree*
  323. [Wed 16:36]     <sandro>        DONE
  324. [Wed 16:36]     <manu-db>       without having stuff dictated to us by a few
  325. [Wed 16:36]     <sandro>        (ADJOURNED)
  326. [Wed 16:36]     * manu-db       chuckles.
  327. [Wed 16:36]     <edsu>  tantek: see, that's progress :)
  328. [Wed 16:36]     <danbri>        :)
  329. [Wed 16:36]     <manu-db>       Yes - both RDFa and Microformats communities agree - sky will be falling, next.
  330. [Wed 16:37]     * danbri        thanks laroyo for the audio
  331. [Wed 16:37]     <welty> thanks sandro
  332. [Wed 16:37]     <welty> ttly
  333. [Wed 16:37]     <manu-db>       (shouldn't say 'communities'... at least Tantek and I agree)
  334. [Wed 16:37]     <danbri>        tantek, we should talk more :) i'll check out the mf 2 stuff
  335. [Wed 16:37]     <danbri>        it looks like we have beginnings of an idea for a common data model too
  336. [Wed 16:37]     <manu-db>       tantek - I'd like to follow up with you, Kevin Marks and Chris Messina at some point.
  337. [Wed 16:37]     <danbri>        one that rdf people can see as triples, and others see as groups of attribute/value pairs ...?
  338. [Wed 16:37]     <tantek>        manu - let's keep blogging
  339. [Wed 16:38]     * manu-db       nods.
  340. [Wed 16:38]     <tantek>        tantek is now offline from Skype
  341. [Wed 16:38]     <manu-db>       Have two in the pipe now
  342. [Wed 16:38]     <danbri>        +1 for public conversations rather than 'follow up withs'
  343. [Wed 16:38]     * danbri        heads off, bye folks
  344. [Wed 16:38]     * manu-db       thanks everyone for making it to the Structured Data in HTML meeting.
  345. [Wed 16:39]     <manu-db>       Great to see so much interest and turnout.
RAW Paste Data