IRC meeting minutes - 03/28

[Svante Schubert <Svante.Schubert@Sun.COM>]

[Wed:17:09:59] <Svante> Agenda:
[Wed:17:09:59] <Svante> 1. Roll Call
[Wed:17:09:59] <Svante> 2. The current discussion of the field(s) we 
want; precise semantics: Update from Svante.
[Wed:17:09:59] <Svante> 3. Do we allow QNames for attribute content? 
Current answer in proposal is "no." : To be discussed.
[Wed:17:09:59] <Svante> 4. Do we include suggestions for serialization? 
To be discussed.
[Wed:17:09:59] <Svante> 5. How do we deliver non-normative 
documentation? To be discussed.
[Wed:17:09:59] <Svante> 6. Deciding on the ODF RDF vocabulary. To be 
discussed.
[Wed:17:15:52] <flr> wrt. 4: "Normal form" of RDF?
[Wed:17:21:11] <Svante> Bruce: suggested to use HTTP URL for namespaces
[Wed:17:22:03] <flr> two types of new elements:
[Wed:17:22:17] <flr> meta:* (reuse of existign urn-based namespace)
[Wed:17:22:44] <flr> manifest:* (new namespace needed, probaly HTTP-based)
[Wed:17:23:13] <flr> Q: use http-based url for meta:about, etc ?
[Wed:17:23:57] <flr> So we need a new prefix: Bruce suggest m:* and 
mm:manifest..
[Wed:17:24:43] <Svante> attributes would be called m:about m:property
[Wed:17:25:39] <flr> m=meta, mm=metamanifest
[Wed:17:26:29] <flr> discussion on 6.
[Wed:17:27:04] <Svante> <rdf:Description 
rdf:about='urn:uuid:224ab023-77b8-4396-a75a-8cecd85b81e3'>
[Wed:17:27:04] <Svante>    <rdf:type 
rdf:resource="http://www.w3.org/2006/vcard/ns#"/>
[Wed:17:27:04] <Svante>    <rdfs:subClassOf 
rdf:resource="http://docs.oasis-open.org/office/2007/03/meta/rdf#element"/>
[Wed:17:27:04] <Svante>    <odf:name>style.xml#some_id</odf:name>
[Wed:17:27:04] <Svante>    <ex:editor>
[Wed:17:27:04] <Svante>        <rdf:Description>
[Wed:17:27:04] <Svante>            <ex:fullName>James Bond</ex:fullName>
[Wed:17:27:04] <Svante>        </rdf:Description>
[Wed:17:27:04] <Svante>    </ex:editor>
[Wed:17:27:04] <Svante> </rdf:Description>
[Wed:17:27:04] <Svante> Styles.xml file:
[Wed:17:27:04] <Svante> <style:master-page style:name="Standard" 
style:page-layout-name="pm1">
[Wed:17:27:04] <Svante>    <style:footer>
[Wed:17:27:04] <Svante>        <text:p xml:id="some_id"
[Wed:17:27:04] <Svante>                text:style-name="fs">Some 
text</text:p>
[Wed:17:27:04] <Svante>    </style:footer>
[Wed:17:27:04] <Svante> </style:master-page>
[Wed:17:28:55] <flr>  <text:p xml:id="style.xml#some_id" 
text:style-name="fs">Some text</text:p>
[Wed:17:30:16] <Svante> Last week:
[Wed:17:30:16] <Svante> [Wed:16:43:18] <EliasT> meta:about="urn:uuid:123"
[Wed:17:30:16] <Svante> [Wed:16:43:32] <EliasT> urn:uuid:123 a odf:Element ;
[Wed:17:30:16] <Svante> [Wed:16:51:32] <EliasT> urn:uuid:123 odf:name 
"content.xml#foo" .
[Wed:17:30:16] <Svante> [Wed:16:53:47] <EliasT> 1. xml:id
[Wed:17:30:16] <Svante> [Wed:16:54:02] <EliasT> 2. meta:about to always 
an absolute URI
[Wed:17:30:16] <Svante> [Wed:16:55:44] <flr> eliast: "encourage" 
programs to generate unique URIs across documents
[Wed:17:34:42] <flr> Action Item for flr: post my understanding of the 
referencing scheme...
[Wed:17:36:59] <flr> <my:property>my:value</my:property> ????
[Wed:17:37:25] <Svante> m:property="vcard:FullName"
[Wed:17:37:51] <Svante> instead of using 
m:property="http://www.w3.org/2006/vcard/ns#FullName";
[Wed:17:38:36] <Svante> Could we use something simnilar as:
[Wed:17:38:39] <Svante> http://www.w3.org/2006/vcard/ns#FullName 
owl:sameAs vcard:FullName
[Wed:17:38:39] <Svante> in RDF/XML?
[Wed:17:40:00] <flr> Action Item for Svante: check what the RDF/XML spec 
says
[Wed:17:40:14] <darcusb> answer: no
[Wed:17:41:19] <flr> Discussion of 2:
[Wed:17:47:52] <flr> (id1, pred1, data1) (id1, pref2, data2)
[Wed:17:48:08] <flr> <meta-field refs to id1>
[Wed:17:48:49] * pdurusau has joined #odf-metadata
[Wed:17:48:51] <flr> <meta-field meta:resource="id1">
[Wed:17:49:18] <flr> id1 is a full URI
[Wed:17:49:46] <flr> what about anonymous nodes?
[Wed:17:52:14] <flr> <meta-field><rdf:Description pred1="data1" 
pred2="data2"/>...</meta-field>
[Wed:17:52:53] <flr> (_:1, pred1, data1) (_:1, pred2, data2)
[Wed:17:54:06] <flr> <text:meta-field field:type="http://odf.org/Citation";>
[Wed:17:54:06] <flr> (
[Wed:17:54:06] <flr>   <text:meta-field meta:about="urn:isbn:34879823">Doe
[Wed:17:54:06] <flr> 1999</text:meta-field>,
[Wed:17:54:06] <flr>   <text:meta-field
[Wed:17:54:06] <flr> 
text:field-style="http://ex.net/short";>2001</text:meta-field>
[Wed:17:54:08] <flr> )
[Wed:17:54:10] <flr> </text:meta-field>
[Wed:17:54:33] <Svante> There are some attributes that are not specified..
[Wed:17:54:50] <Svante> My wish was:
[Wed:17:54:50] <Svante> <text:meta-field xml:id="someID">
[Wed:17:54:50] <Svante> (
[Wed:17:54:50] <Svante>  <text:meta-field meta:about="urn:isbn:34879823" 
meta:property="http://odf.org/Citation#long>Doe 1999</text:meta-field>,
[Wed:17:54:50] <Svante>  <text:meta-field meta:about="urn:isbn:34898789  
meta:property="http://odf.org/Citation#short>2001</text:meta-field>
[Wed:17:54:50] <Svante> )
[Wed:17:54:50] <Svante> </text:meta-field>
[Wed:17:55:44] * flr does not really understand this ;-(
[Wed:17:56:42] <darcusb> Single reference: <text:meta-field 
meta:about="urn:isbn:34879823">(Doe, 1999)</text:meta-field>
[Wed:17:57:12] <Svante> Bruce: Definition is contrary here used 
different than in the other elements
[Wed:17:57:49] <flr> (_:1 http://odf.org/Citation#long 
urn:isbn:34879823) (_:1 http://odf.org/Citation#short urn:isbn:34898789)
[Wed:17:58:02] <flr> and <meta-field ref to _:1> ??
[Wed:18:01:55] <Svante> This can be solved by using RDF/XML for now:
[Wed:18:01:56] <Svante> <text:meta-field xml:id="someID">
[Wed:18:01:56] <Svante> (
[Wed:18:01:56] <Svante>  <text:meta-field xml:id="someID-X">Doe 
1999</text:meta-field>,
[Wed:18:01:56] <Svante>  <text:meta-field 
xml:id="someID-Y">2001</text:meta-field>
[Wed:18:01:56] <Svante> )
[Wed:18:01:56] <Svante> </text:meta-field>
[Wed:18:02:36] <Svante> Just as a reminder, we do not have to define 
everything in ODF 1.2, we continue working on this.
[Wed:18:05:34] <flr> 
<meta:metafield><rdf:Description><citation:long>dd</citation:long><citation:short>...</citation:short></rdf:Descriptin>...</meta-field>
[Wed:18:06:28] <Svante> Why not separating the RDF/XML from content?
[Wed:18:06:31] <flr> 
<meta:metafield><XXX><citation:long>dd</citation:long><citation:short>...</citation:short></XXX>...</meta-field>
[Wed:18:07:40] <flr> <meta:metafield citation:long="..." 
citation:short="...">...</meta-field>
[Wed:18:09:08] <darcusb> <text:meta-field 
m:resource="urn:uuid:10229293829387">(Doe, 1999, 2000)</text:meta:field>
[Wed:18:09:47] <Svante> Is the same as
[Wed:18:09:48] <Svante> <text:meta-field xml:id="foo">(Doe, 1999, 
2000)</text:meta:field>
[Wed:18:09:51] <Svante> using RDF/XML
[Wed:18:10:59] <darcusb> No, it's not, precisely because we've had to 
introduce a new--more complex --indirection mechanism to get aroun 
dproblems with xml:id.
[Wed:18:11:31] <darcusb> Imagine what an xslt would do to transform the 
citation fields with the xml:id approach?
[Wed:18:11:46] <Svante> Using xsl:keys?
[Wed:18:11:58] <darcusb> Yes, but you know that it's not pretty :-)
[Wed:18:12:13] <darcusb> You have to do more than one lookup.
[Wed:18:12:30] <Svante> I think it is even the fastest way and this is 
pretty ;-)
[Wed:18:13:07] <darcusb> OK, gotta go work ... bye.