Hi John,

Thanks for the reply.

On 31 May 2006, at 11:39, john.breslin wrote:
>> The distinction between Site and Forum is awkward,
...
> Of course you don't have to use the Site concept if you don't want
> to, but
> the concept of Site can be used for multiblog sites, BBs with multiple
> forums, etc. Sites could also host different communities of blogs,
> and
> hence the previous discussion about introducing a "Community"
> concept where
> Sites, Forums, Posts, Users etc. could all be "part_of" a community.

As I said before, I'd much rather use Community than Site.

>> sioc:container_of is awkward, sioc:post would be nicer.
>
> Yes, perhaps this is my fault. To not confuse properties and
> classes , and
> to avoid affording any prominence to the concept on either end of the
> property, we used different names for the connecting properties.
> So Forum
> is container_of Post (rather than Forum -> post -> Post or Post ->
> forum ->
> Forum).

Yes, understandable. hasPost, inForum would have worked maybe.

>> I don't like all the duplication of DC and FOAF terms. Not invented
>> here syndrome? You've even reinvented rdf:type.
>
> It may not be reflected in the spec but we've eliminated some of these
> duplicate terms in the NS. But some can remain as subproperties of
> DC, just
> as RSS does...

Yes, that's great. Subproperties are fine.

>> The spec talks in length about a Types module, but I can't find a
>> link to it. http://rdfs.org/sioc/types/ is mentioned in the comment
>> for sioc:type but is 404.
>
> Oops! Not done yet... The idea here is that you could create a
> sioc:Forum
> and say it of type blog, mailing list, USENET group etc.

+1.

>> 2. SIOC is not very suited for modelling other types of fora.
>
> Tell us why, apart from the comments above.

I think the main point is that there are lots of terms that I would
want if I modelled a BB (avatar, number of views, is_closed), but no
terms at the same level of details for blogs (Weblog, comment, feed)
or mailing lists (list address, to/cc). So the vocabulary is biased
towards BBs. The bias could be removed by moving BB-specific stuff
into a separate vocabulary, and adding additional blog-specific and
list-specific and newsgroup-specific and wiki-specific vocabs. (No,
I'm not volunteering ;-)

>> 4. SIOC is not a good RDF citizen.
>
> We're trying to be better! I think we also have to understand that
> non
> Semantic Web people sometimes just don't get the fact that they
> have to use
> three vocabularies instead of one. We're missing the subclassing
> things I
> mentioned above in our mappings file but doing so somewhat appeases
> both
> sides I think.

I'm happy to hear that you want SIOC to be RDF-friendly. My proposed
changes are all motivated by the desire to have interoperable RDF
data and implementations, but I'm perfectly aware that you have to
consider a number of other factors such as acceptability to non-RDF
folks and existing SIOC toolkits and your overall strategy. One has
to find a balance and compromise.

Cheers,
Richard

>
>> In summary, I'd use it again when I have to model a BB but not
>> otherwise. Just my �0.02.
>
> Well, hope to convince you otherwise in this thread! Looking
> forward to
> hearing back from you...
>
> >
>

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "SIOC-Dev" group.
To post to this group, send email to sioc-dev@googlegroups.com
To unsubscribe from this group, send email to sioc-dev-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sioc-dev
-~----------~----~----~----~------~----~------~--~---

Hi Alex,

Thanks for the response! Some comments inline.

On 30 May 2006, at 21:06, Alexandre Passant wrote:
>> The distinction between Site and Forum is awkward, I don't want to
>> introduce a top-level pseudo-forum for the weblog. I'm not convinced
>> that Site and the host properties provide any value.
>
> Even if it can be strange to introduce this Site / Forum
> distinction for a single-blog, this is useful when a site is
> hosting different fora.
> Eg, when modeling a multi-user blog, you have to make the
> distinction between the site (sioc:Site), and users' blogs
> (sioc:Forum).
> BTW, We also think of a Community class, to represent distributed /
> aggregated Sites.

I like sioc:Community much more than sioc:Site. I think it's more
flexible (just about anything can be called a community if you squint
the right way) and more interesting than sioc:Site. I'd be happy with
dropping sioc:Site and using sioc:Community instead. Single-user
blogs don't have a community per se, so I don't need to create two
distinct resources in that case.

> A new version of the ontology should be published soon, where FOAF
> terms have been removed (and so are used directly in FOAF).
> Maybe we should look closer at a DC/SIOC mapping.

Great!

> Well, hope you'll find the new version more "RDF-friendly" :)
> There was a post previously on the ML about the changes, maybe
> you've got some comments / ideas about it. [1]

Thanks for the pointer, I will comment on that in a separate post.

> BTW: Is the SIOC data you created available anywhere ?

Not really. I'm the lead developer of D2R Server [1], a generic
database-to-RDF/SPARQL mapping tool written in Java. I've created an
RDF mapping of the Wordpress DB schema as a demo. The mapping file is
at [2]. It's mostly DC and FOAF and SKOS, with SIOC only for the bits
that are not covered by those. It's quick and dirty and could
certainly be improved, so if there are any Wordpress users here who
feel like playing a bit with a SPARQL service over their weblog
data ... ;-)

I could put the RDF dump of my personal weblog online if that would
help you with interop testing.

Cheers,
Richard

[1] http://www.wiwiss.fu-berlin.de/suhl/bizer/d2r-server/
[2] http://www.wiwiss.fu-berlin.de/suhl/bizer/d2r-server/mappings/
wordpress.ttl

>
> [1] http://groups.google.com/group/sioc-dev/browse_thread/thread/
> b8e3f8ceb73b29f5
> Cheers,
> Richard
>
>
> >

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "SIOC-Dev" group.
To post to this group, send email to sioc-dev@googlegroups.com
To unsubscribe from this group, send email to sioc-dev-unsubscribe@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sioc-dev
-~----------~----~----~----~------~----~------~--~---