> dedicating an entire page to such a small piece of trivia feels like
> overkill.

Try it (you can use ontoworld.org if you like) and see.

This is another topic that's come up before, see the long thread on 
semediawiki-user
   Subject: Modeling "third party" relations on Semantic Mediawiki page?
that morphed into
   Subject: Creating Triples Anywhere in a Semantic Wiki
<http://www.mail-archive.com/[EMAIL PROTECTED]/msg00354.html>

SMW's fundamental model is the subject of a semantic annotation is the 
entity described by the article where the annotation appears.  What's 
the harm of creating pages "Earth-0 (DC world)", "Earth-16 (DC world)", 
etc. so that not only can you generate 
http://en.wikipedia.org/wiki/List_of_DC_Multiverse_worlds, but also 
these articles can participate in Special:Browse, "Pages using the 
property Inhabitants”" lists, etc.?  Wikipedia has a bias against "stub" 
articles but a SMW wiki is going to have lots of pages with just a few 
semantic annotations, whose raison d'ĂȘtre is to provide material for 
semantic searches.

To ease organization you could make these subpages with a slash:
"DC worlds/Earth-0", "DC worlds/Earth-16", etc.  Mediawiki displays a 
breadcrumb back to the top page, and there's a {{List subpages}} 
template on  Wikipedia that you can copy which lists all subpages.  Note 
that by default Mediawiki does not enable the subpage feature in the 
main namespace; read http://meta.wikimedia.org/wiki/Subpages#Administration
for the LocalSettings.php change that enables this.


You're welcome to try to code both your ideas:
* Using the subsection's anchor name as subject of link
* Using "tag which isolates a portion of the page from the rest of it 
for the purpose of semantic annotations"
and see how you get on.  You will no doubt run into issues with factbox 
display, query semantics, parsing, database representation, collisions, etc.

> I'm not (yet) familiar with how the database is set up;

You can use something like phpMyAdmin to look at the wikidb tables in 
your installation, or read function setup() in 
<http://svn.wikimedia.org/viewvc/mediawiki/trunk/extensions/SemanticMediaWiki/includes/storage/SMW_SQLStore.php?revision=HEAD&view=markup>
 
to just see the SMW table format.

>  is there
> anything that forbids hash-marks in page names?

* From a parsing perspective:
   MediaWiki's parsing of a wiki link puts the section after the 
crosshatch in the hyperlink in HTML, so that your browser will jump to 
the right place when you click it.  But both MediaWiki's pagelinks table 
and SMW's smw_relations table don't store this section information.
* From a design perspective:
   Maybe coincidentally, the URI spec 
http://www.ietf.org/rfc/rfc2396.txt states that the fragment identifier 
after the crosshatch is *not* part of a URI, it's part of the reference 
to a URI.
* From an implementation perspective:
   SMW probably assumes annotation subject == (page title === page ID), 
so you'd have to either correct this assumption or add a separate page 
subsection.

MediaWiki is hack upon hack, so don't let any of this stop you trying ;-)

> you might include
> the text '[[Celestial class#Dysnomia::Moon]]', to semantically
> indicate that Dysnomia is a moon. 

That's confusingly different from subject-predicate-object order of RDF 
triples, something like [[#Dysnomia::Celestial class::Moon]] would be 
much better.

Enjoy wiki hacking or code hacking,

--
=S Page

-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Semediawiki-devel mailing list
Semediawiki-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/semediawiki-devel

Reply via email to