On 4/11/07, Alexandre Passant wrote:
> On 4/11/07, Diego Berrueta wrote:
> > El mar, 10-04-2007 a las 22:42 +0200, Sergio Fdez escribió:
> > > What do you think...?
> >
> > I completely agree. Indeed, we should push to get these properties
> > accepted into the SIOC specification ;-)
>
> +1

+1 - apparently there is support for including these properties in SIOC core.
Sergio, is that OK with you?

I've added these properties to the "work" version [1] of the SIOC
ontology. Will publish it once we have agreed on the details. Property
names (according to SIOC naming convention): next_by_date and
previous_by_date. Please look at their descriptions and suggest
improvements.

> It also fits well with the new "Thread" container that aims to order posts.

Thread does not really order posts, it just groups them. Please find
some comments on sorting posts at the end of this email.

> BTW, what about making this property an owl:TransitiveProperty ?
> Could be useful for reasoners.

Done. Can it cause any problems? There were some OWL changes to FOAF
ontology (can't remember what exactly) which were later commented out
due to some problems.

> Finaly, re. the names of the properties: won't previous / next be enough ?
> Indeed, it can be previousByAlphabeticOrder or something else, but in
> the case of posts, previous / next "natural" meaning is chronological,
> so maybe the short name is enough.

"next_by_date" adds useful context, important to understanding how and
why are these items ordered.

In any case, next_by_date / previous_by_date acts as a "shortcut" or a
way of describing results of sorting performed on some container. A
more elaborate way to express sorting on a container would be to
define a "sort_by" property. It could then point to a property [which
data creator suggests] to sort by (e.g., dcterms:created).

How does AtomOWL handle sorting of things?

"Representing Order in RDF" [2] gives an good overview of different
ways how order can be represented in RDF. That may also give some
ideas.

[1] http://sw.deri.org/svn/sw/2005/08/sioc/ontology/ns.rdf
[2] http://infolab.stanford.edu/~stefan/daml/order.html

Uldis

[ http://sioc-project.org/ ]

--~--~---------~--~----~------------~-------~--~----~
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?hl=en
-~----------~----~----~----~------~----~------~--~---

Actually this makes a lot of sense because there is a difference between
true threaded posts where one is a reply to another, and others where
there is no guarantee that the next one by date is actually a reply to
the previous one.

e.g. a true threaded discussion from a mailing list or a BB like
vBulletin with threading enabled could have:

Post1 -> has_reply -> Post2 -> has_reply -> Post3
Post1 -> has_reply -> Post4

whereas an IRC chat or even a phpBB discussion may just have:

Post1 -> next_by_date -> Post2 -> etc.

I support the proposition, and propose we add next_by_date and
prev_by_date to SIOC.

--~--~---------~--~----~------------~-------~--~----~
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?hl=en
-~----------~----~----~----~------~----~------~--~---