W3C home > Mailing lists > Public > www-rdf-interest@w3.org > August 2003

RE: practical problems with rdf:parseType="Collection" implementation

From: Joshua Allen <joshuaa@microsoft.com>
Date: Thu, 14 Aug 2003 12:37:26 -0700
Message-ID: <4F4182C71C1FDD4BA0937A7EB7B8B4C10A44365F@red-msg-08.redmond.corp.microsoft.com>
To: "Garret Wilson" <garret@globalmentor.com>, <www-rdf-interest@w3.org>

Your complaint really seems to be based on variations of the same theme;
namely:

> 1. Empty lists cannot be modified.

I have to admit, I had the same complaints.  But it ends up not being as
big a problem as you first think.  When I implemented an object model
for Rdf Collection, I just implemented IList interface, which for .NET
is basically the same as what you are trying to do in Java.  The only
caveat is that you need to spec that the collection can never be empty,
and then provide an object that your users can use to assign "empty
collection" to a predicate/subject pair when necessary.  Everything else
can behave exactly the same as all other collections on the system; your
list can support Insert, Append, Delete, foreach (or iterator in your
case), and so on.

IMO, the issue that collections in RDF are recursive is really just an
implementation detail.  That's just the storage model used by some data
structures in your app.  There is no reason you have to surface that as
the primary interface for your API/Query interface if you don't want to.
I agree most users are not Lisp-savvy, so it's fine if you want to
provide a friendlier interface..
Received on Thursday, 14 August 2003 15:37:36 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:52:01 GMT