Re: Editors' proposal for membership predicate names

On 2/21/14 9:45 AM, Roger Menday wrote:
>
> does the attached picture help this discussion ??
>
> Roger
>
>
> On 21 Feb 2014, at 14:10, Roger Menday wrote:


Do you have a URL for the source file? Either way, would you be able to 
apply the following tweaks:

1. append (to be ultra obvious) the literals "document" to the labeling 
of each each entity of type: Document
2. use something else (rather than a box) to depict entities that aren't 
of type: Document
3. use an oval to depict predicates that associate relation subjects and 
objects (using a labelled arrow is an inherited inaccuracy that warps 
RDF in really bad ways).

Re #3 it would be:

(#subject) --> (#predicate) --> (#object)  (note: oval for reference and 
rectangle of square for a literal object)

Instead of the overused but utterly broken (which undermines the 
importance and nature of what predicate URIs denote):

(#subject) --- predicate --> (#object) .


Great idea introducing an illustration, there are many murky issues that 
will come to the surface as this illustration evolves.

[1] http://bit.ly/1fluti1 -- RDF described and illustrated
[2] http://slidesha.re/1jkDBXj -- Relationship Roles .


Regards,

Kingsley Idehen	
Founder & CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter Profile: https://twitter.com/kidehen
Google+ Profile: https://plus.google.com/+KingsleyIdehen/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen