Re: blank nodes ….

The major motivation for avoiding blank nodes is that you cannot reliably
refer to those objects from outside the dataset. It precludes any kind of
linking or argumentation. In Bio2RDF, we generate URIs for all objects, and
use the identifier when provided, otherwise we generate one when there is
none to use. The main criteria is that it is unique within the URI
namespace. Bonus if you can generate the same URI for exactly that item in
a subsequent version of the dataset (we recognize that it might not be
feasible)

m.


On Wed, Apr 3, 2013 at 3:30 PM, Kingsley Idehen <kidehen@openlinksw.com>wrote:

>  On 4/3/13 3:12 PM, Jeremy J Carroll wrote:
>
>  > There is, I think, a sustainable argument that blank nodes are not **necessary**.
>
>  I beg to differ. Blank nodes are absolutely necessary.
>
>
> "Horses for courses" again.
>
> Blank nodes are only problematic when misunderstood and used
> inappropriately. No different to misusing a Pronoun due to lack of
> understanding when learning a new language that has Pronouns.
>
> A lot of these issues will whither away as RDF's broader pitch narrative
> improves, which has started to happen in recent times :-)
>
>
>
> --
>
> Regards,
>
> Kingsley Idehen	
> Founder & CEO
> OpenLink Software
> Company Web: http://www.openlinksw.com
> Personal Weblog: http://www.openlinksw.com/blog/~kidehen
> Twitter/Identi.ca handle: @kidehen
> Google+ Profile: https://plus.google.com/112399767740508618350/about
> LinkedIn Profile: http://www.linkedin.com/in/kidehen
>
>
>
>


-- 
Michel Dumontier
Associate Professor of Bioinformatics, Carleton University
Chair, W3C Semantic Web for Health Care and the Life Sciences Interest Group
http://dumontierlab.com