Comment 3 for bug 703933

I confirm that we need two different measures, as it is openly discussed in this paper:

http://toreopsahl.com/2010/04/21/article-node-centrality-in-weighted-networks-generalizing-degree-and-shortest-paths/
https://thetore.files.wordpress.com/2010/04/node_centrality_in_weighted_networks1.pdf

We can provide the generalized degree of Barrat et al (2004), which is the sum of weights, described in this paper:
http://www.cpt.univ-mrs.fr/~barrat/pnas101.pdf