Originally reported on Google Code with ID 396
Right now, an edge action block takes as parameter only the model of the edge.
The model of the edge is mainly targeted to identifying the start and end node, but
it many times is
not enough to do something with the edge afterwards.
By providing the start and end node in the action block, we take advantage of the already
computed
nodes.
To not break existing code, an idea would be to use something similar to glamourValue
through
which the block parameters are optional.
Reported by tudor.girba on 2010-05-05 23:17:38