Not technically a tree

[Scott Blomquist]

I originally proposed that a puzzle's solution can be represented as a tree. Someone in the audience at GC Summit last night corrected me: technically, it's a directed graph that's not necessarily a tree. Since there's fairly broad consensus that puzzles exist in layers, the graphs will frequently be tree-like, but it's entirely possible for nodes to have multiple inbound edges.

--
http://scott.blomqui.st