Of course, you're correct. I chose to ignore that aspect since Boyd's book covers a lot of basic material, and I'm not actually aware of any such text covering non-convex methods in general ( and considering how fast research is in this space, I'm not sure of any general work, in contrast to work like [1] ).