Joshua O'Madadhain

Responding to the notion of using someone else's library for analogues to Dimension/Point2D: (1) As a general statement, I would prefer to keep the _number_ of dependencies that JUNG has...

@jbduncan re: option (1), my basic idea is to have a top-level Builder class that you use to get the basic instance for the algorithm and then have specialized Builder...

Responding to @tomnelson : > I'm starting with the notion that Layout is not part of the jung-visualization module. Maybe there are cases where someone wants to make a graph,...

@tomnelson We've gotten along (slowly) without spatial data structures for this long, we can continue to do so for as long as we need to; it's a question of where...

@tomnelson I'm not particularly advocating for getting rid of the capability to update visualizations while the layout is updating, I'm just trying to [encourage us to] explore the design space...

_Something_ has to hold the graph instance for visualization purposes. I agree that, ideally, there would be only one such instance so that we don't have to worry about keeping...

Thanks for the note; I wasn't aware of this issue. The way that we're likely going to resolve this in 3.0 is to get rid of the `graph-impl` jar entirely;...

@jbduncan I think that this is something that we're likely to do in `common.graph`, so I'd rather do it there and let JUNG use that. (There's actually an internal implementation...

If you've got time, please go ahead and file the issue against Guava (that's what I meant by my comment). Otherwise I'll get to it eventually. Side note: I haven't...

Hey, @jbduncan, good to hear from you. :) FYI, I am still paying attention to this project, and I do still intend to release 3.0 (i.e., not just a snapshot...