Around 2001 I was deeply fascinated by the potential of graphs and their visualization to save the world, etc.
It seemed to me that if you just found the right layout algorithm, all our visualization problems would be solved.
A couple of months ago I read (can't remember where) a cry "No more force-directed graph visualizations please!" and I immediately thought, yes please.
Today, I found an article by hypertext maven Mark Bernstein, that includes the following screenshot:
Really, that doesn't look useful, not even a little bit.
(I'm worried that I come across as an anti-progress apologist, like John Gruber. I'm not. I really like new stuff, and think it's important to push the state of the art, but in this case, I think we have failed.)
Maybe the whole premise is flawed. Maybe hypermedia thrives on invisibility, the fact that you never know where the next link will take you.
And maybe, the desire for general graph visualization is a desire for a deus ex machina, that can never be fulfilled. Looks this way to me.