Network integrity and non-random removal of nodes

Hmm - not sure on who owns the trees, but if anyone does know it would
be very useful. Most of the work I've read say the connection
distribution follows a power law, and is why I did not think it would
follow a linear pattern, but just guess work on my part. There has been
some work at BU that says all the power law stuff is just sampling bias
inherent in traceroute type of data aquisition. In which case
vulnerbility might not be as dire as some of the early reports. How to
get an accurate topology is the real big problem, but no need to reopen
the private sector vs government vs academia debate.