(tye)Re: converting complex netrange to cidr notation

For a single range, the above solution is the minimal one even if you allow arbitrary masks (though I haven't rigorously proved that -- just supporting your suspicion). If you allow multiple ranges, then the problem gets much harder and it becomes possible for arbitrary masks to provide smaller solutions.

        - tye (but my friends call me "Tye")

[dino]

As a footnote I checked the above against
http://www.csc.fi/english/funet/calc/laskin2.html
The results agree ;-)
tbfn
Dino