[ryanwaggoner]

We already do spam and malware filtering at both of those levels...how is this different? I'm not saying that seizure alone would be enough to intervene at these levels, but if Libya decided to hijack these links for nefarious purposes, then yeah, I don't have a problem with it.

[bigiain]

Yeah, if Libya went down the "nefarious purposes" path, we should probably filter at the browser level.

But think down the path you'd need to go if you chose to block bit.ly at the DNS level... Right now, IANA's root name servers delegate all .ly requests to name servers under Libya's control. I'm pretty sure there's no mechanism in place for the root servers to change that to "send all .ly requests to Libya _except for_ bit.ly requests, and serve them ourselves or send them elsewhere", and in my opinion there _shouldn't_ be any such mechanism because of the potential for abuse at government levels - would you want the Chinese government to know IANA has the ability to redirect traffic from domains they don't like?

So the _other_ alternative seems to be to have IANA refer all .ly requests to a new tld authority. Same problem - if you do that once, especially if you do it just to protect a company's profits/business model, how do you refuse a request from China to assume control of all of Tibet's tld?

Suggesting DNS level changes for bit.ly seems to me to open a whole can of very ugly worms...

[ryanwaggoner]

I was thinking more of things like OpenDNS and Google Public DNS, which I believe do malware filtering. I could be mistaken, though...this definitely isn't my area of expertise.