[likpok]

> If you want to talk about confusing, I watch C-SPAN constantly (it's an illness) and whenever anybody in the legislative or executive branch talks about "cyber security" they always talk about IP protection and "preventing a cyber pearl harbor" in the same breath.

The trouble is that the effective, worthwhile and highly damaging cyberattacks all involve IP, in some way or another. There's not much value in taking down Coca-Cola's internal network. Stealing their M&A strategies or product roadmaps can be extremely lucrative/damaging (I recall seeing estimates that billions have been lost as a result).

[Torgo]

No they don't. I think it is extremely confusing to talk about theft of data at the same time as talking about someone hacking a nuclear power plant to go into meltdown or something. When people say things like "cyber pearl harbor" at that time they could be talking about a DDOS that makes it impossible to do online banking or they could be talking about an attack on SCADA systems at a power plant that takes out power for a city. It really drives me nuts because either everybody in government talking about it is a poor thinker or they are intentionally being vague.

[tptacek]

I have no idea what this comment is even trying to articulate. You suggest two kinds of "cyber attacks", one which cause power plants to malfunction and the other that attacks online banking. I am not sure what you think this distinction demonstrates about online security.

On the one hand, the attacks on power plants that you allude to are possible. Utilities have been networked and electronically controlled since the 1970s. Nobody builds networks on telephony or X.25 anymore; it's all IP. IP connectivity to insanely sensitive systems leaks routinely; moreover, application-level data sharing between Internet-connected systems and supposedly air-gapped backend systems is extremely common.

On the other hand, the "less serious" attacks you allude to are very very bad. Google and Hotmail aren't national utilities. But they are attacked by state actors because dissident organizations use them to communicate. For that matter, the Internet backbone is a collection of computers sharing information using a decades-old routing protocol for which policy is controlled by regular expressions.

Finally, if you run a startup and happen to say something I disagree with, such as "I think CISPA is a power grab by the content industry", I could today very easily push you off the Internet with a trivial DDoS attack. The people who extorted online casinos with DDoS botnets were not rocket surgeons. When I attack you for disagreeing me online, and you call your ISP, guess what you're going to hear? "You're on your own". It is always very weird for me to see people on Hacker News, a hub for online startup news, downplaying the severity of DOS attacks. I've spent a decent chunk of my career in DOS mitigation and it is not remotely a solved problem.

[Torgo]

I think the government has a legitimate interest in protecting against computer attacks on public infrastructure that could result in death, and I see a place in there for government involvement. To a lesser degree there is a legitimate interest for government regarding IP theft. But I think how the government is involved and what powers they have, are different for these two scenarios. I understand that they overlap. CISPA is going to give government a much expanded jurisdiction and I don't think the restrictions are fine-grained enough.