I am not a privacy evangelist, but I don’t like deceptions.

I’d say the biggest threat to someone being able to view your personal search history is the government (or a court of law). By removing that part of the IP, you cannot identify a specific user. Sure you can draw conclusions and probably come up with a pretty good guess – but nothing that would pass in court, or likely as intelligence data.

What about, oh, I don’t know, the fact that everything you do on the internet gets routed through the public “air”. The internet is not private and secure.

You know, the treatment from Google is just so unfair. Some product pages get high profile links with targeted anchor text and some blogs just get angry “you mislead people” posts.

I guess you fall somewhere in the middle.

Remember: Only you can prevent Google from getting PageRank. Use NOFOLLOW on all links to all Google properties!

Almost the same as if the full IP was provided and then of course who the provider of the IP is. If the class-C subnet shows that the user comes from AOL, then the class-c information are as good as the full IP (or not good in this scenario).

Different story when it comes to Class-C subnets that belong to smaller companies with not too many people accessing the internet.

Carsten it’s interesting that earlier in the video they explain how at most they could identify the ISP and company you are using so when they do drop the class D part of the IP they only drop the part that didn’t provide any meaningful information in the first place.

That means that there are only 255 possible IP-Addresses the user could have had and the Class-C Subnet information, which they keep, are enough to know the ISP or Company the user was using.

From the Cookie was also not much removed, except for the preferences. What they remove from the logs is nothing and only good for PR, no more and no less.