Shouldn't it be titled "Earth Warming More Quickly Than Thought, New Climate Models Predict ??

But the weird thing is, in order to make the buoy data match the engine water intake data, they decided to "warm" the buoy data rather than "cool" the intake data. That's hard to understand. They could have split the difference, but no. It's all a warming adjustment which makes skeptics wonder why that was chosen when the best guess is that the intake data was too warm, not the buoy data too cool.

If the content network is directly connected to the eyeball network, your analysis is correct; both ISP are being paid by their respective customers to move data their customers both want. But eyeballs are all over the place, and there is often a third network in between that is actually moving the bits the furthest. Those peering relationships are where the problem is.

This is BS. Spamming over IPv6 is not easier. First of all, very few servers can even receive spam (or any email) over IPv6 yet. Second, identifying the source IP is a whole lot easier in IPv6 with its unfragmented IP allocations and a radically smaller routing table. Spam havens would stick out like a sore thumb. Yes, I've gotten my share of spam from HE netspace, but Cogent is responsible for quite a bit too. This isn't about Cogent not wanting its lily-white IPv6 network dirtied with HE packets. Its about Cogent using its tenuous Tier 1 status in the IPv4 world to try to usurp HE's Tier 1 status in the IPv6 arena.

Actually, the IP address is about the only piece of information available in the SMTP dialog that is very difficult to forge, because it is the source IP of the TCP session. It may belong to a trojaned box, but it isn't "forged".