The Digg effect is basically when many people, ususally from Digg, inadvertently perform a DDoS attack on a page they want to view. Why does it happen? Well, it's because the way we host web pages doesn't scale. Shouldn't it be obvious? Plain http cannot perform as well as we need it to. What other situations have we run into this with? Downloads of large files, perhaps? How did we fix that? 

We fixed that with p2p technologies like bittorent. And that is how we must fix the web. Now, you can't just throw a web page on a torrent and expect it to work; there are some challenges to overcome, and there are some limitations to take into account. First of all, any page that requires a secure connection won't work with this; if the page is unique to a viewer, this simply will not be of very much assistance; however -- any file on that page that is common among users will be able to take advantage of this idea. Secondly, you can't trust anyone, when it comes down to it. One must be able to verify that the data they received is what was orignally hosted on the server they wanted to access. This is part of why we have checksums. Thirdly, ISP's will hate us for doing this, because they won't be able to sell as much server bandwidth. They've pulled enough garbage that I don't really care what they think anymore. Finally, we don't have an implementation yet... and this is what I want to fix. The system that we need to create is as follows:

When a person goes to view a web page with a bittorent enabled browser (which doesn't exist yet afaik), the browser should download only a bunch of checksums and file tracker locations from that server. The browser should then begin all torrents and display the page. As long as that page is in the browser cache, the browser should seed the file torrents-- basically, it should help host. No one realized it yet, but that's what browser cache is for. Let us use it properly. Some people who have copyright policy that they must strictly adhere to will not want the files on their page being torrented, or the page itself, of course. There are two sides to this: The first is that it could throw off hit counts and download counts. When this system is implemented, it needs to take that into consideration. The second is that they may dislike p2p and equate it with people bound and determined to violate their intellectual property rights. Now, this is not what I use bittorent for; I used to, and there are many who do, but these days I try to stick to what's already Free. I get my music from Jamendo and my software from Ubuntu repositories. Yes, I'm one of those angels who only uses torrents for downloading Creative commons material and linux distros. I think you've heard of me :) Anyway, sites out there that are creative commons licensed could be great to test out a rough implementation of this idea; the difference in page hits would be tiny at first, and they have nothing against the redistribution of their hosted content. In fact, it is sites like that, as well as all of us internet users of course, that stand to benefit the most from this idea if it is widely put into practice -- or even just among those of us who inhabit these parts of the internet. The more people who see this idea, the more likely it is to find someone with both the ability and the will to code it into existence. If you are that person, please let me know so I can alpha/beta test whatever you cook up.

ethana2@gmail.com -- yes, that is my email address in plaintext. There are, sadly, few things I find more enthralling than camping my gmail inbox and playing whack-a-spam. Average time to click 'report spam' button: 7.3 seconds.