Thad Anderson of outragedmoderates.org is outraged at how the Bush Administration is trashing true American values. He’s set up his site to share government documents that reveal the misdeeds of the current American regime.
He also is providing the documents through peer-to-peer (p2p) networks more famously known for sharing music files. Ernest Miller, who is usually dead-right about most anything he writes on, criticizes Anderson for his use of p2p.
Miller’s complaint is basically that in this instance, since Anderson could just host all the documents on his site directly, using a p2p network makes no sense, and just adds an unnecessary layer of complexity. Miller goes so far as to say, “There are legitimate uses and needs for P2P. Particular functions where it makes sense. This isn’t one of them.” I think he’s dead wrong about that. Here’s why:
What follows is an e-mail I just sent to Anderson of outragedmoderates.org.
[Intro snipped]
Your idea to provide government documents via p2p networks is a good one, and one I have thought about myself. I have a suggestion for you.
If a project such as yours were to remain on a small-scale, then the use of a p2p network would just be a gimmick, for you could host versions of all the documents on your site, and people would not generally have doubts about their authenticity. There would be no need, in such a case, for introducing a p2p distribution system, because it also introduces the questions about document veracity that you address by suggesting people search for your username on the p2p networks.
But, in a large scale government documents project, the bandwidth and hosting space necessary for an enormous amount of government documents becomes more than any single individual, who’s only spent $220, can bear. This is where a p2p distribution mechanism would make an enormous difference. It would allow others to share your load. The problem, in such a case, to solve then, is the problem that end users have of verifying the veracity of the documents.
Fortunately, this is easily solved. Alongside each document that you list on your site, you should publish the md5sum “hash” or “fingerprint” of that file. This number can be generated by the md5sum program, which is available for every major operating system.
For instance, I downloaded the first document linked on your site, Halliburton Contract ‘coordinated w VP’s office’ Email which came in at 39,454 bytes and produced an md5sum fingerprint of: d22b6c8827d6a8437beab1bb66da03ef
You can check this yourself. To run md5sum on Windows, get to a dos prompt and just type ‘md5sum filename’ without the quotes and replacing ‘filename’ with the actual filename, 030503.pdf in this case. In Linux or Mac OS X, you use the same command from any shell/terminal prompt.
If you acted as a central repository of the md5 hashes of each file that you seed into the p2p network, then when you host a document on the p2p networks, the first wave of downloaders will get it from you, check the hash on your site against what md5sum tells them, and then they in turn can host verified copies of the documents on the p2p network, and the next person who comes along can get the file from either of you, check the hash against your site, and also know that the file remains unchanged. You just reduced your bandwidth needs and gained an army of helpers.
Also, by monitoring the p2p networks for verified copies of the files, you could at some point even stop “sharing” them yourself as plenty of other good versions would be out there.
Hopefully this makes sense. Feel free to lift from my explanations above if you decide to adopt this suggestion and want to explain it on your site. In any event, your up to something good! Keep it up.
So, the reason I say Miller is wrong to criticize this idea is that there are lots and lots of government documents and lots and lots of people who might want to download them. Asking one individual to host all those on a single web site is asking them to have a bandwidth-deathwish.
Instead, there is a valuable service someone could provide, perhaps outragedmoderates.org, of being a central repository of all the md5sum hashes for such documents. Then we only have to trust a single site not to alter the documents before creating the hash. After reading his “About Us” page, I think I trust Thad. I hope he, or someone, will adopt this suggestion.
This work, unless otherwise expressly stated, is licensed under a Creative Commons Attribution-Share Alike 3.0 United States License.
Place your comment