Slashdot Mirror
Peer to Peer and Spam in the Internet
RobertDHaskins writes "A very interesting series of papers from Helsinki University of Technology on the topics of P2P and spam. Written by PhD students they are a little long, but some very good coverage of the state of the art."
...for those that don't wanna read the PDF:
Here.
libertarianswag.com
Just a head's up...
Here's what I do: Bitty Browser & Andromeda
- Follow the money
- Block networks who let spammers send traffic on them, no matter if it's SMTP, DNS, FTP or HTTP
Once a few big guys find themselves turned into intranets, they'll start paying attention.I don't have time to read a document hundreds of pages long, especially not one that's packed with information: I need a quick summary.
Could someone post a one line summary? For example,
Linux good; Microsoft bad; SCO evil; RMS god.
John.
From the paper: "The idea was to learn about the disruptive and also annoying phenomena that have become very commonplace over the past couple of years in the internet: namely, the Peer-to-Peer traffic and applications and the unsolicited and unwanted e-mail or Spam."
I think bundling p2p and spam is either totally missing the point, or attempting to influence the opinions of people who don't know better. The users of p2p want what they get for the most part (maybe not viruses and fakes, but the author seems to be targeting p2p due to the copyrighted content).
Its true that a bunch of computers can simulate a server for a game.
If you have 6 computers transfering information to each of them, you can create almost the same environment that 6 computers feeding off a server is.
If you place the anti-cheat code on every computer, you form a community to check against cheats.
If you also store every character's information on every computer, then you can watch for hacks there too.
Given its extrodinarily complicated, and fails to mob rule(conspiracy of hackers to overwhelm the system)... Its something that could be done.
I'm sure theres even more complicated things you can do with P2P, such as organizing nodes for filesharing and so on.
God spoke to me
Maybe I'm missing something, but isn't it curious that these papers from a Finn university are in english?
Treehugger? Treehugger... Treehugger!
In Soviet Russia.. Dissertation writes You!
The more keywords in the file name the lesser chance it will contain anything that makes sense.
In EDonkey it's worth looking at other file names of given share, they often offer some insight. You grab ROTK, check and see 3 other names: FOTR-Extended-Edition, and you may be sure it was some moron who can't tell "1" apart from "3" who renamed it and some more morons download it without checking.
Anagram("United States of America") == "Dine out, taste a Mac, fries"
I think it's worth mentioning this article talks about P2P, then about SPAM.
While it doesn't imply they are somehow related in their functions, the common nature of these two is the bandwidth consumption, which as stated by the author, can be annoying and disruptive.
Can we get the RIAA to shut these guys down?
Why isn't there a service where you can get full-speed from behind a firewall without portmapping? College students everywhere would rejoice. When I'm home I port forward and get the full pipe, but when I'm at college the firewall keeps my download speeds nice and slow. I know this because every once and a while I'll get lucky and some BT seed will connect and start sending me 80kb/s for about five minutes and stop. They made Supernodes to make the network more scalable and to make it work with firewalls. Can they make it work at full speed with firewalls?
Even those who arrange and design shrubberies are under considerable economic stress at this period in history.
The Finns have noticed that no-one understands Finnish, so they've become extremely good at putting things in more popular languages. For example you can get the news in Latin courtesy of Finnish Radio (today's headline: Kerry candidatus democratarum.)
For example, this is in the introduction to the Freenet section:
Um, many people might disagree with that little gem.It might solve the problem, but the social implications are horrific. I am assuming, despite the lack of an April 1st date, that this is a subtle joke?..
Alternatively, if the poster *is* serious, it is reminiscent of the old joke whose punchline is "you can't get there from here".
"The time is always now" - Victor
Still I'm pretty surprised that its already 2004 and the nets still pretty open. Whats to stop industry from rolling something like this out?
I've *never* been lured into sending someone else an e-mail. The only way I could imagine someone doing it is either (a) sending me an e-mail first, or (b) setting up a website with a "click here to e-mail me about ..." In case (a), if someone charged me, I would turn around and charge them for the original e-mail, and we would be even. Listservs could work on a similar basis: a small, one-stamp's-worth registration fee that is forfeited if you ever charge them.
In case (b), people would quickly learn not to click on links that they don't trust, which is a desirable outcome.
Malicious or vengeful charging is also a non-issue. Simply give each "charge" (I prefer to call it a "reusable stamp") a time limit (perhaps one day, or three); then, the amount of damage that any one person can do to another is small and easily caught. Once bitten, twice shy: if you charge me, then I won't e-mail you anymore, which is the desired outcome.
Enforcement is a harder problem. To do this, a third party would have to be trusted to hold small amounts of money from everyone (and larger amounts of money for listservs), and mail would have to be routed through that person. I forsee that putting a serious crimp into e-mail bandwidth. Also, the more centralized the money-holding system, the more tempting it will be for electronic thieves.
Notwithstanding, I think it is very much worth trying to solve this last problem. At this point, there are only two live options for dealing with Spam: either ICAAN and e-mail users decide to self-regulate, or else national governments will step in and regulate for us. Do you want a government to decide what is spam and what isn't?
Out of the self-regulatory options, I like the reusable stamps the best for three reasons:
(1) It gives the end-user the decision about what is spam
(2) Unlike filters, it places the cost of spamming on the sender instead of the receiver
(3) Also unlike filters, it results in reduced levels of spam being sent instead of increased levels of gibberish-coated spam being sent.
Regards,
Jeff Cagle
Human being (n.): A genetically human, genetically distinct, functioning organism.
The professor is clearly biased (or purposely acting biased) against P2P, lumping it together with spam as "parasitic and threaten[ing of] the purpose the Internet was designed for". How he figures sending files to one another is a subversion of the Internet's purpose, I dunno.
But the students' papers are all about how effective and efficient the various P2P architectures out there are and how they might be improved. Heh. Bless you, students.
"A great democracy must be progressive or it will soon cease to be a great democracy." --Theodore Roosevelt
And I'll post using the login New Here and say "No, I'm New Here."
And I'll keep thinking it's funny each time I do it.
to not let her computer get hijacked
Even if it isn't it'd never work. You just can't detect encrypted content. My JPEG file could contain an mp3 once the headers are stripped off, and who'd know the difference, short of parsing the whole file, looking at the result, and realizing that it's a mess.
This technique is already used to fool some web browsers (ie IE) into downloading a file as binary instead of getting it as text...
Lock up the current net, and another one will spring up underneath. It'd probably be kind of like freenet, except that it'd be implemented at a much lower level (cap the necessary technologies over your PPP or eth interface, and talk to that). Sure bandwidth would suffer, but the information would remain free, and uncontrollable.
Ask 8 slackers a question, get 10 awnsers (a citation, but I can't remember from who)
I'm using Cloudmark's Spamnet, which is essentially what they seem to be talking about (although I didn't read all 109 pages). Seems to work well enough. It tosses about 105 spam a day for me and I have about one or two slip through that I myself block.
(Argh, the cat got my tongue!)
I didn't think this would work originally, but this method sounds good!