p2pnet - rss feed: Viralg anti-p2p: Part II
p2pnet.net News:- Yesterday, we reported that Finland’s Viralg “guarantees” to stop file sharing in its tracks.
At the time, we were going on what we saw on the site but Benedek Toth told us about Viralg’s press release. You can read it here, should you so desire. But the headline says it all:
VIRALG: End of illegal peer to peer file sharing
Toth isn’t impressed.
Neither is Ed Felten.
"The finnish company Viralg claiming that their new method against p2p networks is able to fake or manipulate data in p2p networks, even if the users are using a verification site," says Toth in findhash.com, going on, "In their press release they are naming findhash.com as an example.
"Findhash is one of the oldests of its kind, and gained the title ‘trusted’ by many of our users, simply because we’d never use or recommend any technology, method or software which can be corrupted by a third party.
"Most client uses at least one, but usually two different checksum to verify the downloaded data. These methods are developed by real professional organisations (such NSA, Berkeley etc) and often used in other secure transfers, such as your communication with your bank.
"The use of two different algorithm, a secure and a tree-based can provide a very good corruption-handling to the filesharing client, so even if they trying to alter the data, you only have to redownload a smaller part of the file. Most of them do this for you automatically.
“No matter what they claim to be able to do or try to do, the laws of mathematics cannot be changed by them. And these laws tells us that, even using the whole computing capacity of mankind, they’d need millions of years to fake the hash of one movie file.”
‘It’s not the hashes …’
Felten took the trouble to examine the Viralg site in some detail, and here’s what he has to say in Freedom to Tinker
“This shows all the signs of being a scam or hoax. The company’s website offers virtually nothing beyond claims to be able to totally eradicate file swapping of targeted files. The ‘Company’ page has no information about the company or who works for it. The ‘Customers’ page does not mention any specific customers. The ‘Testimonials’ page has no actual testimonials from customers or anybody else. The ‘Services’ page refers to independent testing but gives no information about who did the testing or what specifically they found. The ‘Contacts’ page lists only an email address. There is no description of the company’s technology, except to say that it is a "virtual algorithm", whatever that means. Neither the website nor the Viralg press release nor any of the press coverage mentions the name of any person affiliated with Viralg. The press release uses nonsense technobabble like ‘super randomized corruption’.
“The only real technical information available is in a patent application from Viralg, which describes standard, well-known methods for spoofing content in Kazaa and other filesharing networks. If this is the Viralg technology, it certainly doesn’t provide what the website and press release claim.
“My strong suspicion is that the headline on the Slashdot story – ‘Finnish Firm Claims Fake P2P Hash Technology’ — is correct. But it’s not the hashes that look fake, it’s the technology.”
Something you think we should know? tips[at]p2pnet.net
See:-
stop file sharing - ‘Guaranteed’ anti-file sharing, p2pnet, April 18, 2005
findhash.com - Response to Viralg Oy, April 19, 2005
Freedom to Tinker - Why Does Anybody Believe Viralg?, April 19, 2005
April 20th, 2005 at 8:16 am
I bet the cryptographers for HD-DVD’s will make the same assertion, but there is always that workaround nobody thought of.. a novel new approach.
I doubt however that it’d require more than a simple upgrade to p2p apps in general to quell it.
But this does give me insight into how important my more advanced data structures courses are. Had no idea those headache-inducing things were used as often as in an everyday p2p app, I’ll have to really ingrain those in my memory or keep supplementary notes for later reference.